www.filterforge.com - Creating Filters

Need programmer for bomber

Ouroboros — Sat, 18 May 2013 05:30:47 -0400

Ouroboros, Wed, 15 May 2013 06:22:44 -0400
Hi!
I need to modify bomber, I have no idea if this is possible:

-Make the bomber choose the color of the particle from only one pixel
-Input a second image that determines the angle of the particles

We can work through elance or whatever, thanks

Paul

CFandM, Wed, 15 May 2013 07:53:35 -0400
Well you can't modify the bomber itself but you add other components to it...This will result in final task that you need.....btw what is the final task that you hope to achieve?.....You can contact me through the email button in this post... ;)

jitspoe, Fri, 17 May 2013 00:43:56 -0400
What you're trying to do may already be possible. I'm not sure what you mean by "only one pixel", though. Do you want the first pixel in the image to determine the color of every particle? Through some use of other components (ex: scale and tiles), you could probably blow the entire image up into a solid color from one pixel, then use that to color the particles or pass it into the tint.

For the second part, you can already pass in a component to determine particle rotation, so just pass an image into that...

Betis, Sat, 18 May 2013 05:30:47 -0400
Tint Color and Rotation are the inputs you're looking for! Just plug something into them, that's how FF is supposed to work! :-p

Fractals by Bomber

ThreeDee — Wed, 15 May 2013 03:59:22 -0400

ThreeDee, Sat, 11 Aug 2012 02:47:00 -0400
One can make simple fractals with the Bomber component.

ThreeDee, Sat, 11 Aug 2012 02:48:03 -0400
Changing the Particle will give variations on the theme.

ThreeDee, Sat, 11 Aug 2012 03:13:53 -0400
The "Inverse" kind is also possible.

SpaceRay, Sat, 11 Aug 2012 03:15:16 -0400
WOW!!! We have been working both on the same thing :D

I have just also made also a Fractal images filter based on bomber

ThreeDee, Sat, 11 Aug 2012 03:16:12 -0400
Here are the Bomber settings that are needed to achieve this.

Note especially Layer order "Big on top".

ThreeDee, Sat, 11 Aug 2012 03:18:47 -0400
Hi SpaceRay,

Yup, that's the way to do it!

The other important thing is to feed a particle with transparent background so the particles on top don't completely cover up the ones below. (As you have obviously done.)

SpaceRay, Sat, 11 Aug 2012 03:19:40 -0400
Beautiful examples you have done, and like your version that gives those results.

Is this above made with just ONE bomber with the setting you have shown ?

ThreeDee, Sat, 11 Aug 2012 03:23:42 -0400
Since the Bomber layering works by dividing the image area (repeatedly) into four quadrants, the most obvious choices for particles are things that align to that pattern.

ThreeDee, Sat, 11 Aug 2012 03:26:36 -0400
Yes, it is just one bomber. In fact, the filter for the top image is just 3 components (plus Result) total!

Here is the filter with various particle types.

ThreeDee, Sat, 11 Aug 2012 03:33:36 -0400
When you spherize the above, you get this fractal spherical tiling.

ThreeDee, Sat, 11 Aug 2012 03:34:54 -0400
All fashion of interesting things can be made with it.

SpaceRay, Sat, 11 Aug 2012 03:40:33 -0400

Quote
ThreeDee wrote: The other important thing is to feed a particle with transparent background so the particles on top don't completely cover up the ones below. (As you have obviously done.)

YES, you are right, this is very important to get this effect, because if not is difficult to see the effect and is not the same at all. And yes, I have used the transparency lifesaver :)

Quote
ThreeDee wrote: Yes, it is just one bomber. In fact, the filter for the top image is just 3 components (plus Result) total!

My filter is more complex, made from 4 bombers as I want to be able to mix shapes, images, colors and textures in a fractal way.

This is an example using two of the bombers (you can choose many and how you want to use them)

SpaceRay, Sat, 11 Aug 2012 03:41:39 -0400

Quote
ThreeDee wrote: Here is the filter with various particle types.

Thanks very much for sharing your own version.

Sorry that yet I can´t upload my version as is not good to show it and is yet in beta stage and does not look as good as I want.

Quote
ThreeDee wrote: When you spherize the above, you get this fractal spherical tiling.

WOW!! I have not thought that this could be done this way, this looks awesome and fantastic, great and beautiful examples.

ThreeDee, Sat, 11 Aug 2012 03:42:03 -0400
:D

SpaceRay, Sat, 11 Aug 2012 04:07:40 -0400
Cool and nice fractal multi tree like shown in your last example. Like it.

I have tried your uploaded filter above and love the way you have done it, and like also all the presets showing well what is possible AND LOVE MUCH TOO all the great collection of beveled shapes you have put inside this filter to show as fractal examples. This shapes are very useful also for using them on others filters and to to try them in others ways.

This is VERY useful and interesting, thanks very much for sharing it here

Here I have added a color control to the bomber tint color

ThreeDee, Sat, 11 Aug 2012 04:21:47 -0400
;)

Ghislaine, Sat, 11 Aug 2012 14:22:10 -0400
:)

Ghislaine, Sat, 11 Aug 2012 14:28:04 -0400
;)

Sharandra, Sat, 11 Aug 2012 15:11:09 -0400
cool, nice examples :-)

SpaceRay, Sat, 11 Aug 2012 15:47:08 -0400
I am making an update to your great and well done filter so you can have also my idea of fractal images (like the ones above with the lifesaver) and also 3 more shapes (16 in total using Totte´s Megaswitch) and also add some more controls.

I will upload it when I have it ready

Skybase, Sun, 12 Aug 2012 00:29:19 -0400
That's lovely :D

SpaceRay, Sun, 12 Aug 2012 00:59:06 -0400
This is the text included inside the filter description

Quote
Here you can make a fractal effect multiplying your image source or any of the other 15 shapes (particles) available to choose Threedee has started to make this filter and I (SpaceRay) have upd ate it with 3 more shapes and the ability to use source images and added the controls settings to use this filter much better and have more options I have also created 60 presets for you to use: 1 - the first black and white part are the shapes with 20 ITERATIONS 2- The second part that is colored is using the tint color and also 20 ITERATIONS 3 -The third part that is black and white is using 10 ITERATIONS 4 - The others are modifying the Offset and rotation se ttings with 10 ITERATIONS Also want to give my thanks to: Totte for his well done Megaswitch component for 16 switchs Tim2501 for the Sierpinski Prism filter as I could understand and modify better this filter thanks to his filter.

Quote

Here you can make a fractal effect multiplying your image source or any of the other 15 shapes (particles) available to choose

Threedee has started to make this filter and I (SpaceRay) have upd ate it with 3 more shapes and the ability to use source images and added the controls settings to use this filter much better and have more options

I have also created 60 presets for you to use:

1 - the first black and white part are the shapes with 20 ITERATIONS

2- The second part that is colored is using the tint color and also 20 ITERATIONS

3 -The third part that is black and white is using 10 ITERATIONS

4 - The others are modifying the Offset and rotation se ttings with 10 ITERATIONS

Also want to give my thanks to:

Totte for his well done Megaswitch component for 16 switchs

Tim2501 for the Sierpinski Prism filter as I could understand and modify better this filter thanks to his filter.

Here is the filter

Morgantao, Mon, 13 Aug 2012 14:40:16 -0400
Very cool, you two!

Thank you both for sharing those fractals. One can never have too many fractals 8)

SpaceRay, Tue, 14 Aug 2012 08:32:15 -0400
Thanks Morgantao

Here I have made another different version, is the same as above BUT the main difference is that NOW YOU CAN USE A MASK to render the result so the fractals will be laid around the borders of the mask instead of filling the whole screen, and the 60 presets has been changed to show this mask.

All the presets are using the Lifesaver with transparency mask, but you can use of course your own mask loaded through any image with transparency you may have.

I have put it in the lower part all what is related to the new masking

CFandM, Wed, 15 Aug 2012 04:53:43 -0400

Quote
ThreeDee wrote One can make simple fractals with the Bomber component.

This is one of those methods where you slap yourself on the forehead and say...Geez..It was right there all along... :) Good one ThreeDee..

Morgantao, Sat, 18 Aug 2012 10:41:07 -0400
Hey, ThreeDee, how did you make this one?

I didn't see it in the filter... :?:

ThreeDee, Sat, 18 Aug 2012 11:02:54 -0400

Quote
Morgantao wrote: Hey, ThreeDee, how did you make this one?

Make an image with the two gradients at the top half and transparent in the bottom half. Otherwise same method.

Morgantao, Sat, 18 Aug 2012 11:58:31 -0400
Ah, that was easy :)
Thanks!

ThreeDee, Sat, 18 Aug 2012 12:04:51 -0400
Fractal Tree by Bomber

Strategically places transparency and opaque areas make for interesting results with the simplest of constructions.

ThreeDee, Sat, 18 Aug 2012 12:06:24 -0400
It is really as simple as this.

ThreeDee, Sat, 18 Aug 2012 12:23:13 -0400
Similarly:

Ghislaine, Sat, 18 Aug 2012 13:59:19 -0400
BIG THANK 3D ;)

Morgantao, Sat, 18 Aug 2012 15:46:33 -0400

Quote
It is really as simple as this.

Exactly how I did it after your wonderful explanation on how to do the square gradients :)

SpaceRay, Sat, 18 Aug 2012 21:41:26 -0400

Quote
ThreeDee wrote: It is really as simple as this.

Thanks very much for showing this and is interesting and useful

Also using your shape done in this filter I have taken it to the other filter above and I have got this :)

SpaceRay, Sat, 18 Aug 2012 21:42:16 -0400
Changing the white to images with a color control through the Threshold

SpaceRay, Sat, 18 Aug 2012 21:43:39 -0400
and now changing the black to images with another color control through the threshold

SpaceRay, Sat, 18 Aug 2012 22:04:12 -0400
Or changing the circle for a circular saw shape in another configuration

SpaceRay, Sat, 18 Aug 2012 22:11:32 -0400
Or using a heart with threshold to make a frame for the source image

SpaceRay, Sat, 18 Aug 2012 23:16:53 -0400

What was missing on this one to really make a difference and so they look more separated is some drop shadows and added also shadows to the lifesaver itself, so I have modified the filter to have this two kind of shadows applied to the source image.

I think looks much better than the above one here

SpaceRay, Sat, 18 Aug 2012 23:27:53 -0400
In the same way this one

This one above is too much plain and can´t really see well the different elements, so let´s give shadows that make it pop

Which one looks better from this two ? ;)

SpaceRay, Sat, 18 Aug 2012 23:35:22 -0400
Instead of the Lifesaver I have put a sunflower

SpaceRay, Sat, 18 Aug 2012 23:49:01 -0400
Here is another one using a flower with transparent background so you can see through the hole of the round frame and does not fill it all as in the previous examples

Although I need still to fix the color of the shadow when using this transparent background images as it does not look good (it´s gray instead of black)

Morgantao, Sun, 19 Aug 2012 01:25:00 -0400
Way to go SR, the drop shadow version look much deeper and better indeed :)

SpaceRay, Sun, 19 Aug 2012 05:48:55 -0400
Thanks Morgantao, I was going to upload it here BUT have found an error on the shadows when the border is white :?: and also if the iiner image has transparent background.
I will upload this new version when I have it fixed.

This has tow kind of different shadows to make it look better, inner shadows and outer shadows (like drop shadows) that are all around.

SpaceRay, Wed, 15 May 2013 03:58:13 -0400
I forgot about this filter and now with the Circle Fractal example by CFandM in this thread I remembered it

I have not been able to solve the problem with the shadow that sometimes does not work right so I am uploading the filter as it is now, if someone wants to use it.

SpaceRay, Wed, 15 May 2013 03:59:22 -0400
and here is preset 3 I have done for the filter without any external image as shown on many of the examples shown on the previous page

Note:

As I have just seen that this post is in a new page, so for the filter to make this look at the last post of the page 1

PlanetGen 4

TBH-1138 — Sun, 12 May 2013 20:04:57 -0400

TBH-1138, Tue, 30 Apr 2013 22:47:51 -0400
I haven't gotten clouds or planet rendering done yet, so here's a preview of the default surface.

Skybase, Tue, 30 Apr 2013 23:34:51 -0400
That's verryy nice! :D

Betis, Wed, 01 May 2013 05:54:58 -0400
don't forget to adjust colors/snow based on elevation! looks awesome!

TBH-1138, Wed, 01 May 2013 16:24:36 -0400

Quote
don't forget to adjust colors/snow based on elevation!

I'd do that, but I'd have to add mountain ranges first. As-is, this kind of noise generates mountains that look like someone splattered white paint on the terrain.

TBH-1138, Thu, 02 May 2013 13:03:57 -0400
Mountains.

TBH-1138, Thu, 02 May 2013 15:52:01 -0400
Now it renders the planet. There are clouds, but I've disabled them for this image.

TBH-1138, Thu, 02 May 2013 22:19:27 -0400
Here it is with clouds. I'm not quite happy with them yet. The smaller details are too fibrous, but I'm not quite sure how to fix it.

Indigo Ray, Fri, 03 May 2013 21:42:36 -0400
Glad Earth doesn't have all those giant craters!

Have you tried using my vortex filter to help with the clouds?
Cover the planet in hurricanes! :D

TBH-1138, Sat, 04 May 2013 20:43:23 -0400
I actually was planning on using that for something, but the vortices just don't really look natural. It's either concentric circles or "walls" between each vortex. Another noise distortion might fix that.

TBH-1138, Sat, 04 May 2013 23:20:20 -0400
More noise distortion didn't quite break up the artifacts. I'm thinking of adding city lights or possibly glowing water for lava planets and such.

TBH-1138, Sun, 05 May 2013 01:41:19 -0400
Speaking of glowing water, here it is. This will be disabled by default. Currently the glow is rendered over the planet; clouds don't block it.

TBH-1138, Sun, 05 May 2013 04:44:08 -0400
Lava planet made with glowing water.

TBH-1138, Sun, 05 May 2013 18:12:54 -0400
City lights. Should I add anything else or release as-is?

Betis, Sun, 05 May 2013 20:02:16 -0400
I would refine the clouds before submission, and probably work on the lighting too.

Use some real-references

Notice the heavy blue cast from the atmosphere, the planets true colors are scattered and absorbed and replaced with the atmospheric glow. Also the water and shadows don't get very dark.

Also you might want some real lighting phases in there (I have a filter just for that)
http://www.filterforge.com/filters/8845.html

But this is all constructive criticism, I think it's coming along nicely! :)

TBH-1138, Tue, 07 May 2013 18:15:17 -0400

Quote
Notice the heavy blue cast from the atmosphere, the planets true colors are scattered and absorbed and replaced with the atmospheric glow.

I see what you mean.

Quote
Also the water and shadows don't get very dark.

In this picture, the light is coming from behind the camera. There are no shadows in that kind of lighting.

As you can see here, or by going outside at night, the dark side of a planet is pitch black.

Quote
Also you might want some real lighting phases in there (I have a filter just for that) http://www.filterforge.com/filters/8845.html

That's awesome. The only reason I'm not using that is because I can't generate specular highlights with it.

Betis, Tue, 07 May 2013 18:23:08 -0400
you might be able to get specular highlights by messing with the curve... I'm not sure I haven't tried it for that. but also notice that the photographs hardly appear to have any specular anyway

CorvusCroax, Wed, 08 May 2013 04:16:13 -0400
Those are great looking planets TBH-1138. You should totally release it!

TBH-1138, Wed, 08 May 2013 21:33:57 -0400

Quote
you might be able to get specular highlights by messing with the curve... I'm not sure I haven't tried it for that.

Nope. I tried everything I could think of and nothing worked.

Quote
but also notice that the photographs hardly appear to have any specular anyway

It's a subtle effect, but it definitely enhances realism. It all looks like solid rock without specular.

Quote
Those are great looking planets TBH-1138. You should totally release it!

I would, but the clouds still need some work and I'm rendering some 4k planetary noise from an edited version of ThreeDee's snippet. It's been rendering for ~1.5 days so far.

TBH-1138, Sat, 11 May 2013 15:54:45 -0400
Made some lighting and atmosphere improvements.

TBH-1138, Sat, 11 May 2013 20:47:52 -0400
I think I'll call this done. Here's a planet from the randomizer.

ddaydreams, Sun, 12 May 2013 11:21:49 -0400
Cool! I hope you left the glowing water feature in there. I liked that.

TBH-1138, Sun, 12 May 2013 20:04:57 -0400
It's still in there. :) One of the presets uses it for lava, and another (above) uses it for some kind of bioluminescence in the water.

Let the FF Games begin

ThreeDee — Mon, 06 May 2013 08:07:23 -0400

ThreeDee, Mon, 06 May 2013 05:50:38 -0400
There are a couple of good-looking examples of FF games made in the past, notably Meyendlesss's Nine Men's Morris. Render speed makes it a bit hard to use for actually playing, but it is a nice looking game.

While real-time games need other programming (as in Redcap's Games made with FF), there are a number of things that can be made and played right in FF.

In the hopes for getting things going, I made a faster version of the basic dice that can be "thrown" by clicking the Next Variant button.

Here's the file, it will also be in the Library under Snippets.

ThreeDee, Mon, 06 May 2013 05:52:41 -0400
Games where you advance in a certain predetermined direction -- or freely -- on a game board, are good candidates.

ThreeDee, Mon, 06 May 2013 07:52:11 -0400
Random game board generation would definitely be an FF-oriented feature.

ThreeDee, Mon, 06 May 2013 08:07:23 -0400
Which makes me consider that it would be useful to have two modes in the filter: game board generation and game play modes. Thus one could generate and save the game board in "Generate" mode and then switch to "Play" mode, load the board and play. In such fashion the filter wouldn't have to redraw the board on every move.

Other Interesting Ideas

ThreeDee — Wed, 01 May 2013 15:57:19 -0400

ThreeDee, Fri, 10 Oct 2008 02:51:05 -0400
Seems like I have more ideas than I'll ever have time to try, so I better write them down somewhere in case someone else has use for them.

Multi-pass filters

These are filters that have two or more "modes" (switched by a checkbox or selector), that first generate a starting image or an intermediate image and on the next application generate either the next step or the final image.

Example: Etch-A-Sketch. First application "shake mode" would draw the Etch-A-Sketch and one line. Subsequent application in "draw mode" would draw one more line each time without altering the image otherwise, just adding a line each time. Line beginning and end would be controlled by sliders.

Storing and retrieving information from image

For multi-pass filters, you might want to be able to store information that will be used in the next pass. Using the same Etch-A-Sketch example, you might want to store the x and y coordinates of where the last line ended so you can use it as the beginning coordinates of the next line in the next pass.

There are a couple of ways of doing this. They are based on embedding a couple of pixels that represent the number (0=black, 100=white). A rather neat way would be to use an area that is transparent, so as to hide the pixels from view altogether. For instance, in the Etch-A-Sketch filter, the four corners are transparent. You could store the values there in the RGB channels and retrieve them from there.

As this may not be enough of an explanation, the method to get the stored numbers back out and into use involves enlarging those pixels to cover the entire image area. Then you can use them in any green component input.

Other uses for transparency

Or you can store a whole bunch of information temporarily as transparency. Let's say you want to modify a height map over a number of passes of a filter. You could do a Set Alpha to temporarily use the transparency of the image to store the height map and Get Alpha in the next pass to retrieve it. Since FF does not remove any of the RGB information during this operation, it is non-destructive and can be used as many times as you want.

Boolean Logic

The easiest way to implement "IF ____( AND/OR/XOR ____) THEN" functionality into a filter is to use black and white to represent FALSE and TRUE. Then you can do blends to make boolean logic decisions. Blend modes work as follows:

Maximum = OR
Minimum = AND
Difference = XOR

Then you can use the result as opacity in another blend.

Okay, that may be difficult to comprehend. Here's two examples:

Say you only want to flip an image IF the center pixel is lighter than 50%. Okay, you would enlarge the center pixel to cover the entire image and threshold it at 50. Then you would use the outcome (which is basically TRUE or FALSE) as the opacity in a blend in which Background is flipped and Foreground isn't.

Follow?

Second example: You only want to flip an image IF the center pixel is lighter than 50% AND the transparency of the center pixel is less than 75. In this case, you would do the same with the threshold. Then you would do a similar operation on the transparency (Get alpha, enlarge to cover whole image, threshold at 75). Now you have two outputs, one which tells you whether the center pixel is lighter than 50% (white or black for TRUE or FALSE) and another that tells you whether its transparency is less than 75 (TRUE or FALSE). You want them both to be true, so you blend them with minimum (boolean AND operation) and get TRUE or FALSE. Then feed this into the opacity of the flipped/non-flipped blend.

If you read the above many times enough and try it, I'm sure it will eventually make sense.

Kraellin, Fri, 10 Oct 2008 12:59:23 -0400

Quote
Multi-pass filters These are filters that have two or more "modes" (switched by a checkbox or selector), that first generate a starting image or an intermediate image and on the next application generate either the next step or the final image. Example: Etch-A-Sketch. First application "shake mode" would draw the Etch-A-Sketch and one line. Subsequent application in "draw mode" would draw one more line each time without altering the image otherwise, just adding a line each time. Line beginning and end would be controlled by sliders.

Quote

Multi-pass filters

These are filters that have two or more "modes" (switched by a checkbox or selector), that first generate a starting image or an intermediate image and on the next application generate either the next step or the final image.

Example: Etch-A-Sketch. First application "shake mode" would draw the Etch-A-Sketch and one line. Subsequent application in "draw mode" would draw one more line each time without altering the image otherwise, just adding a line each time. Line beginning and end would be controlled by sliders.

i like this one. it's similar to something i asked for once, looping. goes well with batch processing.

one of the things talked about in the wish lists was scripting. i would think the storing and retreiving ideas might fit with this, also some or all of the boolean. and yes, i've love some decision components. i miss the if-thens.

and i also got some ideas from some of your explanations of black and white uses. you really could make decision legs off that... i think :)

ThreeDee, Fri, 10 Oct 2008 15:26:27 -0400
[Sorry, useless blabber was here -- I need to sleep more]

StevieJ, Fri, 10 Oct 2008 16:34:40 -0400

Quote
ThreeDee wrote: useless blabber was here

That sounds like most of my posts..... :| ..... :dgrin: LOL....

Excellent ideas!!! :devil:

ThreeDee, Sat, 11 Oct 2008 04:59:52 -0400
Boolean by depth

Not altogether a new idea, but I haven't seen it clearly pointed out elsewhere and definitely worth a special mention:

That's how I'll be constructing my 3D teapot...

Here's how it works:

ThreeDee, Sat, 11 Oct 2008 05:34:36 -0400
Negative boolean by depth

This one may be something new.

You can do boolean with any two (or more) height maps, it is in no way limited to just spheres. As you can see, you can have the color/texture defined for each component.

Here's the negative boolean technique (The curves math in the beginning is not part of the boolean, it's there to make the smaller spheres height map):

ThreeDee, Sat, 11 Oct 2008 06:01:01 -0400
Boolean Intersection

I just had to try it...

ThreeDee, Sat, 11 Oct 2008 06:11:32 -0400
Damn, this is getting rather exciting, if I dare say so myself!

Kraellin, Sat, 11 Oct 2008 10:16:09 -0400
way the freak above my head. lol. but i still like looking :)

StevieJ, Sat, 11 Oct 2008 11:42:50 -0400
Very interesting.....you'll make a 3D program out of FF yet!!! :devil:

Craig, Booleans are a simple 3D concept buy which two or more solid forms can be added, subtracted, or differenced with each other. Pretty easy to learn.....and part of almost all 3D programs out there. You would definitely encounter it if you ever got in something like TrueSpace..... ;) :)

ThreeDee, Sat, 11 Oct 2008 15:54:57 -0400

Quote
Kraellin wrote: way the freak above my head. lol. but i still like looking

"Boolean" is just a fancy mathematical word, the idea is really simple. When it comes to 3D objects, they are basically just three different things:

1) Combine objects together.

2) Subtract object(s) from another.

3) Use the space that is in common with two or more objects as new object(s).

Then you can call them "boolean operations" to make it sound real fancy. (George Boole was the mathematician that described these types of operations)

Kraellin, Sat, 11 Oct 2008 18:29:47 -0400
oh, i know what boolean is as far as computers and coding is, i'd just never seen it applied to imaging before.

ThreeDee, Sun, 12 Oct 2008 12:05:33 -0400
Saving multiple alpha channels in one file

As to what you would use them for, I leave it to your imagination. (Think for instance bump, transparency, diffusion, ambience, specularity, metallicity and reflectivity maps. Or making selections for separate objects in the image, rocks separate from sand, grass, twigs, etc.) Here's the theory.

You can of course create a filter that creates an alpha channel, say with a checkbox and a switch, and save it as an image and then later load it as a selection. Not a big deal. But what if you wanted to save more than one selection into one file?

You could save three alpha channels pretty easily by making each of them one of the RGB channels.

But for even more capacity, take a look at this: Filter Forge can save lossless images with 3 channels, 16 bits per channel. If you use 8 bits for full quality anti-aliasing, you can theoretically still have 3x8 = 24 "alpha channels" at photoshop quality for every image. But how would you store them in a single image?

By the upper 8 bits!

Okay, that is probably too technical an explanation, so let me rephrase that in other terms.

You could save one selection by making the alpha channel white 0/256 black and the alpha channel black 1/256 black in one of the RGB channels. You would have used one bit of image information.

Then you can save your second alpha channel in the same RGB channel by making white 0/256 black and black 2/256 black. When you add that to the existing image, you will have used 2 bits of image information. Yet the two alpha channels remain separatable out, for they fall in distinguishable black ranges.

The next alpha channel would be added to it with black at 4/256, the next one with black at 8/256, 16/256, 32/256, 64/256 and 128/256. At this point you would have used up one of the RGB channels storage capacity.

Thus, if you save the image in 16-bit format, you should be able to store up to 24 alpha channels in one image this way.

Separating them back out is a bit of a subject, too, but I'll return to it eventually, or if someone wants to try this.

Sounds good in theory, it'll have to be tested to see how it works out.

ThreeDee, Sun, 12 Oct 2008 12:22:13 -0400
Object construction by multi-pass filters

This may be my next thing to try.

The boolean method can be used over and over, adding new pieces to construct an object in several passes. You will have to store the bump information (and, for fancier applications, other information) separately from the RGB channels. And you will have to have a separate surface-type filter to generate the final shaded object.

How would that work?

Say you did that same image I made the boolean example for. One big purple sphere and four smaller green spheres.

You would have a filter that would output a spherical height map and a colored circle of the same size. The filter would have controls for radius and x, y and z coordinates of the sphere. So you would make a big purple sphere in the center. Then you would save the output, making the height map either a separate file (with a checkbox and switch) or make it the transparency channel of the image.

With this new image loaded back in (or filter applied and FF evoked again, if you are using it as a plug-in), you would create a smaller green circle in the top right corner and choose one of the boolean operations (add, subtract or intersect) and save that image and the new height map, separately or together.

Then do the same for the remaining three green spheres in the other corners.

(You could continue this with spheres of different sizes, colors and boolean operations infinitely as long as you saved the image in 16-bit format.)

In the end you would call up a surface filter that would use the height map and create a dimensional object out of it (that looked like one of the boolean examples above).

ThreeDee, Sun, 12 Oct 2008 12:45:30 -0400
Separating objects by index number

Another idea for multi-stage image construction. Say you want to "build" and object but haven't decided what textures to apply to the different parts. How about assigning an index number to each of the parts and deciding later what texture to apply to each? It's like having separate modeling and texturing steps that anyone who has done 3D modeling is familiar with.

And how that would be possible is this:

With each separate "object" that is created, the filter can create a grayscale image that matches the objects with a certain flat gray value (again, checkbox-and-switch approach would work for rendering this out separately). I better make a picture here to show you what I mean. I'll use the same boolean image example.

Here is the diffuse color:

Here is the height/bump map:

And here is how the index channel could look:

You see how that could be used later? You could retrieve any one object out of that index image and assign a texture to it, then another object, assign a texture to it, etc. This could be done for example with the multiple-image-input workaround method.

But this would allow you to add, change and experiment with different textures altogether separately from object construction.

---

PS. The index channel idea could, BTW, be used as a simplified multi-selection storage method. It is limited in that you can't have overlapping selections, but could be useful for some purposes.

Sphinx., Sun, 12 Oct 2008 12:54:43 -0400

Quote
ThreeDee wrote: Then you can save your second alpha channel in the same RGB channel by making white 0/256 black and black 2/256 black. When you add that to the existing image, you will have used 2 bits of image information. Yet the two alpha channels remain separatable out, for they fall in distinguishable black ranges.

I think that will be very difficult, if not impossible since FF internally works in floating point mode, you can't work with packed composites as you do with 32 bit unsigned ints. If you work out a way to do this I'd be very interested in the approach ;)

Task: pack two map inputs into one channel by only reducing bitdepth and not samplerate.. tough one!

ThreeDee, Sun, 12 Oct 2008 13:55:04 -0400
Custom shapes library

Thought of another use for saving multiple alpha channels in one image. You could make a custom shapes library that way. Think about it: 24 custom shapes in one file!

Didn't InfiniView or someone talk about some such thing?

Man, now I really want to try that idea out!

ThreeDee, Sun, 12 Oct 2008 13:57:40 -0400

Quote
Sphinx. wrote: I think that will be very difficult, if not impossible since FF internally works in floating point mode, you can't work with packed composites as you do with 32 bit unsigned ints. If you work out a way to do this I'd be very interested in the approach Wink Task: pack two map inputs into one channel by only reducing bitdepth and not samplerate.. tough one!

True, I might be wrong about that since I haven't tried it, but I will be doing so soon enough.

ThreeDee, Sun, 12 Oct 2008 14:36:10 -0400
Okay, wasn't quite as easy as that, and probably will only work for 2-bit images, for you get the residue from the antialiasing of the other channel(s) in the lower end. Nevertheless, my first attempt came out fine with a bit of cleanup at the end (.1 blur and levels to clean up the residue).

Here's the 2-image packing example:

ThreeDee, Sun, 12 Oct 2008 14:36:37 -0400
And here's the extraction method:

CorvusCroax, Sun, 12 Oct 2008 14:43:49 -0400

Quote
ThreeDee wrote: You could save three alpha channels pretty easily by making each of them one of the RGB channels. But for even more capacity, take a look at this: Filter Forge can save lossless images with 3 channels, 16 bits per channel. If you use 8 bits for full quality anti-aliasing, you can theoretically still have 3x8 = 24 "alpha channels" at photoshop quality for every image. But how would you store them in a single image? By the upper 8 bits!

Quote

ThreeDee wrote:
You could save three alpha channels pretty easily by making each of them one of the RGB channels.

But for even more capacity, take a look at this: Filter Forge can save lossless images with 3 channels, 16 bits per channel. If you use 8 bits for full quality anti-aliasing, you can theoretically still have 3x8 = 24 "alpha channels" at photoshop quality for every image. But how would you store them in a single image?

By the upper 8 bits!

We actually use the RGB multi-alpha channel trick in games. It works pretty well. All sort of inter-related data packed into an RGB layer. We've talked about breaking each layer into it's respective bits, but it's hard to work that way. Not to mention messy; It's very easy to get a little bit of anti-aliasing from one area into another area accidentally. I've tried separating it all off with pshop threshold layers, but it's hard to work that way and your work habits have to be pristine.

Clever idea for FF, though!

ThreeDee, Sun, 12 Oct 2008 17:52:49 -0400

Quote
CorvusCroax wrote: We actually use the RGB multi-alpha channel trick in games. It works pretty well. All sort of inter-related data packed into an RGB layer. We've talked about breaking each layer into it's respective bits, but it's hard to work that way. Not to mention messy; It's very easy to get a little bit of anti-aliasing from one area into another area accidentally. I've tried separating it all off with pshop threshold layers, but it's hard to work that way and your work habits have to be pristine. Clever idea for FF, though!

Quote

CorvusCroax wrote:
We actually use the RGB multi-alpha channel trick in games. It works pretty well. All sort of inter-related data packed into an RGB layer. We've talked about breaking each layer into it's respective bits, but it's hard to work that way. Not to mention messy; It's very easy to get a little bit of anti-aliasing from one area into another area accidentally. I've tried separating it all off with pshop threshold layers, but it's hard to work that way and your work habits have to be pristine.

Clever idea for FF, though!

Yes, that makes sense in games, trying to cram as much data into as small a package as possible, sort of like I'm trying to do here since FF can take 5 channel inputs max at a time (R,G,B, selection and transparency).

Well, I managed to make something somewhat useful out of it. Check this out.

If you download the below image (right-click, save as) and apply the attached filter to it, lo and behold, you get 12 different pictures out of it. Yes, they're a bit raggy, but usable with some cleanup or by reducing their size.

I managed to pack and extract 21 channels in exploring this, but it was getting too slow to be that useful. 12 seems like an optimum of sorts at the moment.

ThreeDee, Sun, 12 Oct 2008 18:27:40 -0400
Importing font outlines into FF?

If you add a selection input to the above, you get 16 alpha channels.

Further, if you put, multi-image-input style, 4-up on each channel, you have 64 inputs -- or enough inputs to get the upper and lower case alphanumeric character outlines into FF.

Not that I can think of any sensible reason to do so, but maybe someone will -- or maybe I can come up with some nonsensical reason to do it anyways.

ThreeDee, Sun, 12 Oct 2008 18:48:54 -0400
Oh, and I forgot to make the originals 2-bit, which would obviate having to do the cleanup of the antialiased edges remnants...

ThreeDee, Sun, 12 Oct 2008 18:52:23 -0400
I was in the middle of another too-ridiculous-for-words filter but now I have to go to sleep. Not fair.

ThreeDee, Sun, 12 Oct 2008 19:05:39 -0400
Secret messages!

Okay, just one more crazy idea, then I'll go to sleep...

You could make image scrambling filters that you could process an image with at the sending end and then have another (or the same) FF filter that would unscramble it at the receiving end. For secure image transmissions. Or just for kicks.

ThreeDee, Mon, 13 Oct 2008 06:29:27 -0400
Here are couple of the ideas in action.

The image on the screen of the Etch-a-Sketch was drawn one line at a time in FF during several passes with the filter, "rotating the knobs" with two sliders.

You can see two small gray squares at the top corners; that is where the previous pass X and Y coordinates are stored and retrieved from.

Actually the biggest problem was figuring out how to draw lines between two points. It's not working correctly yet, for the lines disappear if they are drawn horizontally or vertically.

Plus, it is painfully slow to draw anything with it -- just like the real Etch-a-Sketch! :D

I think I have an idea for how to draw the lines easier.

Kraellin, Mon, 13 Oct 2008 09:21:49 -0400
hehe, you're going to have to embed that cyber sketcher on a web site somewhere :)

very clever, 3D. just fascinating watching your output here :)

CorvusCroax, Mon, 13 Oct 2008 11:58:48 -0400
Ah, steganography!

You could set it such that you have to know the correct variation # to decode!

Some techniques use the last 2 bits, ThreeDee:

http://en.wikipedia.org/wiki/Steganography

ThreeDee, Mon, 13 Oct 2008 16:33:00 -0400

Quote
CorvusCroax wrote: Ah, steganography! You could set it such that you have to know the correct variation # to decode!

Was thinking along the same lines...

Quote
CorvusCroax wrote: Some techniques use the last 2 bits, ThreeDee: http://en.wikipedia.org/wiki/Steganography

Ooh, most interesting. They've even got a sample image. Let's see if I can make a "reader" for steganographic images! Should work the same as my alphapacking/unpacking.

ThreeDee, Mon, 13 Oct 2008 18:21:08 -0400
I give up -- it is impossible. You can probably only do it as a bitwise operation.

ThreeDee, Mon, 13 Oct 2008 18:55:32 -0400
Custom shaders

HDRI shading is nice but there are instances where building the shading in a Simple filter works better. It gives you more control and additional options to do it that way. For example:

-Casting shadows from the bump map
-Multiple reflections or refractions
-Multiple-layer surfaces
-More realistic metals with blended reflections (sharp and blurred)
-Anisotropic reflections
-More controlled specularity, specular halos
-Reflectivity by angle
-Deeper shadows, more contrast
-Internal diffusion
-Spectral refractions
-Glow/self-illumination
-Custom lighting sources
-Custom environments
-Colored lights and ambient lighting
-Fuzzy objects

You can build some of these effects into Surface filters too, but the HDRI map tends to override and conceal them.

The key component for making such effects in Simple filters is Refraction, for it doubles as reflection, and the derivative simulated normal mapping which can be used for custom lighting.

And if you work with a 3D program, you can make switches to render these maps for use in a 3D program as applicable. Or keep them as faster-rendering simulations of the same effects permanently embedded into the diffuse map. Why recreate a bump map shading -- or anything else for that matter -- newly in 3D unless your lighting scenario is likely to change radically?

ThreeDee, Mon, 13 Oct 2008 18:56:18 -0400
What important shading types and effects am I forgetting?

StevieJ, Mon, 13 Oct 2008 22:06:59 -0400

Quote
ThreeDee wrote: there are instances where building the shading in a Simple filter works better

That's for sure.....I'd love to be able to use different lighting without the raised surface artifacts..... :eek:

Quote
ThreeDee wrote: What important shading types and effects am I forgetting?

Point source lighting would be nice.....but need FF to go 3D..... :devil:

CorvusCroax, Tue, 14 Oct 2008 15:09:14 -0400

Quote
ThreeDee wrote: What important shading types and effects am I forgetting?

Ambient Occlusion ;) (is sort of Deeper shadows, more contrast / casting shadows, but not exactly)

ThreeDee, Thu, 16 Oct 2008 12:45:30 -0400
I finally discovered Sphinx Shader V0.2 where he has already done some of these things.

StevieJ, Thu, 16 Oct 2008 13:02:31 -0400

Quote
CorvusCroax wrote: Ambient Occlusion

This works.....to a certain extent.....

http://www.filterforge.com/filters/4517.html

CorvusCroax, Thu, 16 Oct 2008 13:55:54 -0400

Quote
StevieJ wrote: This works.....to a certain extent..... http://www.filterforge.com/filters/4517.html Steve

Does it work realistically on an arbitrary shape? (ie not a photo, but a heightmap)

ThreeDee, Thu, 16 Oct 2008 14:40:02 -0400
Meanwhile, here's my first stab at a custom "surface filter" constructed as a simple filter. Got reflections, transparency, refractions and a (sort-of-a) lightsource. The blob represents an object. Or a blob.

ThreeDee, Thu, 16 Oct 2008 15:08:08 -0400
Same with teapot.

Kinda cute. Remarkably slow.

ThreeDee, Thu, 16 Oct 2008 15:17:01 -0400

Quote
CorvusCroax wrote: Ambient Occlusion

Quote
StevieJ wrote: This works.....to a certain extent..... http://www.filterforge.com/filters/4517.html Steve

Okay, I get the idea. The object itself occludes ambient lighting in its self-generated shadow areas. Like reverse radiosity.

CorvusCroax, Thu, 16 Oct 2008 16:15:36 -0400

Quote
ThreeDee wrote: Okay, I get the idea. The object itself occludes ambient lighting in its self-generated shadow areas. Like reverse radiosity.

Yes, that's correct. It's self-shadowing the ambient light. In a 3d app, an HDRI image for lighting would take care of this in the render, but since FF doesn't shadow, you don't get the self-shadowing of the ambient.

Ambient Occlusion (AO) goes a long way making an object look more realistic. Typical 'ambient' in computer graphics should almost never be used, IMHO, b/c it's just added everywhere, and makes an image look exceedingly fake looking.

AO is different from radiosity. Radiosity is like sunlight bouncing off a red wall, which makes the other things around it red.

ThreeDee, Thu, 16 Oct 2008 16:49:39 -0400
Yes, I get what you mean.

StevieJ, Fri, 17 Oct 2008 13:14:54 -0400

Quote
CorvusCroax wrote: Does it work realistically on an arbitrary shape? (ie not a photo, but a heightmap)

Good question.....only used it as a "fake out" on a normal map.....I should check that out.....

Quote
ThreeDee wrote: here's my first stab at a custom "surface filter" constructed as a simple filter. Got reflections, transparency, refractions and a (sort-of-a) lightsource.

That's a challenge.....getting a surface result refraction and shine out of a simple result.....but I believe it's possible.....

ThreeDee, Fri, 17 Oct 2008 18:38:37 -0400
I just realized the flaw in my logic with the idea of being able to store alpha channels in RGB. 16 bits isn't really any more than 2 times 8 bits, in terms of storage capacity -- duh! So with RGB, you could store a maximum of 6 alpha channels with 8-bit quality. But you could still store up to 24 1-bit alpha channels in an 8-bit RGB image, or 48 of them in a 16-bit image. At least theoretically.

ThreeDee, Sat, 18 Oct 2008 11:33:09 -0400
Steganographic Decoder returns

Hahaa -- steganographic decoding solved!

Here's the original image from wikipedia:

Run through Steganographic Decoder filter:

What made it work was using a Step Curve to make the original work like an 8-bit image even though FF does its calculations at floating-point accuracy!

(This method of faking 8-bit processing is bound to improve alpha-packing and unpacking success, too.)

So here's my question: is this just too obscure a filter to submit?

Or a sure-fire HU when you batch every PNG image on the web in search of more steganographically hidden images...? :D

Okay, I think I know the answer: if I make a Steganographic Encoder as well, it will be worth it, right?

CorvusCroax, Sat, 18 Oct 2008 14:28:47 -0400

Quote
ThreeDee wrote: So here's my question: is this just too obscure a filter to submit? Or a sure-fire HU when you batch every PNG image on the web in search of more steganographically hidden images...? Big grin Okay, I think I know the answer: if I make a Steganographic Encoder as well, it will be worth it, right?

Oh, you should submit it, just for coolness factor!

Kraellin, Sun, 19 Oct 2008 00:24:30 -0400
stegano what? i see some trees and i see a cat.

jffe, Sun, 19 Oct 2008 01:22:51 -0400
You turned a tree into a cat ?! :banana: Awesomeness. :D

jffe

Carl, Sun, 19 Oct 2008 03:24:43 -0400
You've encoded the cat pic with the tree pic but to make a decoder to get it out again that is a challenge - hats off if you can acheive that :D

Free Gradient angle rotation

Burt — Tue, 23 Apr 2013 20:20:18 -0400

Burt, Fri, 19 Apr 2013 20:36:03 -0400
This may be a baby question but what is the best way to get a Free Gradient to rotate in 90 degree increments the same way that the Profile Gradient does? I want to replace all the Profile Gradients in a filter but don't want to try to sort out all those rotations. I want play around with the XY Anchor available on the Free gradient to see if I can create a landscape mode version of a filter - right now only square and if I change the image there is an obvious line in the midddle. I guess I can feed all the Profile gradients into a Free Gradient but not sure if that is the best or only way as it would double the number of gradients.

CFandM, Fri, 19 Apr 2013 20:53:26 -0400
You can create it this way for one.... ;)

Burt, Mon, 22 Apr 2013 20:26:55 -0400
Seems like there are so many ways to do the same thing in FF (in most great tools I guess). Thanks for the sample. :)

CFandM, Tue, 23 Apr 2013 05:55:20 -0400
You are welcome.. :)
Yep thats one of the things I like to do is figure out how many ways to do the same thing...For instance there is another way to get a profile gradient...Although this one is not a non-square.... ;) :)

Burt, Tue, 23 Apr 2013 20:20:18 -0400
Cool. Yes it is easy to get caught up with FF. :)

Suggest a new filter.

uberzev — Sat, 20 Apr 2013 09:36:32 -0400

uberzev, Fri, 30 Mar 2007 02:59:08 -0400
Lets here some ideas.

Sjeiti, Fri, 30 Mar 2007 03:11:05 -0400
feathers

Crapadilla, Fri, 30 Mar 2007 04:42:40 -0400
pizza
swiss cheese

StevieJ, Fri, 30 Mar 2007 12:55:51 -0400
Kraellin's "Monster" filter got me thinking of filters that would be able to select and change skin to another texture. For example, change just the skin of a persons portrait to, let's say, Vlad's organics or Larryb's Malachite texture. I think that gamers would love filters like this for their characters.....and make it easy to bypass all that work to get a much quicker result....

Maybe filters that also form-fit selected areas of a figure....say using "MetalStorm" to quickly equip figures with alien body armor etc....and be able to puff it out with a 3d appearance.....kinda like the Borg on Srar trek :)

Just a thought :)

StevieJ, Fri, 30 Mar 2007 13:07:31 -0400
Also, filters that create fiberous hair on things seems to have been mentioned quite a bit lately....

LOL.... the "Wig Creator" filter :D I have some friends that would like something like this to touch up pictures of themselves :D

Kraellin, Fri, 30 Mar 2007 14:06:57 -0400
excellent ideas, steve and you just gave me one. and since i doubt i could pull it off myself, i'll throw it out there before even trying myself.... tatoos!

also, some of what you're talking about, steve, is displacement mapping. but, this could be done with FF too. hmmm, i forget the name of this now, but i made a request a while back for an 'extract 1 color' type filter and they produced it. it used a set alpha to extract just one color or 1 color range from a given image. you could do something like that with skin tones and do what you're suggesting with the displacement mapping thing.

and yes, hair is ALWAYS a big deal with retouchers and artists.

you also got me thinking of another type. how about a facial redesign filter? or maybe an age regressor or age progressor type? you know, where you take someone's photo and subtract 10 years of age from it or add 10 years of age to it. in my case, i'd like to remove about 30 :)

StevieJ, Fri, 30 Mar 2007 14:39:06 -0400
Kraellin, those are excellent ideas!!! :) Age Regressor/Progressor is a really good one....

Zephos, Mon, 02 Apr 2007 20:11:27 -0400
Such excellent ideas! The one that caught my attention was Sjeiti's feather one actually -- well, as in the thing that falls off of birds as oppose to selection feather in Photoshop (As I understand).

I was working on a Feather filter at one point when I first started FF as a sister filter to Sakurama, but unfortunately, feathers were too complicated a shape to make appear naturally in clusters to my satisfaction, so I have yet to bring it to a level that I have confidence in.

Crapadilla, Tue, 03 Apr 2007 05:05:34 -0400
I'd like to see feathers too, but it appears we'll have to wait for that to become possible/reasonable with FF.

Kraellin, Tue, 03 Apr 2007 08:58:22 -0400
wow, dilla; i like that one. fire feathers!

Crapadilla, Tue, 03 Apr 2007 10:13:31 -0400
It's a Genetica texture though... ;)

Kraellin, Tue, 03 Apr 2007 22:32:46 -0400
well, it's still a good image :)

uberzev, Wed, 04 Apr 2007 00:55:13 -0400
Raw steak...

Crapadilla, Wed, 04 Apr 2007 02:40:03 -0400
uberzev is right. Pictures!

Crapadilla, Wed, 04 Apr 2007 02:43:46 -0400
And...

Crapadilla, Wed, 04 Apr 2007 02:48:50 -0400
Burned paper

uberzev, Thu, 05 Apr 2007 03:49:40 -0400

Crapadilla, Fri, 06 Apr 2007 06:24:04 -0400
Tiffany Glassworks

Crapadilla, Fri, 06 Apr 2007 06:25:49 -0400
De Stijl

Crapadilla, Fri, 06 Apr 2007 06:31:16 -0400
Lego mosaic

Crapadilla, Fri, 06 Apr 2007 06:31:51 -0400
Lego mosaic

Crapadilla, Fri, 06 Apr 2007 06:37:56 -0400
RPG style random city maps (should be quite challenging to do right)... Sjeiti? ;) :D

Ken, Fri, 06 Apr 2007 20:47:43 -0400
Hi.

I had an idea to make a Picture Cube.

I looked at CF&M's 3D Cube but that did not contain an image so I started again with it.
I can't quite get it to work right so I anyone would like to finish it I think it could be quite a useful filter.

I also tried using the 'Frame Component' but I again had trouble with the Top

Maybe this could also be made into a 'Book'

Ken.

CFandM, Fri, 06 Apr 2007 23:01:20 -0400

Quote
Ken wrote: I looked at CF&M's 3D Cube but that did not contain an image

I just posted the one that I had with an image but never posted it..Maybe something useful in there...I haven't thought about this one for a while.. :)
http://www.filterforge.com/forum/read...10&TID=747

StevieJ, Fri, 06 Apr 2007 23:06:07 -0400
Dilla, Tiffany Glass is a good one.....and the Logo Mosaics is an excellent idea!!! I think that Zeph is going to like that second one.... :)

Torley, Sat, 07 Apr 2007 10:45:29 -0400
I'd like to see more cool techno pattern ones — like one that does a simulated Star Trek LCARS interface, with those sloped lines and the distinct color scheme!

StevieJ, Sat, 07 Apr 2007 12:31:28 -0400
That would be cool! I didn't know you were a "Trekkie" :)

This got me thinking that 'type' components might be very useful.....

Torley, Sun, 08 Apr 2007 03:24:56 -0400
@StevieJ: Indeed, I am big into Star Trek! And I think Mike Okuda is a graphic design genius; so many of his innovations have been copied into fandom and beyond, used in "serious" capacities. I was surprised upon reading the [url=http://en.wikipedia.org/wiki/Lcars]Wikipedia entry for LCARS[/urk] of such initiatives.

Torley, Sun, 08 Apr 2007 22:54:01 -0400
Another somewhat general suggestion: I'd like to see more filters with a minimalist glitch art aesthetic, or sleek techno-stuffs. For example, this video titled "Trioon" has really beautiful, simple shapes synced to the music. I know our filters aren't animated, but generating things that looked like still frames out of this can't be too hard, yet it'd be quite aesthetically pleasing.

More examples of what I have in mind:

» http://www.meso.net/video/video.php
» http://www.dienststelle.de/visuals/

Note the soft blurring 'round the edges, these aren't totally "hard" visuals.

I'm also heavy into the simple-and-colorful visuals exemplified by such arcade classic remakes as Tsunami 2010, which has such a spacey, tripped-out vibe to it.

Also, a well-rounded natural-media plugin that could do a lot of what Alien Skin Snap Art can do would be distinguished, and very popular, I'd think! (We already have several similar filters, but there remains to be so much exploring done in this area!)

StevieJ, Mon, 09 Apr 2007 00:15:16 -0400
Wow!!! I checked out all of it and I'm intrigued.....and makes me feel like I've been leading a somewhat sheltered digital life :)

I like gaming too. When I get the urge, I enjoy strategy games like Armored Core....or some older ones like Command & Conquer. I see the visual draw on Tsunami 2010.....

I had started programming filters with the Photoshop SDK when I discovered Filter Forge almost three months ago.....and that was the end of the SDK :)

Kraellin, Tue, 10 Apr 2007 09:12:13 -0400
well, we've done fire, but where are the explosions? that's always one that's in demand.

Crapadilla, Sun, 15 Apr 2007 05:56:19 -0400
Carbonite... would be sweet as an effect filter, although it'd probably work better on a true heightmap rather than on your average image...

Crapadilla, Sun, 15 Apr 2007 05:57:46 -0400
Side orders, anyone? ;)

Kraellin, Sun, 15 Apr 2007 12:22:56 -0400
dilla,

that 'carbonite' image is intriguing all by itself. having a filter that could do that would be rather amazing. you'd probably have to have a plotting component to do it right, though.

StevieJ, Sun, 15 Apr 2007 13:09:27 -0400
I think the "monster" could cover the carbonite effect, eh? :D Probably wouldn't do so well with the bacon LOL.... :)

Torley, Mon, 16 Apr 2007 22:38:11 -0400
Ahhhh I have a practical idea, I'd like to see someone make a tilt-shift photography filter. If you're not familiar with what that is, it's a technique that lets you create the illusion that a full-scale photo is a miniature model! There are some examples here:

» http://hame.ca/tiltshift.htm
» http://www.flickr.com/photos/tags/tiltshift/

Here's how to do it in Photoshop:

» http://recedinghairline.co.uk/tutorials/fakemodel/

Surreal and beautiful at the same time. :)

uberzev, Mon, 16 Apr 2007 23:12:24 -0400
The tilt shift effect requires a masked controlled blur which isn't a FF feature currently.

Crapadilla, Sat, 28 Apr 2007 11:54:10 -0400
There are some really sweet Genetica textures around. Anyone feel like 'filterforgetizing' them? ;) :D

Old Dollhouse

Crapadilla, Sat, 28 Apr 2007 11:55:20 -0400
One for StevieJ: Rose Crystal Assortment.

Crapadilla, Sat, 28 Apr 2007 11:57:23 -0400
Garnet

Crapadilla, Sat, 28 Apr 2007 11:58:15 -0400
Girl's Best Friend

Crapadilla, Sat, 28 Apr 2007 12:03:48 -0400
Blood Spattered Wall.

Kraellin, Sat, 28 Apr 2007 12:04:05 -0400
cool, dilla. yes, some gemstones would be great.

Crapadilla, Sun, 29 Apr 2007 16:22:58 -0400
Crystalline Fissures

Crapadilla, Sun, 29 Apr 2007 16:24:39 -0400
Fluorite Cavern

Crapadilla, Sun, 29 Apr 2007 16:26:50 -0400
This one rocks!

Melting Rubber

Crapadilla, Sun, 29 Apr 2007 16:30:51 -0400
Excellent...

Smashed Gray Marble

Crapadilla, Sun, 29 Apr 2007 16:32:55 -0400
Funny concept...

Battered Robot

Crapadilla, Sun, 29 Apr 2007 16:35:18 -0400
This one is called 'Gold Rush', but reminds me of iced coffee...

Crapadilla, Sun, 29 Apr 2007 16:38:00 -0400
In case anyone wants to take a further look, there is more goodness here. Some excellent textures, I must say. Seems we got some breadth development work to do... ;) :D

FWFF (or foolin' with Filter Forge)

Burt — Mon, 08 Apr 2013 18:47:07 -0400

Burt, Thu, 14 Feb 2013 11:48:28 -0500
I'm just playing, having fun with FF and wanted a place to post. Here's one that is sort of like ocean rock. In fact the last one even seems to have a few tube worms popping up so will have to look and see where they are coing from.

Sharandra, Thu, 14 Feb 2013 16:29:27 -0500
Oh nice, coral reefs! Looks promising :-)

ddaydreams, Thu, 14 Feb 2013 20:40:47 -0500
I like the top image best.

Burt, Thu, 14 Feb 2013 21:01:57 -0500
Thanks. I imagine its the same for everyone but I was doing something else, saw an interesting pattern and cut it out to start a new filter. I've got a whole bunch of these little snips now allw waiting to become filters - whether I get to them or not is another question. :)

Burt, Fri, 15 Feb 2013 10:04:18 -0500
Decided to create some 'base rock' to grow the coral on.

Trying to figure out a few things and place some little sea urchins, plants, shells etc 'on top' of the rock but right now every thing is embedded.

Burt, Mon, 18 Feb 2013 04:48:31 -0500
Played with the 'base rock' again today but has grown a litle heavy. I was working on some shrimp and othe critters but had to simplify them as it was crawling.

Sharandra, Mon, 18 Feb 2013 08:11:55 -0500
Looks great! :-) I like the rock with it´s little caves, can almost see little fishes swim in and out! :D

Burt, Mon, 18 Feb 2013 12:46:41 -0500
Thanks. I actually thought of putting some eyes in the dark areas but will have to optimize it some as it was starting to crawl. :)

Burt, Thu, 21 Feb 2013 03:17:19 -0500
Well I see there is a user gallery s maybe this thread should have been there -not sure. Anyway trying to mix the coral and the rock above now. Still needs to be optimized, a few small critters and som lighting effects I hope. Fun though I'm getting even less slep than usual playing with FF. :)

Skybase, Thu, 21 Feb 2013 03:47:32 -0500
I really love your color scheme! These are looking wonderful! :)

Burt, Thu, 21 Feb 2013 13:31:11 -0500
Thanks. I am having fun with this.

Burt, Tue, 26 Feb 2013 15:57:39 -0500
Well still poking at this -almost lost it after making some changes as I was looking at some nodes, but not the final render -and at some point I had broken it. I was able to go to an older backup and get it mostly the same but still need to rework the lighting, solve a few placement issues (getting some floating polyps to attach to the right spots) and then optimize. This image took 45 minutes to render!

Burt, Tue, 26 Feb 2013 19:21:22 -0500
Here's one more (again 45 minutes yikes. making copies and optimizing soon). It's ugly squashed top down view but you can see the shrimp anyway.

Sharandra, Mon, 04 Mar 2013 14:22:05 -0500
Cool! Love the shrimps! :D

Burt, Mon, 04 Mar 2013 18:18:51 -0500
Well trying to trim it down. Put everything including the kitchen sink in there and it is slow. Trying to optimize and cutting out a few things which probably will have to be seperate filters. I wanted to build a full ocean replicator but guess I may have overstepped my ability and RAM just a bit. ;)

Burt, Mon, 04 Mar 2013 21:09:44 -0500
I'm thinking I will have to submit this soon and do a second round. I have a lot to learn about optimizing and found the excellent Wiki articles after I built this monstrosity. I removed some sections and have tried to use some of the techniques but wil likely have to rebuild it section by section to take advnatage of some of the techniques. I'm down from about an hour rendering to 40 minutes. Here is one in progress.

Burt, Wed, 06 Mar 2013 18:40:22 -0500
Thanks to Skybases's answer to my grid question Forum Post I now have some schooling fish.

Burt, Thu, 07 Mar 2013 02:50:48 -0500
Well finished for now though far from perfect. Some trouble uploading so will try tomorrow. In the mean time here are a few renders.

Burt, Thu, 07 Mar 2013 02:51:08 -0500
Here's another.

Burt, Thu, 07 Mar 2013 02:51:26 -0500
and one final one for tonight...

Sharandra, Thu, 07 Mar 2013 09:26:44 -0500
Very cool, especially the last one. The school of fish is great! :-)

Burt, Thu, 07 Mar 2013 09:41:50 -0500
Thanks. Lost some things in all the changes like the polyps in the tubes. Also some of the nice detail when I switched it to a Simple filter but learned a lot and will fix some of those things in the next version.

Burt, Thu, 07 Mar 2013 21:14:21 -0500
While waiting to get past this upload error I took a sample Indigo Ray provided and simplified to add a basic 'Camera Mode' -zoom, rotate, up/down/left/right.

Burt, Thu, 07 Mar 2013 22:44:05 -0500
I was just playing with some curves and got some interesting patterns. Not enough to submit to the Library so posting it here.

Burt, Thu, 07 Mar 2013 22:45:14 -0500
Here's one render

Burt, Thu, 07 Mar 2013 22:48:18 -0500
Here's another

Burt, Sat, 09 Mar 2013 21:19:42 -0500
Here's that same curve with Pavement.

Burt, Mon, 11 Mar 2013 18:17:03 -0400
Well just added a simple motion trail and turned this into a snippet filter. It doesn't offset the image in any way so lots can still be done to it but I though the effect was cool.

Burt, Tue, 12 Mar 2013 19:18:15 -0400
Well I'm just fooling around -actually tring to understand how the coral filter generates its height - I got it but want to be able to reproduce at will. In my poking I came up with this effect which looks like sci-fi buildings - or maybe I'm reaching and just want to see it. :)

Burt, Tue, 12 Mar 2013 19:18:37 -0400
here's another

Burt, Tue, 12 Mar 2013 19:18:51 -0400
and one final one.

Burt, Tue, 12 Mar 2013 20:22:31 -0400
If I don't screw this up I think I will try to make a sci-fi background generator.

Sharandra, Tue, 12 Mar 2013 21:32:44 -0400
Oh cool! I did some greeble experiments earlier, because I originally wanted to make a death star in the foreground of my space scene filter, but I failed so far lol. :-)

Burt, Wed, 13 Mar 2013 10:39:28 -0400
This started as the unexpected result of an experiment. I'll work on it a bit and add some controls before submitting -right now still manually plugging and unplugging some connections to see different results.

Burt, Wed, 13 Mar 2013 17:38:28 -0400
Well no controls yet, still randomly plugging and unplugging inputs, trying different gradients, shapes and patterns as inputs and just fooling around to see what might be possible - not really a linear thinker (to my dismay).

It's cool because on some renders I can almost see tiny windows, ceiling and floor patterns, doors and other features - I know I'm probably losing it. Wonder if they wife would let me take a week off work with no pay so I could play with FF? ;) Anyway I thought this one came out cool too.

Burt, Thu, 14 Mar 2013 00:56:01 -0400
I was poking at some way to generate anemones or soft corals for the coral filter and I missed the obvious - the lowly polygon does a pretty good job.

Burt, Fri, 15 Mar 2013 12:17:39 -0400
Hope I'm not boring anyone with my babble -just having lots of fun with Filter Forge (wish I could fin a way to do this for a living :) ).

More play with Bomber, Smudge and curves and ThreeDee's answer to my question based on Sphinx's improved Smudge bit -already messed up my first example above with too much tweaking

I have also been experimenting with using the Frame component on several filters as a square gradient of sorts - some weird and interesting results now and then.

Burt, Fri, 15 Mar 2013 12:19:59 -0400
a few more.

SpaceRay, Sat, 16 Mar 2013 01:29:48 -0400
Good and nice smudge tests you have done with those greeble experiments :)

Burt, Sat, 16 Mar 2013 18:44:31 -0400
This is a simple sample of greeble 'city' based on ThreeDee's sample which as based on Sphinx's sample.

Burt, Sat, 16 Mar 2013 18:47:41 -0400
Here's a shot -nothing fancy.

Sharandra, Sun, 17 Mar 2013 09:32:45 -0400
Lol weird distortions :-) I like the cities.

Burt, Sun, 17 Mar 2013 12:21:25 -0400
I'm going to try to clean up and post some of the other versions I posted earlier about, see if I can add at least some control.

Burt, Mon, 18 Mar 2013 17:41:04 -0400
Playing with the greeble city idea some more. I'm not sure if this will actually be useful to people working with textures or not but it is fun.

Burt, Fri, 22 Mar 2013 01:16:18 -0400
Well I went off on a real tangent with this greeble stuff - and it was just an accidental result that derailed me. The brain is mush. Anyway this popped out today and thought it looked cool.

Burt, Fri, 22 Mar 2013 01:16:31 -0400
here is another

Burt, Fri, 22 Mar 2013 01:17:45 -0400
and yet another - machine or engine like I think

Burt, Fri, 22 Mar 2013 01:18:11 -0400
Final one for now.

ThreeDee, Fri, 22 Mar 2013 06:47:13 -0400
Really cool sci-fi machinery, Burt!

Sharandra, Fri, 22 Mar 2013 12:31:04 -0400
Wow! Very cool! :D

Shar´s Forge

Sharandra — Sun, 31 Mar 2013 23:48:09 -0400

Sharandra, Tue, 24 Jul 2012 23:22:37 -0400
Actually it´s more of a garden right now :D

Growing some ivy.

Sharandra, Tue, 24 Jul 2012 23:38:49 -0400
:-)

Skybase, Wed, 25 Jul 2012 00:42:58 -0400
This I like a lot :D

Sharandra, Wed, 25 Jul 2012 02:02:12 -0400
Someone needs to cut the ivy! :P

Indigo Ray, Wed, 25 Jul 2012 08:11:21 -0400
Oh wow! 8)

Morgantao, Wed, 25 Jul 2012 08:45:40 -0400
Very cool!
How do the vines behave? Do they follow the image, or do you have to use a mask?

Sharandra, Wed, 25 Jul 2012 12:23:24 -0400
They misbehave and grow all over the place! :D

Using a frame/Gradient to crop them atm.

Would be cool to have some diffusive growth patterns.
Math hates me, else I´d try to script it. Would prolly be very slow tho.

Sharandra, Wed, 25 Jul 2012 12:59:45 -0400
Playing with the edge detector component.

Sharandra, Wed, 25 Jul 2012 13:29:05 -0400
Doesn´t work with all images yet tho.

Sharandra, Wed, 25 Jul 2012 18:23:49 -0400
Speedpainting filter - ruins your images in record time!

Sharandra, Thu, 26 Jul 2012 00:52:42 -0400
Almost there. :-)

Morgantao, Thu, 26 Jul 2012 02:20:36 -0400
Oh, last one is very nice!

Quote
ruins your images in record time!

Ghislaine, Thu, 26 Jul 2012 09:35:09 -0400
Love this Sharandra ! in particular the second and the third one. I allow me to make a photo-montage with your last image combined with my solar eclipse. You can see this here

Sharandra, Thu, 26 Jul 2012 18:54:02 -0400
Added the option to use your own texture for the leaves. I´m planning on adding some more shapes aswell, but for now I´ll leave it as it is.
Have to do a bit of testing, but I think it´s fit for release :-)

One with custom leaves and Gigi´s eclipse :-)

Ghislaine, Thu, 26 Jul 2012 19:21:53 -0400
Nice return Sharandra ! Love this kind of game :)

OK, I'm glad that you added this option. Your filter will be very useful. You always make good and beautiful filters.

As you said, "Have to do a bit of testing" on my eclipse filter too.

Morgantao, Fri, 27 Jul 2012 01:49:51 -0400
Looking forward to seeing it in the library :)

Sharandra, Fri, 27 Jul 2012 03:06:26 -0400
Well I just send it on it´s journey through the depth of the submission queue :-)
So hopefully you can play with it soon :-)

Sharandra, Sun, 29 Jul 2012 16:09:43 -0400
Trying to make a stonewall with irregurlar stones. at least I can grow some ivy on it to cover up how hideous it looks :P

SpaceRay, Sun, 29 Jul 2012 22:36:07 -0400
Very well done and nice stones, and very good for the ivy to grow over it, surely it will look beautiful

Sharandra, Mon, 30 Jul 2012 00:42:35 -0400
Thanks. :P Not entirely what I am after but it looks decent enough on my model, that´s the important thing.

Just need to add some moss and stuff. :-)

Sharandra, Mon, 30 Jul 2012 15:00:33 -0400
Added Moss, looks nicer now :-)

Sharandra, Tue, 31 Jul 2012 03:43:41 -0400
Getting there :-) Think I am more or less done :D

Sharandra, Tue, 31 Jul 2012 03:56:47 -0400
Bending the tiles to my will was challenging :D

SpaceRay, Tue, 31 Jul 2012 04:05:06 -0400
This is MOSST beautiful ;) :D

Very well done and very realistic moss, like it.

Sharandra, Tue, 31 Jul 2012 08:43:41 -0400
Glad you like it :D

Ghislaine, Tue, 31 Jul 2012 10:07:09 -0400
Very good looking Sharandra. This is also my cup of tea. :)

Skybase, Tue, 31 Jul 2012 12:27:39 -0400
:)

Sharandra, Tue, 31 Jul 2012 20:24:09 -0400
:D

Well I´ve called it done for now and submitted :-)
Not entirely happy with the speed, but the results are quite nice. I´ve optimized it for randomizer use, so not a lot of fiddling with settings, makes up for the slowness :P

I´m gonna make a forest floor spin-off from this one (hopefully a bit faster :S), but I need to finish another filter first :-)

Sharandra, Wed, 01 Aug 2012 13:07:08 -0400
Filter Forge Garden :D

SpaceRay, Wed, 01 Aug 2012 13:16:13 -0400
WOW, this is really beautiful and very well done, like much the mix of the ivy and the stones, moss and great effect with the water, and the added realistic lighting. Great!!!

Morgantao, Thu, 02 Aug 2012 01:03:12 -0400
Very nice garden!

Sharandra, Thu, 02 Aug 2012 04:32:12 -0400
Thanks :-)

I used Add Atmosphereto enhance the lighting, apart from adding atmosphere it also enhanced the form of the wall :-)

For the water I used Sea Level and Waterfallofc :-)

Indigo Ray, Thu, 02 Aug 2012 07:40:24 -0400

Quote
Filter Forge Garden

Holy crap! :eek: 8)

Kraellin, Thu, 02 Aug 2012 15:34:36 -0400
Yep, I had to log in just to say holy crap, also! That's bloody amazing! I DO hope you publish this!

Sharandra, Fri, 03 Aug 2012 11:11:21 -0400
Aww thanks :blush: :)

It´s not a filter, was just a quick test of ivy on the wall and then I got carried away playing with some effect filters :-)

SpaceRay, Fri, 03 Aug 2012 11:25:25 -0400

Quote
Kraellin wrote: That's bloody amazing!

I agree and is true that is a great work

Quote
Kraellin wrote: I DO hope you publish this!

As said by Sharandra above is not a filter is a amazing work of mixing different filters to create this artwork, mainly used her Ivy and Boulder wall and others too

Quote
Sharandra wrote: I used Add Atmosphere to enhance the lighting, apart from adding atmosphere it also enhanced the form of the wall For the water I used Sea Level and Waterfall ofc

Sharandra, Fri, 03 Aug 2012 21:39:55 -0400
Was missing some water lilys in my pond :P

SpaceRay, Fri, 03 Aug 2012 22:10:50 -0400

Quote
Was missing some water lilys in my pond

This is the pond that is at the bottom of the Filter Forge Garden? :D

WOW, this looks very well and like your water lilys !!

Sharandra, Fri, 03 Aug 2012 22:54:44 -0400
Hehe, thx Ray :-)

here´s a better version. and it´s fast for a change...yet! >-< :D

Now let´s see if I can do a perspective version of that.

Ghislaine, Fri, 03 Aug 2012 23:06:32 -0400
Very nice ! It will be one more in my collection :)

Sharandra, Sat, 04 Aug 2012 11:40:58 -0400
Not entirely where I want it yet, but getting there :-)

Sharandra, Sat, 04 Aug 2012 14:15:00 -0400
Sadly doesn´t work on non square images :-(

Ghislaine, Sat, 04 Aug 2012 14:34:29 -0400
Yes it is a problem but I'm hoping that this will be possible for the next upgrade.

Sharandra, Sun, 05 Aug 2012 10:55:08 -0400
Water lilys don´t grow in midair... :-)

Sharandra, Sun, 05 Aug 2012 12:42:06 -0400
after more tweaks to the water,a test with image background :-)

Ghislaine, Sun, 05 Aug 2012 12:59:26 -0400
Cute ! The wall complete the scenery. Those two (water and this wall) goes well together. :)

SpaceRay, Sun, 05 Aug 2012 17:44:24 -0400

This looks very good and nice, and realistic, the shadows under the leaves, and the beautiful 3D water lilies, and together with your boulder wall, is complete.

Sharandra, Sun, 05 Aug 2012 22:45:01 -0400
Hehe thx.
They´re less realistic than I would like, but at least they make a great base for fake paintings :D

Sharandra, Mon, 06 Aug 2012 02:09:42 -0400
Weeeeeee.....flying island! XD

I like Sphinx´s improved spherical extrusion thingy :-)

SpaceRay, Mon, 06 Aug 2012 03:19:12 -0400
WOW, nice flying island ! Although the top part could be a little less crowded in relation to the size of the island

I have made a search in google images "water lilies pond" and got something similar to what you have done, so I think that your result is very good.

Stroke

Burt — Sun, 24 Mar 2013 14:07:27 -0400

Burt, Sun, 24 Mar 2013 02:21:38 -0400
I would like to apply a stroke effect to this filter sample - like the stroke in Photoshop that applies a line around an image (I really only need to retain the gray line and everything else transparent). The way I did it here with the Min and Max works in about a second here but I am adding this a more complex filter and it grinds to a halt which ThreeDee explained is is a result of those components rasterization http://filterforge.com/forum/read.php...&PAGEN_1=2.

I tried several ways -like Blur one, then Blend with Multiply mode, and Threshold but failed to get a well defined stroked border. I am wondering if someone has any other 'lite' ways of generating a border like this?

ThreeDee, Sun, 24 Mar 2013 12:58:14 -0400
Here goes

Burt, Sun, 24 Mar 2013 14:07:27 -0400
Thank you ThreeDee. I really appreciate all of your help. :)

FF math: 30+1=163

Morgantao — Sat, 23 Mar 2013 19:24:10 -0400

Morgantao, Tue, 17 Apr 2012 15:07:52 -0400
Can someone please explain this to me?

I have a filter that takes 30 seconds on avarage to render an image. Let's call it Romancing the Gorrilla by Carl. ;)
I take the resulting image and feed it as input to a socond filter, that takes around 1 second to render. Let's call this filter M_Shadows and Highlights by Morgantao :D

That took a total 31 seconds of render time, right?

So how it is that if I take the second filter AS IS and put it at the very end of the first filter, it takes 2 minutes and 43 seconds to render the exact same image?? :|

I even added a switch so I can just turn my filter on/off inside of Carl's filter, and I get the exact same results... When the switch is set to pass through my filter, it takes alsmost 3 minutes, and with the switch set to bypass my filter it takes only 30 seconds. :(

What gives?! :?:

Carl, is it ok with you if I upload the combined filter here?

Nebukadnezar, Tue, 17 Apr 2012 15:21:29 -0400
Is there some kind of cache-component?
The difference is perhaps that in one case the source is a bitmap.
What happens if you plug a small blur between the two filters? (For testing.)
I read somewhere blur is one of the few components that renders a bitmap.

SpaceRay, Tue, 17 Apr 2012 15:30:43 -0400
I have tried it myself and have put your filter after the Switch component and is true that it is much slower after this :?: and even more weird and strange is that if you put a switch to turn off your filter, THE TIME IS THE SAME :?: :?:

There something fishy in this mix of filters as this can´t be possible

Morgantao, Tue, 17 Apr 2012 15:35:55 -0400
I added the blur as you suggested, and it takes the render time down a little, but it's still waaaay slower than the original filters...
If it is indeed a caching issue, I think it's time that FF inc. either solve the issue, or at least add a bitmapping component, kind of a result node that has an output (SpaceRay, sounds familliar?).

Nebukadnezar, Tue, 17 Apr 2012 20:18:25 -0400
I think there might be something like "too much" precision.

The Filter Forge algorithm seems to perform very poorly with long filter chains.
One reason might be to safe RAM, but there are times when it would clearly benefit from simply rendering each step in the chain and passing it on further down, instead of doing whatever it does. (Which has probably its advantages, too.)

Try linking several edge detection components and witness your system come to a grinding halt. Try running the same number of those edge detection components in parallel and it's no problem at all. For many cases that extra precision is not needed at all and there should be the possibility to switch to a more stupid/less precise/forward algorithm.

I imagine something like this happens:
A point of the last component needs 4 points of the second last component to be calculated.
Each of those 4 points needs 4 points of the third last component to be calculated
Each of those 4 points needs 4 points of the fourth last component to be calculated.
And so on.
Exponential growth, rendering frenzy, grinding halt.

While with a forward approach (with caching/setting back the precision after each step) the rendering time would increase more linearly. Each time it's the "same" four points that are used to render the next step, instead of a ever growing/finer mesh.

Both ways have their advantages so either the program or the filter author should be able to decide which way to go.

(I basically don't know what I'm talking about so bear with me :D )

Thinking about it, if the information about each pixel/point in each step is kept and the resolution stays the same down the chain there should be no difference. But even more RAM would be needed than building a cache for each step and flushing the older ones down the toilet. (Ok, you could render buckets either way like FF seems to do.)

But if you get finer and finer the deeper you go...

SpaceRay, Tue, 17 Apr 2012 23:34:46 -0400
I want to put here how I have mixed both filters from the first post and do not know if this is the best way to do it or there could be a better alternative to mix them.

The Switch is the very last component of Carl´s filter before the result component, so I changed the result for Morgantao´s filter.

I have reduced the way Morgantao have arranged the filter so it can fit everything in the screen and can make a screenshot that is big to see it well.

SpaceRay, Tue, 17 Apr 2012 23:36:44 -0400
Nebukadnezar thanks for the theory and suggestion, well done

Good explained, fr om all of you have said I have a question

Does this mean that if FF could have more RAM would be faster and do not need to economy RAM ?

______________________________________________________

Could the Morgantao´s filter be put in another place than the end of Carl´s filter and this way would be faster ?

Although really I do not wh ere to put it because from the construction of Carl´s filter I think that you can´t put it before the last switch component

Morgantao, Wed, 18 Apr 2012 04:02:24 -0400
SpaceRay, while this is an interesting qeustion, does it really matter?
We both know that one of the main problems with FF is the custom memory manager, and we both know there's no plans of updating the 6 year old manager with something better.
As have been said on another thread, the FF inc. has programed itself into a corner. What they did may have been ground breaking 6 years ago, but it lacks the ability to evolve and be adapted to newer hardware.
What FF needs now is nothing less than a complete overhaul. It should be rebuilt from the ground up, with lessons learned from the past.
FF inc's official response to this was: No way! Too much work.

Sphinx., Wed, 18 Apr 2012 04:34:07 -0400
While Nebukadnezars observation about sample branching is a very important aspect to focus on regarding performance, I don't think that is the problem in this actual case:

The memory usage vs FF allocation restrictions seem to cause the bitmap buffers to get flushed (in blur, sharpen etc). I'm not sure about the details here. If this is correct, you should see a significant performance improvement at some point if you reduce the input image size (try e.g. 256x256 and see if the same strange performance penalties show up).

That is also why the blur buffer trick won't work (actually it will just use more mem.) - the reason you see a slightly increase in speed is most likely related to the anti alias process getting simplified by that blur..

Nebukadnezar, Wed, 18 Apr 2012 12:25:25 -0400
That's the nicest way anyone ever told me I was wrong. :D

SpaceRay, Thu, 19 Apr 2012 02:48:58 -0400
Thanks Sphinx for the answer but I am sorry to say that I really do not have the enough knowledge about FF to understand your answer.

Morgantao, very well said, and totally agree with what you have put in your last post.

It would be really nice to have the opinion or any suggestion from FF inc. itself as they know really best how FF works and WHY this is happening and if there could be a good way to solve this if possible.

Sphinx., Thu, 19 Apr 2012 02:53:42 -0400
Ok, the essence is that the memory restriction is the likely culprit. You can verify this claim by testing with lower resolution images: at some given lower size I predict that the strange additional overhead disappears :-)

Morgantao, Sun, 22 Apr 2012 13:33:44 -0400
Well, I just tried testing with a 100x100 image, and pure gorilla takes around 8 seconds.
With my addition it's close to 40 seconds.

Do you still think it's just a memory thing?
Maybe there's anything else you could think of? :|

SpaceRay, Sun, 22 Apr 2012 22:32:14 -0400
This is really strange, as you have confirmed that at a very low resolution of 100x100 keeps THE SAME speed so, the problems seems not to be what Sphinx suggested well, and there is another thing that we are missing and is happening and we do not see it.

Why the mix of this components give this much slower result ? :eek: :?: :?: :?:

SpaceRay, Sun, 29 Apr 2012 00:22:48 -0400
As the good Sphinx RAM theory could have been possible it seems that as Morgantao have showed it is not true, so does anybody else have another suggestion about WHY this is happening ?

We should call Sherlock Holmes to solve this mistery ? :?: :D

SpaceRay, Wed, 09 May 2012 10:25:40 -0400
Has been discovered the hidden mystery behind this ? :?:

Could this also happen with other mixes of filters or selected components and we do not know it unless we make a benchmark test like Morgantao have done ?

Morgantao, Wed, 09 May 2012 14:32:34 -0400
I sure wish someone could test this out and see if he can find out what the problem is...

SpaceRay, Wed, 09 May 2012 22:27:14 -0400
Perhaps you could contact the FF support and as ask to the experts at Filter Forge and see if they can give some light over this problem and why this happens.

It could be that it´s a UIB (Unidentified Invisible Bug) and then they could know about this and could perhaps find a fix.

Morgantao, Wed, 09 May 2012 23:31:38 -0400
OK, I will try this as soon as GMM informs us that the ticket system is back online :)

SpaceRay, Wed, 09 May 2012 23:38:28 -0400
OK

If this could be happening because of a bug inside FF no one would be able to help you other than the FF team.

GMM, Thu, 10 May 2012 03:29:23 -0400

Quote
SpaceRay wrote: you could contact the FF support and as ask to the experts at Filter Forge

I've already asked the developers to see into this.

Carl, Thu, 10 May 2012 05:08:54 -0400

Quote
GMM wrote: I've already asked the developers to see into this

Just don't blame the poor endangered Gorrilla :D

Morgantao, Thu, 10 May 2012 07:24:32 -0400
I think we should zap the gorilla with a cattle prod a few times to see if it makes it work faster :D

SpaceRay, Sun, 10 Jun 2012 14:36:39 -0400
today is one month since the last news about this "problem", bug or whatever

Please, is there any NEW information about this ??? :?:

Quote
GMM wrote one month ago I've already asked the developers to see into this.

As it has already gone one month since this, have they been able to test this and see why this is happening ?

Thanks very much for your help

SpaceRay, Mon, 11 Jun 2012 01:52:13 -0400
I think this would interesting to be able to optimize filters more, because perhaps this is some kind of bug or unusual thing that happens only when combining certain components or mixing them and makes them very slow and could help to make them faster if it is discovered why this happens.

Kraellin, Mon, 11 Jun 2012 12:55:29 -0400
morgan, post up an example filter. easier to digest that way.

Morgantao, Mon, 11 Jun 2012 14:52:41 -0400
OK, here it is...
First, look at the top post, and get the 2 filters. Test them individualy, and see how long things take to render on each of the filters.

Then download the filter attached here, and test it. All it is, is Carl's gorilla filter, only with an option to have the result run through my shadow\highlight filter.
If you check the "++" option, you use my filter too. If you uncheck "++", you use Carl's filter as is.

SpaceRay, Fri, 22 Jun 2012 08:01:34 -0400
Still no news about this, I think we will have to call Sherlock Forger to be able to solve it. :D

SpaceRay, Wed, 06 Feb 2013 16:57:07 -0500
Anyone knows the phone or email of Sherlock Forger so he could investigate and know what is happening when you mix two filters and you may probably get unusually much higher render times?

I have seen that the same as it happens here, it can happen also with others combinations of filters that are good separate and when joined in one filter, the render time is much higher than the sum of the two.

To be able to detect and investigate and know ourselves what could be happening, I think that perhaps one way could be to have the ability to know the render times of each component in the filter editor as suggested very well by Betis in this thread --> Workspace Rendering Times

Vladimir Golovin, Thu, 07 Feb 2013 04:52:09 -0500

Quote
Try linking several edge detection components and witness your system come to a grinding halt. Try running the same number of those edge detection components in parallel and it's no problem at all.

Nebukadnezar, you're completely correct regarding the exponential growth.

This particular case is a side effect of FF's sample-based architecture:
http://www.filterforge.com/more/help/...cture.html

It has nothing to do with memory and caches since Edge Detector doesn't store any caches, it's a 100% sample-based component.

Here's what happens:

To evaluate a single output sample, the Edge Detector shoots multiple samples into their Source input. Let's say the number is 8, and let's assume the components are chained sequentially.

The renderer requests a single sample from the first Edge Detector in the chain.

The first Edge Detector begins to evaluate the color at requested sample coordinates and shoots 8 samples into its Source input.

The second Edge Detector does the same (shoots 8 samples into its Source input), which repeats 8 times.

Et cetera, et cetera.

What you have here is essentially an exponential recursion which grows, roughly speaking, at O(8^n) where n is the number of Edge Detectors sequentially chained together.

Vladimir Golovin, Thu, 07 Feb 2013 05:17:59 -0500
Another possible source of slowdown with merged filter trees is the evaluation of image information beyond output image boundaries. Blur and Noise Distortion are two examples of components that require a larger "input area" than their output.

By passing the intermediate result as a saved bitmap, these beyond-the-border calculations are replaced by a crude bitmap mirroring / tiling trick (because there's no other way to obtain beyond-the-bounds information from a bitmap), and this trick is extremely fast.

SpaceRay, Sun, 24 Feb 2013 06:39:51 -0500
Thanks very much for the answer Vladimir, will have to check and try this out.

Any possible suggestions and advice to avoid this great slow down when you mix components ?

What you should take care and avoid to use combined?

I mean what combination of possible components will make the mix make FF slow down and what alternative could there be more optimized?

Thanks for any help

SpaceRay, Sat, 23 Mar 2013 07:11:42 -0400
Please, anyone else is also having any possible misterious problems and the render time rises very much after combining fast components or mixing two filters?

Sharandra, Sat, 23 Mar 2013 08:40:08 -0400
Ray, there is an entire article on that topic in the Wiki. I suggest you go read that.:-)

SpaceRay, Sat, 23 Mar 2013 19:10:18 -0400

Quote
Sharandra Ray, there is an entire article on that topic in the Wiki. I suggest you go read that

Sorry Sharandra that I did not know, please could you be so kind to tell me which article of the wiki are you refering to?

Thanks

Betis, Sat, 23 Mar 2013 19:24:10 -0400
SpaceRay, Here's this Wiki Page and also this one.

Spiral Border

xirja — Sat, 23 Mar 2013 16:16:21 -0400

xirja, Wed, 20 Mar 2013 15:52:31 -0400
Trying to get something like this:

Sphinx., Wed, 20 Mar 2013 16:07:53 -0400
Just hook up the color node and select white. Anything else? (I guess you ran into the image limitations in the forum)

xirja, Wed, 20 Mar 2013 16:24:27 -0400

Using the Solid Fill mode technique as seen in both:
Cell Area
Local Vortecies

So far, I have this:

xirja, Wed, 20 Mar 2013 16:36:35 -0400
But how to fix the boundary between the curves, i.e. what sort of curve needed to attach to the gradient?

Also, it would be nice to be able to 'squarify' to get:

xirja, Wed, 20 Mar 2013 16:40:40 -0400
Looking to further this along the lines of:

http://en.wikipedia.org/wiki/Navier–S...Turbulence

ThreeDee, Wed, 20 Mar 2013 17:22:50 -0400
I don't quite follow what you're trying to achieve. Maybe a photoshop approximation of the intended results?

xirja, Wed, 20 Mar 2013 18:42:25 -0400
So, I have 3 goals:

1. To get the bulge smoothed between the spirals (using the above ffxml attachment), so that the 2nd graphic looks like the 1st with respect to the transition between spirals

2. To be able to 'squarify' the same filter to achieve the 3rd graphic with respect to the general squared morphology. I know about masking and coloring, etc.

3. To then be able to move beyond the square grid of vortex centers and have a smooth iterative flow. Thinking about smoke stream eddies.

xirja, Wed, 20 Mar 2013 19:43:02 -0400
As close as I can get:

xirja, Wed, 20 Mar 2013 19:43:24 -0400
#2

xirja, Wed, 20 Mar 2013 19:44:32 -0400
Would hate to submit this with such glaring imperfections. :|

xirja, Thu, 21 Mar 2013 09:45:19 -0400
Having some fun altering the boundaries for some ornamentation though:

xirja, Fri, 22 Mar 2013 18:38:06 -0400
Using the last ffxml attachment, these areas circled in red are the problems I am looking to straighten/smooth (preferably with a curve and not a blur) to look like the very first graphic in this post:

xirja, Fri, 22 Mar 2013 18:39:27 -0400
Undulating areas more clear to see in the 'squarified' version:

xirja, Fri, 22 Mar 2013 18:51:51 -0400
Prior fun with altered boundaries using this filter:

xirja, Fri, 22 Mar 2013 19:05:03 -0400
oh, and if the ffxml examples are unloadable, they were made in FF4.0 Beta 1&2. ;)

Indigo Ray, Sat, 23 Mar 2013 08:22:46 -0400
I'm not sure what all that stuff feeding into the bomber was doing... In my method, the circles work fine. The diamonds have little squares at the connections, which come from the gradient we're rotating. See what I mean (FF4 Beta):

xirja, Sat, 23 Mar 2013 16:16:21 -0400
That method is great for constant connected spirals.

The logic for the tint color was so that there would be counter rotation and amplitude differences based upon the distance from the center line. Thinking ahead about fluid dynamics emulation (a crude example):

Strangely using the bomber is 67% faster than using a blend(multiply) method.

This might be one of those problems that can't be solved without relaxation of splines or something. See how noticeable the abrupt transition is on the loosely would spirals on the left? My only idea is to do an offset at the intersection points.

The Secret of Lookup

Burt — Thu, 21 Mar 2013 11:00:01 -0400

Burt, Wed, 20 Mar 2013 20:10:03 -0400
I am poking at some samples which ThreeDee has graciously posted in this thread and I just don't fully understand what Lookup is doing and what data it is passing. I am trying to get one Lookup component, which is storing the results of a Smudge operation with scale abd rotation etc and use that data to scale, rotate and smudge a second object so that I can blend with the results of the first -basically multiple types of buildings generated by seperate Smudge operations but I want to avoid using several Smudge components.

I have played with the examples and even think I understand but then the implmentation is wrong so I really don't understand what is going on. After the Smudge is fed into Lookup what data does it contain? Does it contain all the scale, rotation and smudge information that precedes it in the chain?

I want to apply that same scale, rotation and smudge to a simple polygon to start. I assume the results from the first Lookup need to be passed into a second Lookup or other Transform component that accepts X,Y values. I tried pluggin a simple polygon into the second Lookup in each input and tried every permutation I could think of but can't seem to get it exactly right.

I then created a second Smudge chain and it works, so then I tried to remove the second Smudge component, assuming the 'chain' up to taht point was good and substitute it with a Lookup (and each of teh transform components, to pass the data from the first chains Lookup -again, never exactly right.

Anyway I'd appreciate more info on how the Lookup component works as it seems like a vital component in any significant filter construct. I can just cut and paste the existing example by ThreeDee but I need to understand it better.

Sorry for the many posts.

ThreeDee, Thu, 21 Mar 2013 05:02:00 -0400
Lookup simply looks at the Source and picks up what image is at the coordinates given by X and Y inputs. Zero (black) is the top and 100 (white) is the bottom in Y coordinates -- left and right respectively in X coordinates. Gray values are the positions in between.

In it's simplest form, if you have a gradient from black to white, vertical for Y and horizontal for X, the lookup result is same as Source. If you flip the gradient, you get a flipped image. If you curve the gradient, you get a curved image, etc.

So those two gradients are the coordinate system (x and y coordinates) for Lookup. Anything you do to them, you can then do to any source image by Lookup.

Does that clarify?

Burt, Thu, 21 Mar 2013 11:00:01 -0400
It does - thanks. I did some reading about mapped inputs and have a better grasp of what is going on now. I was imagining the components passing data as distinct values rather than a grayscale map ie rotation:30 scale: 33% smudge:trail:20 etc

Hints and Tips on how to get the most out of the Loop component

SpaceRay — Thu, 21 Mar 2013 04:28:55 -0400

SpaceRay, Wed, 20 Mar 2013 02:48:28 -0400
Now that the FF 4.0 Beta 2 included the very good, awesome, interesting and useful Loop and iteration component, and having seen that many of you already know well how to use it in many ways and have many ideas about what to do with it I want to ask the following:

Please, could you suggest and give ideas, suggestions and Hints & Tips on how to get the most out of the Loop component?

Quote
(from beta 2 page) Loop is immensely powerful

As it is very powerful, it would be good to give ideas, sugesttions, hints, tips and comments on how to be able to use the new loop component at its best and what really can be done with it.

Also thought that here could be possible to have some filters examples using the new loop component uploaded to know better and learn how to use it better.

Quote
(from beta 2 page) However, proceed with caution. At the core of the Loop component lies recursion, a concept adapted from programming languages with all its benefits and hazards (such as potential exponential growth in rendering time). Unlike many other Filter Forge components that produce visible results no matter how you connect them, Loops require a more deliberate approach. That's why we put them into the new Advanced component category, together with Scripts

Quote

(from beta 2 page)

However, proceed with caution. At the core of the Loop component lies recursion, a concept adapted from programming languages with all its benefits and hazards (such as potential exponential growth in rendering time). Unlike many other Filter Forge components that produce visible results no matter how you connect them, Loops require a more deliberate approach. That's why we put them into the new Advanced component category, together with Scripts

I also thought to put this thread after reading this text from the beta 2 page, as it is perhaps a "dangerous" things if you do not know well how to use it and to avoid in some way to run into problems

Also as said this component requires more deliberate approach as is not simply connecting components to it, and would be good to know how to use it in the best possible way

LOOP COMPONENT OPTIMIZATION

Would be good to know about how to optimize the process of the iteration so it keeps inside an efficient and good level and don´t go into problems or becomes too slow.

What components would be good to use connected to the loops, and what other components would be perhaps a bad idea to use with this.?

Skybase, Wed, 20 Mar 2013 03:22:59 -0400
Let me keep it very simple: Don't over use it, use only when needed. If you know something that renders faster, use that instead. No need to be fancy at all, be clear and simple.

Quote
What components would be good to use connected to the loops, and what other components would be perhaps a bad idea to use with this.?

FilterForge already solved that for you but not letting you use bitmap based components.

Basic filter production rules otherwise apply.

Vladimir Golovin, Wed, 20 Mar 2013 04:21:42 -0400
Loops do add some overhead. As I just tested, on a simple Loop subtree (Free Ellipse + Blend) on 64 iterations the Loop version rendered about 1.5 times longer than the unrolled version.

However, with a heavier subtree, Loop seems to be faster than the unrolled version, about 8% faster for the attached example.

In other words, don't be afraid to use Loops (but watch out for exponential recursions!)

SpaceRay, Wed, 20 Mar 2013 05:28:11 -0400
Thanks for the comments

Quote
Vladimir Golovin Loops do add some overhead. As I just tested, on a simple Loop subtree (Free Ellipse + Blend) on 64 iterations the Loop version rendered about 1.5 times longer than the unrolled version

Good to know, so it seems that not always is good to use the loop component instead of the unrolled (non-loop) version

Quote
Vladimir Golovin However, with a heavier subtree, Loop seems to be faster than the unrolled version, about 8% faster for the attached example

So, considering this, it is only better to use loop on more complex iterations and not on simple ones, but then where would be the limit between the best way to use the loop component or the unrolled version?

Perhaps I think that the only way would be try and test the render speed of both and see wich one works faster, or there is a number of minimum iterations needed for the loop component to be faster anyway?

Vladimir Golovin, Wed, 20 Mar 2013 05:40:56 -0400
Unless the rendering time is already too long, I'd just use the Loop anyway, as it's more flexible, compact and convenient. Unrolling is ugly.

SpaceRay, Wed, 20 Mar 2013 05:49:34 -0400
Thanks Vladimir

Quote
Vladimir Golovin Unrolling is ugly

yes :D is much better to see the compact version than the unrolled one :D you feel overwhelmed with the great amnount of components.

LOTUS PATTERNS FILTER - USE LOOPS INSTEAD?

I have this filter Lotus Patterns by Tim2501 and this would be a great candidate for using the new loops component, BUT, would be useful and better to use the new loop component and would be able to get any better render speed and be more efficient, or would not be good and justified in this specifical case?

Vladimir Golovin, Wed, 20 Mar 2013 09:07:35 -0400

Quote
would be useful and better to use the new loop component and would be able to get any better render speed and be more efficient, or would not be good and justified in this specifical case?

Captain Obvious here: Try it yourself and see if it's better or not.

(I deleted some posts below. Spaceray, please don't clutter up the discussion.)

Crapadilla, Wed, 20 Mar 2013 13:42:40 -0400
Speaking of Unrolling:

Will there be a way to 'bake' a loop's subtree into a corresponding 'unrolled' filter tree? Via an 'Unroll Loop' menu option, perhaps?

I'm asking because this could potentially be a great way for filter authors to debug loops.

Vladimir Golovin, Thu, 21 Mar 2013 04:28:55 -0400
Dilla, no unrolling is planned. For debugging, you can use the Iteration (Preview) slider.

The Secret of Smudge?

Burt — Wed, 20 Mar 2013 10:52:00 -0400

Burt, Thu, 14 Mar 2013 13:20:34 -0400
I used Smudge in a few things but don't truly understand it. I am playing with a 'greebely' type filter but having a hard time understanding what inputs will get me nice well defined blocks and elevation. I saw the 'noise buildings' filter and he seemed to get great height and flat tops. My results sometimes are good and sometimes not but laregely just chance and tweaking until something happens. So I am trying to 'raise' a simple square (attached). Can someone explain or show me a simple example of how to convert this into a simple, well defined (sharp angled) cube?

Hopefully the dumbed down explanation will allow me to apply it better.

Thanks.

ThreeDee, Thu, 14 Mar 2013 15:39:14 -0400
Here's my advice.

You can't really get totally clean edges with smudge as it doesn't really smudge at 100% density even at 100% density, so the best approach (as discovered by Sphinx) is to smudge the coordinates used by the Lookup component and use Lookup to distort the actual image.

Some further notes in the filter itself.

Burt, Thu, 14 Mar 2013 16:04:29 -0400
Thanks for the sample. I'm going through it now. I'm going to have to explore the Lookup component a lot more too it seems. Thank you.

[PS. Just noticed those disconnected controls were used to add instruction for me - thansk for that detail]

Burt, Thu, 14 Mar 2013 19:12:07 -0400
Well going to take a while to really understand this I see - right now I'm like a monkey sticking his finger into the electric socket to see what happens. But I did have some fun using your code sample and plugging stuff in - looks like a ciy in there. Thanks.

Burt, Mon, 18 Mar 2013 13:32:53 -0400
I have a followup question to this if anyone cares to answer and if it is possible. I am trying to use the basic operation from the file posted by ThreeDee above, but want to generate additional objects - say with the polygon tool- and then use the output from the Lookup component to smudge and then correctly orient these new objects so that they can be stacked on the original smudge results. So in the example above we have one cube, can a second or even a third smaller cube be stacked on top (something like this image) using the output of the existing Lookup component so as to use only one Smudge operation?

Thanks.

ThreeDee, Mon, 18 Mar 2013 15:46:07 -0400
You mean like this?

It is really a matter of what you feed into the smudge. It is really just a height map of sorts.

Burt, Mon, 18 Mar 2013 17:17:33 -0400
Thanks for the response ThreeDee. Not exactly what I meant. It is a heightmap but correct me if I am wrong, the total max height of the 3 cubes will be the same as the total max height of 1 cube - white represents some sort of constant maximum height is that right? I want to build higher -so these buildings I might want to add an antenna or a cooling tower, or a gun turret looking things. So I want to do the max, then do it again and combine.

Burt, Mon, 18 Mar 2013 18:09:14 -0400
I posted this is another thread but you can see some solid white objects on top of these which I did by limiting the max color for the other structures but I was hoping to try some small solid white squares to get antennas, maybe try to get something that looks a little like pipes, and eventually skylights (and dare I hope windows though that would be another direction). Just enough to fool the eye a little more.

ThreeDee, Tue, 19 Mar 2013 02:58:20 -0400
With Max Trail set at 100, you can get the height all the way to the top of the image (Unless you touch the "Size, pixels" slider). It's not totally clean and you start losing small details, but it does work.

Burt, Tue, 19 Mar 2013 10:20:53 -0400
I didn't realize that thanks ThreeDee. Dare I ask one mroe related question? There isn't enough info in the Help file about the LookUp component. Is it possible to say save the results of a Smudge of a Noise component in the Lookup and then use that LookUp to smudge a second or even a third Noise component? Basically wondering if it is possible to avoid using multiple Smudge components.

Thanks again.

ThreeDee, Tue, 19 Mar 2013 11:17:58 -0400
Yes, when done this way -- i.e. by Smudging the coordinate system for the Lookup component (those two gradients that are fed into the X and Y inputs of the Lookup), you can use the result for any number of image sources.

One Smudge to rule them all.

Burt, Tue, 19 Mar 2013 12:33:12 -0400
Thanks a million ThreeDee. Just what I needed. :)

Burt, Tue, 19 Mar 2013 20:45:56 -0400
Well looks like I will be playing with this for a while so I uploaded the most basic version to the Snippets section of the Library -before I make changes and break it. :)

I attached a copy of the filter and do have more Smudgy questions if anyone cares to answer - don't want to abuse. :)

Right now the filter won't save a non-square image correctly - 600h x 1200w. I removed the one Rotation component because of what I read about Orthagonal view and it gets a little closer but still not right - besides I hate to lose the various angles but maybe that is unavoidable? I tried setting the various components XY Achor to Image Bounds thinking that might be to blame but the image then looks like a mountain range - all the 'city' definition is gone.

My only thought is maybe I can 'trick' it by making a 1200 x1200 image then somehow doing a crop somehow before the render so the output image is 600x1200 (or 1200x600 never remeber which goes first).

Thoughts?

ThreeDee, Wed, 20 Mar 2013 02:55:42 -0400
I replaced the profile gradients with free gradients set to "image bounds" and set the Lookup component to "image bounds" as well. Seems that doing that with Rotate component doesn't have the desired effect; it rotates around the first square origin point, but otherwise looks alright.

Burt, Wed, 20 Mar 2013 10:37:19 -0400
Thanks for looking into this. I can't open the file - need newer version. Are you using the 4.0 Beta? I can recreate it using what byou said though so no worries. Thanks again.

Burt, Wed, 20 Mar 2013 10:52:00 -0400
Yep. thanks that did it. Woohoo!

Please, How to control the SPACING between particles in Bomber FF 3.0?

SpaceRay — Wed, 20 Mar 2013 06:05:44 -0400

SpaceRay, Sat, 18 Feb 2012 15:17:04 -0500
Hello, I have been looking at the help guide to see if there is a reference to how to determine and control the spacing between the particles of the Bomber of FF 3.0 component, BUT regretably I have not been able to find it.

I want to make that none of the particles are over or below any of the others particles, and that all are at the same levels, and could get joined toghether in the best possible space.

There is Size, rotation, chance, opacity, randomization, offset X Y, Squash, tint, color, repeat, density, ... But as far as I know NONE of this controls how much space should be between each particle.

I suposse that probably you could use another external component to be able to control this but do not know how.

Is also like the 3D option of "colission detection" so none of the particles invades the space of another.

I have used the www.shapecollage.com software and there is a slider for controling the spacing between all the "particles" or photos.

Also I have been looking for a filter that could have this inside to see how it is done, but I have not found it yet.

The example I can give for this is this one

Morgantao, Sat, 18 Feb 2012 15:57:10 -0500
This is a good question. Would be cool if someone post a snippet or filter that let's you use a bomber in this way.

SpaceRay, Sat, 18 Feb 2012 21:43:23 -0500
Thanks Morgantao.

I have tried many ways with the bomber setting but as far as I know and my tests you can´t control the spacing with any of the MANY settings available in the bomber component.

So I think that if this is possible, it probably be done through another external component (or group of components) connected to the bomber.

I have been also looking through lots of filter in the library using the the Filter Library Browser but I have not been able to find one filter that could have bomber graphics that are NOT overlaid and mixed toghether.

SpaceRay, Tue, 21 Feb 2012 09:36:21 -0500
Loking at many filters to see if I could find a way to have a control over the spacing between the particles I have not found yet any solution, becuase on all the filters seems to have always a random component, and none of them have a specific control over the spacing between the particles, even not looking at the components involved in the bomber making, at least I did not see it.

It is strange that the Bomber component has the greatest amount of controls and options BUT does not have a control for the spacing between particles :?:

I am beginning to think that this is because all the bomber is based on RANDOM calculations so it would not be possible to add a spacing control if the particles are NOT themselves controlled.

SpaceRay, Tue, 28 Feb 2012 16:12:57 -0500

Quote
This is a good question. Would be cool if someone post a snippet or filter that let's you use a bomber in this way.

Is a very good question, but one that seems that nobody knows the answer, and FF does not want to answer too because from my investigation and study of many filters and how the bomber works I think that this is NOT possible with the actual bomber as it based mainly in Square results in grids that are defined in layers inside the Bomber, but the particles are not controlled in a way that you can measure the SPACING between them and is really something random.

I wish I and hope that I would be wrong and as you said, would be cool if someone could put a snippet or filter to show that this IS POSSIBLE.

I will ask this same question to the Genetica forum and see what they can answer (not about FF)

Morgantao, Tue, 28 Feb 2012 16:29:54 -0500
Do tell us what they say.

SpaceRay, Tue, 28 Feb 2012 16:51:05 -0500
I have already posted the question in the Genetica Forum, let´s see if someone there knows something about this and I will tell here.

SpaceRay, Wed, 29 Feb 2012 09:00:11 -0500
Until know I have got one answer only, and he think that this is possible, but he only gave a quick answer "You might like the regions node for this, that doesn't overlap."

I will have to see it in the user guide how this works and how can this be used, as I do not know what the region node is.

CFandM, Wed, 29 Feb 2012 13:40:48 -0500
See the problem is...They might give an answer using genetica not FF....There is no region node in FF....Before you put the image or shape into the bomber you can create a way to rotate. shrink and move that image or shape...Then you have a way to do it manually and it shows in the bombers results...Then the bomber is just an advanced multi-blend,,,,,

SpaceRay, Wed, 29 Feb 2012 14:29:31 -0500

Quote
Spaceray wrote: I will ask this same question to the Genetica forum and see what they can answer (not about FF)

Quote
CFandM wrote: See the problem is...They might give an answer using genetica not FF....There is no region node in FF....

As said above, I went to the Genetica Forum to ask if this was possible to make inside Genetica and not inside FF. I would not ask genetica problem solving questions in the FF forum. If this can be done in Genetica would be the same as in FF.

BUT perhaps the solution could help probably in some way to "copy" it to FF if there could any similar or equivalent component. If the solution is through "region node" I do not think would be possible in FF.

Quote
Before you put the image or shape into the bomber you can create a way to rotate. shrink and move that image or shape...Then you have a way to do it manually and it shows in the bombers results...Then the bomber is just an advanced multi-blend,,,,,

YES, I thought someting similar to do it before Bomber, but then I do not know how to make that they doesn´t overlap after the blend with bomber (or anything else) none of the letters should be overlaid over the other.

USE SHAPE COLLAGE SOFTWARE

I think that the best way could be to use the Shape Collage software that HAS SPACING and can accept unlimited photos, letter images...

I will try it this way and see what I can do.

Thanks very much for your help

Morgantao, Wed, 29 Feb 2012 18:12:18 -0500
SpaceRay, even in Shape Collage it's not always possible to get the effect you're after.
You may set the spacing, but then you'll have areas with bug "holes" between the images, while in other areas the images will still overlap eachother. By the time you add enough spacing to have no overlap in the "crowded" areas, the areas with less images will have alot of white space around the images.

The only way to do it then is to get the Pro version, that allows you to mnualy drag images from place to place, and get some of the overlapping images in the crowded areas to the less crowded areas. But that could be alot of work, and what you want is something that does it automaticaly.

SpaceRay, Wed, 29 Feb 2012 22:22:38 -0500
THIS EXAMPLE ABOVE HAS BEEN DONE MANUALLY ??? :?:

I agree and think and have seen that the typographic patterns shown in the WORDS Typography artistic Collage thread can be made by hand and would not be so difficult and would perhaps would be possible to make,

BUT HOW would you do the above example MANUALLY BY HAND and position one by one EACH of the letters ? I do not think that this has been done ONLY manually. How could you position, rotate, and make them in such a organized way ? How much time would it take ?

THIS ABOVE EXAMPLE IS A SEAMLESS PATTERN :!: :!:

There something MORE important, misterious and suspicious, HOW can you make MANUALLY and placing each letter by hand to make A SEAMLESS PATTERN ????? ;)

BY the way, THE WORDS pattern in the thread I have put are ALSO SEAMLESS!!! :?:

This is why I believe that there must be some kind of software making for this that can be able to make this SEAMLESS pattern, and this is WHY I though first on Filter Forge

Please look below the example in this post to show that this is a SEAMLESS PATTERN so I doubt much that this is made by hand position each one of the letters

Quote

Morgantao wrote:

SpaceRay, even in Shape Collage it's not always possible to get the effect you're after.
You may set the spacing, but then you'll have areas with bug "holes" between the images, while in other areas the images will still overlap eachother. By the time you add enough spacing to have no overlap in the "crowded" areas, the areas with less images will have alot of white space around the images.

YES, very well said, and I agree and you are totally right, and although it has the spacing control, the distibution is totally random, and can´t control the overlap, BUT is a little easier with PNG with transparent images because Shape Collage takes the photo as a whole.

Although there is a VERY important thing I have forgot and missed to tell

Quote
Morgantao wrote: The only way to do it then is to get the Pro version, that allows you to manualy drag images from place to place, and get some of the overlapping images in the crowded areas to the less crowded areas. But that could be alot of work, and what you want is something that does it automaticaly.

EXACTLY, this is what I forgot to tell, it would be really very difficult to do it with the normal version, BUT I have bought the PRO version last year when it had the price of 25$ (now have raised to 40$) and so you can get a distribution that
that the least overlapping or have more space than needed between the photos and then arrange them back in the right way inside Photoshop.

Of course that this is NOT the right way I want and as you said it woould still need some work, but at least is a POSSIBLE solution and available until there could be a better alternative.

As you say I wish that it would be any kind of "automatic overlapping detection" so when loading a bunch of images they would be distributed all over the empty space available and fit with a tight spacing and use the space in the best possible way.

I have searched "fill a shape with letters" and "text or letters plugin" and there is none that could works and look like the the first example above, and also any kind of tutorials that could show how to make this, but without sucess yet.

Morgantao, Thu, 01 Mar 2012 04:45:11 -0500
I'm not saying there's no software out there, somewhere that can do those typographical seamless tiles. I'm just saying I haven't found one yet, and I've been searching for a long time.

Most of the typographical tiles that you can find online are available as VECTOR files too. That suggests that they were made with something like Adobe Illustrator.
I can tell you there's a nice little trick you can do inside Illustrator that will help you make seamless tiles easily enough, and a tool that let's you randomize size and rotation of the letters (Sort of...). But you still have to manually move overlapping objects from eachother.
As far as I know there's nothing that will do the whole work for you automatically.
You can google search "Word cloud", that will help you find some online tools that can take words and arrange them randomly withought overlap, but they are not seamless.

I guess one could use a word cloud generator, save as SVG, import SVG into Adobe Illustrator, then make it seamless, but again, that's far from a fully automatic process.

Morgantao, Thu, 01 Mar 2012 05:47:16 -0500
Here's an example of a word cloud that I made using Illustrator. Note that it's a seamles tile.
As you can see in this example, I haven't moved any of the overlapping words yet, it's just the beginning of the process.
It's a semi-automatic process, as it's not totally automated, but it's not entirely manual.
A letter collage should be easier with this method, as it will be easier to manipulate the overlapping letters, than whole words.

Morgantao, Thu, 01 Mar 2012 14:33:50 -0500
Here's another example made with Illustrator.
This is also a seamless tile, like the last example, but this time I made sure there are no overlapping letters.
It's a little crappy, compared to the image SpaceRay post here on the top, but with patience it can be done better.

SpaceRay, Thu, 01 Mar 2012 17:35:53 -0500

Quote
Morgantao Most of the typographical tiles that you can find online are available as VECTOR files too. That suggests that they were made with something like Adobe Illustrator.

Perhaps is true that you have to use a combination of Adobe Illustrator, Photoshop and a word cloud online tool, but do not know if this would be for seamless.

Please look at this video I found ---> http://www.youtube.com/watch?v=2KfltVCwA_s

Thanks very much for all the suggestions and help you have done in the first of the 3 last posts.

From all the words on that post this ones below have the weight in GOLD filled with diamonds, don´t get me wrong, the others are also interesting and helpful, BUT this ones have helped much than any other thing could have helped. How can I thank you enough for having helped me with this words revealing the real name of "Word CLous" and the meaning of this.

Quote
Morgantao You can google search "Word cloud", that will help you find some online tools that can take words and arrange them randomly withought overlap, but they are not seamless. I guess one could use a word cloud generator

With this simple words that you have written, you have untied and solved a whole problem that have been sleeping and buried since 6 years ago !!! And you have opened a huge great door that I thought I will not be able to open in the way you have done.

First before explaining why I have said this, let me tell you that I am NOT searching for a seamless way to make a typographic pattern, I am sorry if you could understand it this way, and could be a confusion, sorry I am not good perhaps explaining it.

When I said above that the examples I have put are seamless patterns did not mean that I wanted to have seamless patterns, I wanted to show with this that they could NOT have been made MANUALLY, I can´t see anyone making this textures manually, so I had the idea that there SHOULD be any kind of software for MAKING this. I was sure that there must be some kind of tool to make this automatically.

AFTER SEARCHING ON GOOGLE "WORD CLOUD"

I could not believe my eyes and thought that this was a dream and that when I would wake up it would not be true. BUT no, is REAL and is TRUE, this two words are the exact definition and perfect name of what I have been searching for 6 years without any sucess !!!

There are not only one tool to make this, there are LOTS of them !!!! And they are very flexible, customizable, wonderful, useful and amazing.

And with this two words there are lots of websites that have reference to this !!!!

AND IS NOT ONLY ABOUT WORDS --- IS ALSO USEFUL FOR SINGLE LETTERS !!!!!!!

I have found that some of the tool for generating and creating work clouds can ALSO make "letter clouds" and would be with careful design and planning very similar to the example I have put in the first post.SO YOU HAVE SOLVED TWO PROBLEMS IN ONE

Quote
Morgantao I'm not saying there's no software out there, somewhere that can do those typographical seamless tiles. I'm just saying I haven't found one yet, and I've been searching for a long time.

I di not say that there was no sotware available either, and the same as you, I also haven´t found one yet, BUT the difference is that I was NOT looking for a seamless typograhic pattern maker, I wanted a normal typographic pattern texture creator that could be customizable and could follow any shape you could put, but I was happy it was just good with a plain square or rectangle.

I have seen examples in many places since 6 YEARS ago and was in love with these kind of typographical art, and always have been asking in forums, to typographic software companies, searching through lots of places all over internet, investigating HOW could be this found and whcih name could this have if there was one.

At the beginning of last year I was already very tired of this and I let it die the idea and decided to spot searching as was not possible in anyway, and nobody could help me find this.

Now 1 year after, recently by chance looking for another totally different thing I found again the example above and the words textures, and this was the ignition to make the flame desire for this to come alive again, and decided to ask it here in the forum.

AND THE SOLUTION TO ALL THIS GREAT AND LONG QUEST IS TWO WORDS "WORD CLOUD"

Thanks a million for having helped to me finding this !!!!!!!!!!!!!!!!

I will put some examples after

SOFTWARE FOR MAKING AND GENERATING AUTOMATICALLY WORD CLOUDS

http://www.tagxedo.com/

http://www.wordle.net/

although there are many more

Morgantao, Fri, 02 Mar 2012 04:44:01 -0500
Glad I could help :)
If you need seamless tileable typographic word or letter you can see the best result was using Illustrator, but for Non seamless those websites are perfect. They are easy to create and you don't have to do anything manualy.

SpaceRay, Fri, 09 Mar 2012 17:59:39 -0500
Later I will show ALSO how the some of the WORD CLOUDS software can ALSO make LETTER CLOUDS, specially the www.tagxedo.com :)

It has excellent control on spacing and tightness between the letters and even more

ThreeDee, Tue, 20 Mar 2012 05:41:53 -0400
In terms of trying do do something like this in FF, I would try and approach it with several bomber components one after another, wherein none of the letters touched in any one bomber, and subsequent bombers chance control would be determined by a (blurred and) thresholded version of the previous step. (Blurring would enable one to give a certain minimum distance to the particles.) Or one would have to use a "collision test" by first blending the outputs and then routing that to the chance input of another bomber particle with the same settings, taking out those particles that did overlap.

In theory.

Morgantao, Tue, 20 Mar 2012 10:03:52 -0400
I think it's safe to say that while in theory what you just said made a lot of sense, I didn't understand much of it and couldn't possibly start working on it in FF.... :D

SpaceRay, Tue, 20 Mar 2012 13:03:07 -0400

Quote

ThreeDee wrote:

In terms of trying do do something like this in FF:

I would try and approach it with several bomber components one after another, wherein none of the letters touched in any one bomber,

and subsequent bombers chance control would be determined by a (blurred and) thresholded version of the previous step. (Blurring would enable one to give a certain minimum distance to the particles.)

Or one would have to use a "collision test" by first blending the outputs and then routing that to the chance input of another bomber particle with the same settings, taking out those particles that did overlap.

Quote
Morgantao wrote: I think it's safe to say that while in theory what you just said made a lot of sense, I didn't understand much of it and couldn't possibly start working on it in FF.... :D

I think the same as Morgantao :D , this makes a lot of sense what you say, BUT I can´t understand HOW you can do this and try it in the filter editor. Sorry for my possible ignorance.

Skybase, Tue, 20 Mar 2012 21:59:58 -0400
I think when 2.0 beta was coming out uberzev made some test filter that allowed for distinct, non-colliding bomber example that pretty much can be used for this. Been a long time experimenting with stuff like this. Uberzev's example dealt with 4 bombers with really insignificant amounts of tint with distinct hue differences per bomber. It was then fed into a long chain of thresholds to separate out objects and what happens is clean, non-colliding particle. It probably will break if you feed in anything else though, meaning you can't put colored bombers, and it's not a simple construction. :(

ThreeDee, Thu, 22 Mar 2012 14:49:40 -0400
I can imagine that. I just put together this test with 6 bombers. The collision detection (at least the way I did it) is also doubling the bombers, since I first use it to detect collisions and then a copy of the same with the collision areas removed...

The main problem is that even with 6 bombers the letters are not filling the empty spots in the background and each subsequent one fills in less, as there is less and less empty space.

Anyways, maybe there is a way to improve it. Here is what I've got thus far.

ThreeDee, Thu, 22 Mar 2012 15:05:47 -0400
And here's the filter itself.

The controls for the required 'safe areas' for different bombers are manual, as I haven't figured out the math for repeat-amount-to-blur, so you can easily make the particles collide by setting the 'safe area' sliders at a lower setting.

TD

SpaceRay, Thu, 22 Mar 2012 16:12:04 -0400
Thanks ThreeDee will test it and see what you have done.

I think it will be really very difficult to make it really well

Skybase, Fri, 23 Mar 2012 11:04:43 -0400
ThreeDee, wouldn't min/max do just as well as long as input is black and white? This makes me wish for feedback / iterator nodes.

Edit: I meant minimum and maximum in the processing nodes. Not the math nodes.

ThreeDee, Fri, 23 Mar 2012 18:26:28 -0400
Hi Skybase / SpaceRay,

Yup, min/max would also work. Just have to be fully black to cull out the overlaps.

I also tried to do this by feeding the first bomber output (with blur and threshold) into the next bomber chance inputs, but it doesnät work as well, takes out many of the non-overlapping particles as well.

Here's another potential idea:

There could be a script-based computational approach where you generated an array for the bomber particles (matching the array size to the repeat value), gave each particle a random offset and rotation, and then calculated the maximum size the particle could be while not overlapping with the previous one to the left or above. The output of the script would then be routed to the bomber offset, rotation and size inputs. This might be just one script and a single bomber.

A problem arises from the particle shapes which the script could not easily take into account, so it would work better the more the particles filled their respective output squares.

TD

Morgantao, Sat, 24 Mar 2012 16:54:00 -0400
Sounds like a plan.
Call me when it's finished :D

ThreeDee, Sun, 10 Mar 2013 13:17:21 -0400
Just another attempt at making non-overlapping bomber particles. This time two tiles components are used to control the x and y offsets of particles. When adjacent offset go over a certain amount, the offset amount is limited so that the particles don't overlap.

ThreeDee, Sun, 10 Mar 2013 13:19:16 -0400
The filter. Note that three possible scenarios in x and y offsets are checked, as the particle can overlap with diagonally adjacent particles as well.

ThreeDee, Sun, 10 Mar 2013 13:21:46 -0400
Now it would be nice to be able to fill those empty spots as easily.

Skybase, Sun, 10 Mar 2013 13:30:29 -0400
Guessing there's no practical way doing it without having it hardcoded.

I was about to suggest doing it iteratively, as in have the next bomber use the x y offset coords used in the first bomber to determine where not to put the next set of particles. But that doesn't look like it'll work. Like this whole thing always seems pretty impractical to do in FilterForge to begin with. Like even with a script it seems limited.

ThreeDee, Sun, 10 Mar 2013 13:44:52 -0400
Yes, I was thinking about that too, but it gets rather complex fast. On the next iteration you'd have to check for overlaps with 9 different offsets.

Still trying to come up with a solution without a script, for by the time you want tens of thousands of dots in a script, you've run out of memory. And it is slow going in any case.

SpaceRay, Mon, 11 Mar 2013 04:30:48 -0400
What would be REALLY GOOD AND USEFUL is that there could be a NEW updated or different Bomber component that could have already this non-overlapping controls inside the component built-in already by default, and so you could decide to have as now happens overlapping shapes or not.

Would it be so difficult, complex or hard work to include a non-overlapping feature inside the bomber component?

Surely the great experts of the Filter Forge team could find a way to do it, although it depends if Vladimir wants to have it or not.

Burt, Mon, 11 Mar 2013 18:23:39 -0400
I'm not sure how the 'Chance' setting works in bomber but with a pure black and white chance map will a shape only be placed if it fits completely in the white area? If so maybe with your first iteration you generate a chance map (like the image above) to use on a second Bomber and then you can fill in some of the white areas?

ThreeDee, Tue, 12 Mar 2013 12:31:10 -0400
Unfortunately the problem is significantly more complex than that. My letter cloud attempt from 22 March 2012 utilizes a similar approach to the one you suggest, but even with 6 bombers in succession the fill-in is less than desired.

I don't currently have any method to do this with built-in components. I can only consider a script approach, but even that fails due to memory issues when I want to do many small particles. So, unless we get a built-in Bomber that can do this, it is going to be a rather complex scripting matter as far as I can tell. At the moment it seems that you'd have to try and separate every rendering thread (meaning those separate "boxes" that show up at the beginning of the render cycle) to consume memory. Not a simple task.

Burt, Tue, 12 Mar 2013 14:53:30 -0400
Well then we may have to wait until the Bomber with collision detection arrives.

Burt, Tue, 12 Mar 2013 15:15:23 -0400
Well I'm a little (or a lot) out of my depth here but maybe you could fake it? So you create several prebuilt combinations of small and large letters like my simple sample below. It seems that they don't overlap.

ThreeDee, Tue, 12 Mar 2013 15:33:59 -0400
Considering one fairly often has to "fake it to make it" with filters, all ideas are welcome. With a clever arrangement some such approach as you pose may even work. The biggest problem is all that empty space that we wanna get rid of while maintaining the perceived randomness of the placement of individual particles. Perhaps there is some simple solution along those lines -- or some other line -- that has just not occurred to me yet.

SpaceRay, Tue, 12 Mar 2013 15:41:42 -0400
Burt, good try and well done, but regretably I think that this is no solution, as I have seen that all are circular shapes although is true that they do not overlap, there is a lot of space inbetween the shapes and this is because the density is 1, if you raise the density to 2, they will fill the gaps but also will overlap.

Quote
Burt Well then we may have to wait until the Bomber with collision detection arrives.

I agree that perhaps the best would to wait and see if Vladimir or the FF developer team finds a way to make inside bomber a collision detection so you could choose non-overlapping if wanted.

Also think that maybe without this perhaps there is no easy way, or even is not possible without scripting, and perhaps even with scripting can be hard and complex, although really I do not know.

The collision detection should be for regular and irregular shapes, and then join them toghether as tight as possible as you can see from the letters shown here in this letters collage example

This could be the result of a high density bomber result with collision detection

Would be really cool and awesome that the bomber could have collision detection controls, spacing and be able to customize how tight the particles can be and so you could match and make the letters collage above here, OR any other thing that could be used as shapes as for example the FF life saver with transparent background.

SpaceRay, Tue, 12 Mar 2013 15:57:14 -0400
Burt, good try and well done, but regretably I think that this is no solution, as I have seen that all are circular shapes although is true that they do not overlap, there is a lot of space inbetween the shapes and this is because the density is 1, if you raise the density to 2, they will fill the gaps but also will overlap.

Quote
Burt Well then we may have to wait until the Bomber with collision detection arrives.

SpaceRay, Tue, 12 Mar 2013 16:03:29 -0400
And also continuing with examples that would help having collision detection and spacing control for bomber I can have this one here below

Also has been made by hand on photoshop one by one.

SpaceRay, Tue, 12 Mar 2013 16:04:56 -0400
And also one more tight packing sunflowers distribution that the source is just one and only image

Also made by hand in Photoshop

SpaceRay, Tue, 12 Mar 2013 16:21:32 -0400
It would be incredible, amazing, really great and awesome if there could be possible to have a Shape of Things (by CFandM) filter where the shapes loaded could be distributed and arranged in NON-OVERLAPPING way as shown in the two examples above here.

here is one example with the now available very overlapping version

Burt, Tue, 12 Mar 2013 17:32:45 -0400
@SpaceRay I am a fan of the idea of course. If you could have collisions also based on the depth map or a collision map you could have objects flow around things also or create negative spaces that were defined by the ojects flowing around their borders.

@ThreeDee wish I knew more as I like solving problems.

Burt, Fri, 15 Mar 2013 12:04:31 -0400
This may be another lame brained comment but it seems that applying curves to noise - like the Blocks- seperates and might be able to keep particles from overlapping? Maybe I am reading it wrong but this image shows the Blocks component with 100% contrast without a curve and then with one. It seems that by adjusting the curve you affect the distance around an object and avoid overlapping (only managed a quick example as I am stealing a moment at work to do this so some overlap but tweaking the curve changes that) - or am I reading the results wrong? If a curve could be used to define the space around an object to some degree then maybe adding a curve node to the Bomber would be useful?

SpaceRay, Wed, 20 Mar 2013 05:55:44 -0400
As I said already in the previous page, I would really love that the Bomber component could be upgraded and added a spacing control between the particles and also a possible collision detection (or whatever you want to call it) so it could be possible to make non-ovelapping particles available

Or perhaps find a way using a determined group of components already available to be able to have this spacing control and collision detection with bomber and have it as a group component

ThreeDee, Wed, 20 Mar 2013 06:05:44 -0400
This may be easier to achieve with the new loop component, but there will obviously be a limit to the number of particles.

Store in Curve>pattern>possible?

Burt — Sun, 17 Mar 2013 20:25:11 -0400

Burt, Sun, 17 Mar 2013 18:14:49 -0400
I noticed yesterday that if you input stuff into the green nodes on a curve that information is stored with the curve somehow. What I am wondering if it is possible to store a pattern in a curve and then use that to 'texture' say the individual blocks of a Block component? I know this might not make sense the way I am saying it but I was playing with a greeble 'city' of sorts and was wondering if I could 'texture' the individual house roofs for example. Anyway here is a simple image. It 'seems' like the gradient is applied to individual blocks but I'm not sure. I thought it might be possible to increase the complexity of the greeble output somehwta.

Anyway hopefully you understand - the ideas are clearer in the brain but always come out mushed when I try to explain.

Burt, Sun, 17 Mar 2013 20:25:11 -0400
Ok forget it. When I got back I tried building a roof pattern with shapes and it is only applied once. I thought it might be possible with some sort of curve magic that I'd get a nice little roof on each 'house.' Oh, well. Really do need a delete button on the forum. :)

Smear 3D

ThreeDee — Sat, 16 Mar 2013 11:05:00 -0400

ThreeDee, Mon, 30 Jul 2012 14:54:26 -0400
Thought we could have a thread on the subject of creating faux 3D images with the Smear component and related stuff.

ThreeDee, Mon, 30 Jul 2012 14:58:21 -0400
Here's the basic idea: Set to density 100, the Smear component "offsets" the image a given amount from its original location. When the Trail amount is controlled by an image, the offset amount varies accordingly, basically sculpting a sort of a 3D object. In such wise:

ThreeDee, Mon, 30 Jul 2012 14:59:41 -0400
My bad: the actual name of the component is Smudge, not Smear...

Sharandra, Mon, 30 Jul 2012 15:03:04 -0400
I´ll make a request to the FF team to change the name of the component for you :P

ThreeDee, Mon, 30 Jul 2012 15:11:47 -0400
In here, the image that controls the Trail amount is analogous to a height (or displacement) map.

Two other useful things to Smudge offset are:

1) the height map itself

2) a gradient

Both of these give you a "3D" bump map that matches up with the offset image in the first example.

ThreeDee, Mon, 30 Jul 2012 15:15:32 -0400
For a Simple (as opposed to Surface) filter, you could blend the output of the offset image and the offset height map and get this:

ThreeDee, Mon, 30 Jul 2012 15:20:09 -0400
For a Surface filter, you can use the Smudged Height map as height.

ThreeDee, Mon, 30 Jul 2012 15:23:58 -0400
Which enables you to control the lighting.

Since the Smudge component uses bitmap information, you get some artifacts, but they can usually be pretty well disguised by other means.

Sharandra, Mon, 30 Jul 2012 15:37:30 -0400
Nice tutorial ThreeDee :-) Been following along with something simple. Lots of fun in this :-)

SpaceRay, Mon, 30 Jul 2012 15:39:13 -0400
WOW!!! Thanks really very much for this interesting and useful tutorial, and is easy to follow and to make.

Please, Is this way how you have made the Vladimir´s Crysta and the Asteroid in the "other interesting ideas" thread?

ThreeDee, Mon, 30 Jul 2012 15:39:25 -0400
Another very handy 3D tool is the scale component.

If you make a gradient with 0 at the top and 2 (HDR) at the bottom, you have an instant perspective transformation tool.

If you change the percentage of the Scale component, you get anything between no perspective to full horizon-at-the-top-of-the-image perspective transformation.

ThreeDee, Mon, 30 Jul 2012 15:40:39 -0400

Quote
Please, Is this way how you have made the Vladimir´s Crysta and the Asteroid in the "other interesting ideas" thread?

I'll get to that eventually...

ThreeDee, Mon, 30 Jul 2012 15:50:11 -0400
Now, let me jump ahead and combine some of these things.

Notice that I have blended a vertical gradient (with 100% multiply) on the perspective image that feeds into the Trail inputs. This is so that the trail gets less and less as we go toward the horizon.

All of a sudden, we have a complete "3D" image!

ThreeDee, Mon, 30 Jul 2012 15:51:27 -0400
Which looks like this.

ThreeDee, Mon, 30 Jul 2012 15:55:53 -0400
In case some of the settings are unclear from the screenshot, here's the above filter.

ThreeDee, Mon, 30 Jul 2012 15:56:34 -0400
More later.

SpaceRay, Mon, 30 Jul 2012 16:05:27 -0400
Great !!! I will try to use all this you have shown here

:ff: :loveff:

Thanks very much again for making this

Ghislaine, Mon, 30 Jul 2012 20:06:21 -0400
Thanks a lot ThreeDee.

Sharandra, how did you made your beautiful image above ? It is so nice...

ThreeDee, Tue, 31 Jul 2012 02:10:29 -0400
Different particles give you different 3D shapes.

ThreeDee, Tue, 31 Jul 2012 02:11:13 -0400
Result of the above filter.

ThreeDee, Tue, 31 Jul 2012 02:14:20 -0400
Notice the Rotate component right after the bomber. It enables you to rotate the scene around.

ThreeDee, Tue, 31 Jul 2012 02:16:04 -0400
Here is the filter so far.

ThreeDee, Tue, 31 Jul 2012 02:52:20 -0400
If you make the background (the space between the objects) transparent (instead of black), you can reverse the Smudge with the result of turning the objects upside down. In this case, you will have to both change the angle of the smear by 180 degrees and select the Reverse box of the Smudge component.

ThreeDee, Tue, 31 Jul 2012 02:55:47 -0400
Since there are more than a couple of things I had to do, I'll just give you the new version of the filter to study.

I used Threshold to make the background transparent.

I added the same rotation to the checkered pattern to keep things aligned.

I also had to add a gradient to the bottom of the Trail input image so that the image would not wrap around from the bottom to the top.

ThreeDee, Tue, 31 Jul 2012 03:05:07 -0400
The nice thing about reversing the Smudge is that you can then combine the two and get objects that have height information in both directions.

Here I have additionally made the filter a Surface filter so I can use the height information to control lighting.

ThreeDee, Tue, 31 Jul 2012 03:29:16 -0400
Add color, ambient occlusion (wait) and behold.

ThreeDee, Tue, 31 Jul 2012 03:32:31 -0400
Here's the filter. I turned off AO as it doesn't contribute much. Without it, the filter is even reasonably fast.

SpaceRay, Tue, 31 Jul 2012 03:40:32 -0400
WOW !! This is amazing what you have done, and I was wondering how would be possible to make a FULL shape and not just half of it and you have done it now!!

Thanks very much for this very useful and interesting tutorials

I tried texturing the shapes with images, but the image is very distorted and wrongly mapped over the 3D shape :( I thought that perhaps is because the scale of the image must be reduced so is not shown as big as the checker pattern.

Sharandra, Tue, 31 Jul 2012 03:59:32 -0400
Cool, I´ll play with that tonight :-)

Thanks for taking the time to make these :-)

SpaceRay, Tue, 31 Jul 2012 04:03:43 -0400

Quote
Ghislaine wrote: Sharandra, how did you made your beautiful image above ? It is so nice...

I think that probably that could be have been made with a spiral shape, like for example the snippet Crapadilla´s Spiral Gradient

Here is something I have tried BUT is much less nicer and well done as Sharandra´s example

The Spiral snippet inside comes from another different version done by Crapadilla

Sharandra, Tue, 31 Jul 2012 04:28:26 -0400
Yup, was done with a render from Spiral Gradient :-)

Indigo Ray, Tue, 31 Jul 2012 08:37:21 -0400
Making 3d perspective filters into surface filters (see ThreeDee's silver and colorful shapes) isn't the most accurate way to do lighting. It makes the objects look flatter as they get farther away. Instead, we can use the derivative component to simulate light shining on the scene. Use the derivative component before using the scale component.

Indigo Ray, Tue, 31 Jul 2012 08:40:15 -0400
Keep experimenting and you can make something like this:

Indigo Ray, Tue, 31 Jul 2012 08:42:35 -0400
The filter for the first image of mine.

Ghislaine, Tue, 31 Jul 2012 10:03:55 -0400
Wow Indigo ! Another wonder ! Looking forward for the snippet of the second image (if you want to share it with us). The first one seems very good for creating sea floors. The second one should be very useful for creating beautiful landscapes.

Anyway, all I see in this thread is my cup of tea :)

Thanks ThreeDee, Indigo Ray and Sharandra for sharing with us. You are generous persons.

Crapadilla, Tue, 31 Jul 2012 10:10:21 -0400
Awesome thread! :beer:

ThreeDee, Tue, 31 Jul 2012 11:23:48 -0400

Quote
Making 3d perspective filters into surface filters (see ThreeDee's silver and colorful shapes) isn't the most accurate way to do lighting. It makes the objects look flatter as they get farther away. Instead, we can use the derivative component to simulate light shining on the scene. Use the derivative component before using the scale component.

Yes, thanks Indigo. It is definitely better to do the shading in other ways, as Surface filter both slows down the rendering and makes the artifacts much more visible. Derivative is a very good way to get fast, accurate shadows while keeping the filter a Simple filter.

SpaceRay, Wed, 01 Aug 2012 13:03:09 -0400

Quote
SpaceRay wrote: Please, Is this way how you have made the Vladimir´s Crysta and the Rawn´s Asteroid in the "other interesting ideas" thread?

Quote
ThreeDee wrote: I'll get to that eventually...

Thanks very much, I will be waiting to see how wonderful way you have found to make this amazing and awesome 3D texture mapping with this filters you have shown.

Do not worry, take your time and do it whenever you want and can, I am not in a hurry at all.

ThreeDee, Fri, 03 Aug 2012 05:51:56 -0400
Spherical 3D Displacement Previewer

Here is the displaced sphere texture previewer. I can't submit it to the library as it has external images as color inputs. In brief, you use it as follows:

1) Output (save) the diffuse, bump and specular maps from a filter of your choice
2) Load the diffuse, bump and specular maps into this previewer and adjust lighting to taste.

I submitted a "light" version of the same to the Library. It uses the current image only and creates the bump and reflection/specular maps from it.

SpaceRay, Fri, 03 Aug 2012 08:38:29 -0400
WOW!!! Thanks very much for this, I will make some tests and see how this works and what can be done with this.

:ff: :loveff: :ff:

I have put a new thread for showing examples made with this filter and so do not clutter and spoil this very useful, good and interesting thread.

Examples of Spherical 3D Displacement Previewer by ThreeDee

Quote
ThreeDee wrote: I can't submit it to the library as it has external images as color inputs. In brief, you use it as follows:

FF gives error when trying to find the external images and are not found in the filter, and if they are just plain colors, sorry that I do not understand this can´t be submitted to the library without the external images and only the color control using the color.

SpaceRay, Fri, 03 Aug 2012 09:36:00 -0400
HOW TO USE THIS FILTER OF Spherical 3D Displacement Previewer

Quote
ThreeDee wrote: In brief, you use it as follows: 1) Output (save) the diffuse, bump and specular maps from a filter of your choice 2) Load the diffuse, bump and specular maps into this previewer and adjust lighting to taste.

Thanks to the help given by Skybase I know now how to use this maps, and want to make here a mini tutorial for others users that probably do not know how this can be done

1 - Select a filter of your choice (MUST be one that have a "Lighting" tab next to "Settings" tab, so this means that this is a surface filter)

2 - Select the preset you want and make all the settings as you want

3 - When finished and ready go to the the following menu: Filter -> Render Maps -> Diffuse map

4 - Filter Forge will then render ONLY the diffuse map

5 - Go to File -> Save Image as and save it with the name you want but knowing that this is the diffuse map

6 - Now go to the menu Filter -> Render Maps -> Bump map

7 - Save this also in the same way

8 - Now go to Filter -> Render Maps -> Specular Strenght Map

9 - Save it also

10 - If needed you can also go to menu Filter -> Render Maps -> Specular Exponent Map

11 - Choose the size resolution you want the final result to have by using the File -> New Image

12 - And then load back these maps into the ThreeDee´s filter and adjust lighting and settings as you want.

A FOLDER FOR EACH GROUP OF MAPS

I suggest that you make a new folder with the name of the filter and then save inside the folder the 3 maps with the diffuse map, bump map and specular map names so it will be very easy to load them inside the filter.

RENDER MAPS GRAYED OUT ?

If the Render Maps in the Filter menu are not available is because this filter does NOT use the surface and lighting (does not have a "Lighting" tab next to "Settings" tab), so you will have to choose another one.

Here below is how it looks when render maps are not available

Lareieli, Fri, 03 Aug 2012 22:36:44 -0400
Thank you folks for all of this!

Sharandra, Sat, 04 Aug 2012 01:43:19 -0400
Threedee, with your spherical mapping, is there a way to rotate the sphere with the texture on it?

More specifically, I´m trying to figure out a way to change the view of my water lilys from a top to a perspective view.

Sharandra, Sat, 04 Aug 2012 03:23:16 -0400
Like this. Is that somehow possible?

ThreeDee, Sat, 04 Aug 2012 03:38:07 -0400
Hi Sharandra,

It is pretty easy to add rotation, but you won't see the back side since the filter doesn't render it.

Anyways, here's the filter with some modifications that allow you to rotate the sphere, scale down the image and invert the height map.

TD

ThreeDee, Sat, 04 Aug 2012 03:39:02 -0400
Here's what the lily looks like. I didn't rotate it to the back side.

ThreeDee, Sat, 04 Aug 2012 03:40:01 -0400
Crysta goes round'n'round.

Sharandra, Sat, 04 Aug 2012 03:42:22 -0400
Ah cool, thx a lot! :love: I´ll have a look at it :-)

Sharandra, Sat, 04 Aug 2012 03:44:04 -0400
Hehe, the rotating crystal is cool :-)

ThreeDee, Sat, 04 Aug 2012 04:05:31 -0400
I fixed the problem I noticed from SpaceRay's examples, the spherizing formula should be properly spherical now.

Hyperbolic Tiling Challenge

uberzev — Sat, 16 Mar 2013 08:14:56 -0400

uberzev, Tue, 07 Aug 2012 06:53:39 -0400

http://www.josleys.com/show_gallery.php?galid=325

http://en.wikipedia.org/wiki/Hyperbolic_geometry
http://www.bugman123.com/Hyperbolic/index.html

https://www.google.com/search?q=hyperbolic+tiling

SpaceRay, Tue, 07 Aug 2012 10:02:30 -0400
WOW!!! :eek: :eek: :eek:

This hyperbolic tiling examples look really amazing, jawdropping and awesome for me, love them very much!!

Thanks very much for putting this, love geometric things like this

Hyperbolic Tessellations

[img]http://aleph0.clarku.edu/~djoyce/poincare/tess_5_4.gif[/img]

[img]http://aleph0.clarku.edu/~djoyce/poincare/tess_4_5_l.gif[/img]

This Hyperbolic tiling seems to be more simpler and perhaps easier to do than the much more complex above, although may it be based the same formulae so perhaps is not

____________________________________

Uniform tilings in hyperbolic plane

WOW, really LOVE MUCH all this kind of hyperbolic tilings shown in this link here and here below!!!!

Is there any software than can make this kind of tilings for making creative graphics ?

Would be possible in Filter Forge ??

SpaceRay, Tue, 07 Aug 2012 10:36:50 -0400
part 2

SpaceRay, Tue, 07 Aug 2012 10:37:16 -0400
part 3

SpaceRay, Tue, 07 Aug 2012 10:42:06 -0400
part 4

SpaceRay, Tue, 07 Aug 2012 10:42:32 -0400
last part

SpaceRay, Tue, 07 Aug 2012 10:51:23 -0400
see this also ->Hyperbolic Planar Tesselations

SpaceRay, Tue, 07 Aug 2012 10:51:52 -0400
and the second part

SpaceRay, Tue, 07 Aug 2012 10:55:26 -0400
Tilings of Hyperbolic Spaces

lots of links about this topic

Skybase, Tue, 07 Aug 2012 12:04:33 -0400
I'd want to chew into this so badly

Check out: http://www.amazon.com/The-Beauty-Geom...0486409198

Great book on this stuff.

And yes it's very possible in this program. It was suggested earlier back a year ago or something... didn't get anywhere unfortunately....

Skybase, Tue, 07 Aug 2012 12:11:44 -0400
Otherwise for inspiration there's Artmatic 5 which does a lovely job with these kinds of things that don't exist in FilterForge. But annyway....

SpaceRay, Tue, 07 Aug 2012 13:49:51 -0400

Quote
Skybase wrote: Otherwise for inspiration there's Artmatic 5 which does a lovely job with these kinds of things that don't exist in FilterForge.

I have seen the web of this very interesting software and that can make very good artistic results
BUT THE GREATEST AND MAIN PROBLEM FOR ME IS THAT IS MAC ONLY :cry: :cry: :cry: :cry:

Why do not make them a windows version of this awesome software ??? :?: :cry:

Quote

From the Artmatic FAQ page

We are frequently asked whether our products are available for the WINTEL platform. Currently, there are no Windows versions of our programs. We understand that there is a desire for these programs on the Windows platform, but we are expert Mac programmers with little Windows experience.

We have decided to stick to what we do best which is create great Mac programs for creative artists. In the time that it would take us to port a single program to Windows we can create several great programs for the Mac.

We have not ruled out the possibility of porting to Windows but there are no plans at this time. We will update this item if we decide to create a Windows version of any of our apps.

uberzev, Thu, 09 Aug 2012 01:41:17 -0400
http://mathworld.wolfram.com/ConformalMapping.html

SpaceRay, Thu, 09 Aug 2012 02:04:48 -0400
Interesting link to HOW to make this Conformal Mapping BUT the problem is that this sounds like chinese to me :D

Conformal Mapping examples shown here below

In the above link go down on the page and you will find this beautiful and very good conformal mapping

uberzev, Thu, 09 Aug 2012 04:03:26 -0400
http://www.hypatiasoft.fr/Folder_Seam...ker_A.html

SpaceRay, Thu, 09 Aug 2012 04:05:46 -0400
Here I have found some more information about this topic that I think can be useful

Google images search for "conformal mapping" gives some interesting results (at least for me) and think is worth it.
_____________________________________
Bending Hyperbolic Kaleidoscopes by Vladimir Bulatov

Here you can find this conference about this topic with 119 PAGES of examples of this kind of Bending Hyperbolic Kaleidoscopes technique that looks very good and beautiful and is well done

Direct link to the 119 pages PDF

I recommend to download this PDF as is interesting to see al this examples
_____________________________________

Conformal Mapping history since 2006 to 2009

This is a good graphical timeline about this topic

What is curious and I did not know, is that it seems that the Droste Effect is done with Conformal Mapping !!!

Skybase, Thu, 09 Aug 2012 04:31:35 -0400
They're both drawn on complex planes. Which is why they're similar. Just the formulas for deriving at them is different. (Now I feel dumb typing that because I just stated the obvious.)

uberzev, Fri, 10 Aug 2012 05:22:52 -0400
http://www.youtube.com/watch?v=0z1fIsUNhO4
http://en.wikipedia.org/wiki/Moebius_Transformation

ThreeDee, Sat, 11 Aug 2012 00:45:42 -0400
Actual hyperbolic tiling is seriously tricky to achieve, but since we gotta start from somewhere, here's my first attempt to fake something similar.

ThreeDee, Sat, 11 Aug 2012 00:56:24 -0400
Attempt #2

ThreeDee, Sat, 11 Aug 2012 01:07:02 -0400
True hyperbolic tiling doesn't distort the image this much. It is more like a fractal.

ThreeDee, Sat, 11 Aug 2012 01:15:03 -0400
So, let's make it more so:

ThreeDee, Sat, 11 Aug 2012 01:20:31 -0400
At a different circular repeat:

ThreeDee, Sat, 11 Aug 2012 01:21:16 -0400
Even higher setting.

ThreeDee, Sat, 11 Aug 2012 01:28:43 -0400
This may be better.

ThreeDee, Sat, 11 Aug 2012 01:34:52 -0400
And one more, better balanced for less distortion.

ThreeDee, Sat, 11 Aug 2012 01:36:52 -0400
It is still fractal circular tiling, not hyperbolic.

SpaceRay, Sat, 11 Aug 2012 02:33:57 -0400
Cool and well done, although as you say is not hyperbolic, but I like much this results you have shown, specially the last 4 ones. It must come an expert like you to be able to make this. :)

uberzev, Sat, 11 Aug 2012 03:54:08 -0400
I'm obsessed with figuring this out. I'm lost though. I don't understand the "complex plane" math.

SpaceRay, Sat, 11 Aug 2012 04:20:49 -0400

Quote
uberzev wrote: I don't understand the "complex plane" math.

If you that can have the ability,experience and skills to be able to understand this must be really complex to make and learn. For me this is like telling me that I should go the moon and come back :D

Perhaps this is why they call it "complex plane" math, if it would be easier to understand would be "easy plane math" ;) :D

ThreeDee, Sat, 11 Aug 2012 06:37:34 -0400

Quote
Uberzev wrote: I'm obsessed with figuring this out. I'm lost though. I don't understand the "complex plane" math.

I don't know how to do this without scripting, and it would take days to write the script.

Even making a fake version with actual hyperbolic tiling (but without going into infinity at the edges) would be quite a challenge without any scripting.

I will be amazed if someone does it.

ThreeDee, Sat, 11 Aug 2012 06:40:14 -0400
The math for hyperbolic to euclidean transformation looks deceptively simple. I have no idea how to implement that in FF.

Besides, it is not clear from that article if the transformation is from and to the Poincare disc, which is what the original posting image is based on.

ThreeDee, Sat, 11 Aug 2012 06:52:50 -0400
By the way, it seems that this would be the "easiest" hyperbolic tiling -- or at least the most useful for square images -- as each of the tiles are hyperbolic squares, thus the image would tile seamlessly.

If you can make it, that is.

ThreeDee, Sat, 11 Aug 2012 06:56:02 -0400
I tried -- without any great success -- figuring out how to make x and y input maps for the lookup component that would effect this transformation. I got as far as the middle square :|

SpaceRay, Wed, 15 Aug 2012 03:59:06 -0400
I think that most probably to make this is much more complex than it seems, at least to program, and would need a complex custom software to make it probably, OR perhaps could be made in FF but would need someone that takes much time to build a script that could make this kind of complex maths and knows really much about this complex maths theory.

Skybase, Wed, 15 Aug 2012 09:28:06 -0400
It's all in the real and imaginary!!!!

If 100 people saw what we're doing about 80 of them will wonder what the and 18 will be going "I get it" and 2 of them will be saying "I CAN CLEARLY DO THIS IN FILTERFORGE. I KNOW I CAN. I JUST CAN'T."

Sounds like a case where scripting better deals with stuff like this.

Although if not in FilterForge there's UltraFractal.

ThreeDee, Wed, 15 Aug 2012 13:05:12 -0400
Righty. It gets so complex (no pun intended) the question becomes whether it is worth the time and trouble, when there are programs that already do this.

I am of the opinion it might be better to try and achieve something adequately similar without having to script it.

ThreeDee, Thu, 16 Aug 2012 02:32:08 -0400
Hyperbolic Tiling Success!

After relaxing the objective and considering other approaches, a solution emerged. What was making the solution escape me was attempting to get a mathematically accurate resolution, when we're really talking about just hyperbolic tiling without the necessity of making it "perfect."

There is a surprisingly simple solution for (4,4) quadrilateral tiling if you don't aim to make it mathematically perfect in every aspect.

Computer says "Yes!" :D

Morgantao, Thu, 16 Aug 2012 02:41:39 -0400

Quote
Computer says "Yes!"

I say "Hell yeah!"

Way to go ThreeDee, you're still the master :)
Can you use any existing tileable texture with it, and show us an example?

ThreeDee, Thu, 16 Aug 2012 02:53:42 -0400
Hi Morgantao,

Well, it's not actually quadrilateral tiling. There are always four tiles at each crossing point, but it is a mixture of 4-sided and 5-sided polygons.

I have a theory of how to get an image into those tiles, but I have yet to produce it in real life.

This is really just fresh out of the oven. The solution literally came to me when I was going to sleep last night.

I'll make some diagrams which explain the method.

TD

ThreeDee, Thu, 16 Aug 2012 03:35:18 -0400
Here's the construction of this pattern in visual form.

Take an upwrapped (unspherized) sphere with four quadrants.

Take the perimeter of the circle (the bottom line) and split in in three in each quadrant.

Draw overlapping circles centered on the meeting point of the perimeter and the lines that divide it into quadrants that go through these three splitting points.

Divide the resulting 12 sections of the perimeter into three.

In the same way draw circles that match these new points.

Repeat as many times as you deem necessary.

Spherize the result and behold hyperbolic tiling!

---

Since unspherized hyperbolic lines are not actually circles, you get some distortion which you have to deal with, but that is a smaller issue.

This method can be improved upon in various ways, but this is the basic idea.

CFandM, Thu, 16 Aug 2012 05:46:31 -0400
Hurray :hammer: :ff: ..Good work ThreeDee....I was also stuck on this one..... :)

Sharandra, Thu, 16 Aug 2012 05:53:29 -0400
Wow great job!! :D

uberzev, Thu, 16 Aug 2012 05:54:52 -0400
Here's a nice large image that could be good for experimenting with.

Skybase, Thu, 16 Aug 2012 06:04:53 -0400
WELL WELL WELL what do we have here :D

uberzev, Thu, 16 Aug 2012 06:08:10 -0400
heh, that image wasn't made with FF. :D
There's a simple windows program for making hyperbolic tiles

I'm using that image as a template for now.

If I can understand the pattern in which the arcs are placed and sized I can make an efficient FF version. What I've been doing is manually position everything.

ThreeDee, Thu, 16 Aug 2012 13:50:32 -0400

Quote
Uberzev wrote: If I can understand the pattern in which the arcs are placed and sized I can make an efficient FF version. What I've been doing is manually position everything.

Unfortunately it is not a repeating pattern other than around the symmetry axes of the center polygon. You would have to calculate every arc size and location with a script in any case. It seems to me after some study of the math behind proper hyperbolic tiling that uniform tiling (equal corner angle hyperbolic polygons) would require a recursive calculation for each subsequent arc once you know the previous one.

That is where I gave up on accuracy and settled for visual similarity.

I have managed to get the input image to (sort of) tile in this nonuniform hyperbolic tiling now.

It is anything but optimized or accurate, but probably as close as I will ever have the patience to get it.

Morgantao, Thu, 16 Aug 2012 15:18:43 -0400
WOW ThreeDee, how did you do that?
How did you write the visual guide in a language even I can understand?! :D

SpaceRay, Fri, 17 Aug 2012 00:14:43 -0400
WOW!! This is really great and beautiful ThreeDee, very well done.

Quote
uberzev wrote: heh, that image wasn't made with FF. There's a simple windows program for making hyperbolic tiles

Thanks for the link is very interesting and useful

SpaceRay, Fri, 17 Aug 2012 00:23:52 -0400
Sorry to ask this very basic question but, how do you convert the 2D line into a 2D circle?

Of course that I am very far away from all the knowledge about this that many of you have

In all the filters available in FF to make this are for 3D Spheres and not for 2D ones, and they distort the images as they try to simulate the 3D effect on a ball instead of a plain 2D image

Seam

Burt — Thu, 14 Mar 2013 16:02:32 -0400

Burt, Wed, 13 Mar 2013 12:02:04 -0400
Well one of my new filters posted and when I went to look at it I see that there is a seam http://filterforge.com/filters/11290.html Can anyone tell me what the most likely culprit is? I've been digging through the components and not sure what is generating that. Thanks.

ThreeDee, Wed, 13 Mar 2013 15:40:40 -0400
Hi Burt,

I found two problems:

1)the smudge component interacting with the profile gradient created the vertical lines at the edges in the corals. It helps to replace the profile gradients with free gradients. (Also, you don't need to smudge the x-displacement gradient, as you're not displacing in that direction.)

2) The fish scaling was also breaking the tiling. They have to be scaled to 1/2 or 1/3 or 1/4 the size (50%, 33,333% or 25%) so that they continue to tile at the edges. The "perspective" fish are even trickier, they have to be scaled 100% at every EDGE so they tile.

I fixed those two issues here but you may want to tweak the look of the schooling fish some more.

TD

PS. Looks like you need to fix the creator name back as well after I edited the file :)

Burt, Wed, 13 Mar 2013 15:58:24 -0400
Wow, thank you so much. That is really above and beyond. I really appreciate it. I'll take a look and do a quick update. :)

CFandM, Wed, 13 Mar 2013 20:14:47 -0400
Looks like you beat me to it ThreeDee......I took a quick look but then had to go do something else before I could post... :)

ThreeDee, Thu, 14 Mar 2013 15:57:51 -0400
Looking at the other presets now there seems to be other complications, as the fish do get cut off. I really just looked at the default preset.

Burt, Thu, 14 Mar 2013 16:02:32 -0400
ah thanks - so did I.I'll have to break out the fish and play with them for a bit.

Improved Smudge 3D Quality

Sphinx. — Wed, 13 Mar 2013 22:31:22 -0400

Sphinx., Sun, 05 Aug 2012 05:02:41 -0400
Here is a simplified and improved way of using the Smudge component for volumetric extrusion. The example rendering is based on the Spherical Previewer.

Instead of extruding the actual bump / surface / ?? material maps, I transform the coordinate space. Since we only extrude in one direction (vertical), we can let the horizontal space remain procedural. This alone is a 50% improvement in rendering quality.

Moreover since we are transforming coordinate space, it makes sense to use the build-in anti-alias and lower the Density parameter a little (somewhere btwn 95-99, depending on rendering context). I.e. we don't risk blurring the sources or flattening (the wrapping effect) the bump map.

Finally since we transform the space, we can reuse the result for as many maps we want. Thus only one Smudge is required.

Feel free to adapt this method in your 2.5D "tracers" :-)

Sphinx., Sun, 05 Aug 2012 05:05:38 -0400
Here is an example rendering. Notice the crisp edges, the lack of extrusion "banding", nice anti-aliasing and non-flattened look :)

SpaceRay, Sun, 05 Aug 2012 05:06:19 -0400
Thanks very much for improving on this topic, I will try it and see how it works.

Nice example you have done, great textures and cool

Indigo Ray, Sun, 05 Aug 2012 08:56:55 -0400
Good idea, and nice result! It's a small step to add other maps. 8)

ThreeDee, Sun, 05 Aug 2012 16:34:04 -0400
Brilliant -- massive improvement in image quality!

I assume this is also faster since the (likely) most calculation-intensive step is only done once.

Sharandra, Wed, 13 Mar 2013 19:43:10 -0400
I just noticed that the sphere doesn´t seem to be really round. Is there a way to fix that?

Burt, Wed, 13 Mar 2013 22:31:22 -0400
Awesome. Have to try this on a couple of filters -thanks.

Fast cast shadows

ThreeDee — Tue, 12 Mar 2013 16:27:37 -0400

ThreeDee, Mon, 23 Jul 2012 10:21:37 -0400
This should prove useful for a number of applications: fast cast shadows.

Here is an example:

ThreeDee, Mon, 23 Jul 2012 10:26:02 -0400
Here's how:

Skybase, Mon, 23 Jul 2012 10:57:06 -0400
I love you.

CFandM, Mon, 23 Jul 2012 11:13:24 -0400
Niiiccceee.. :) :pimp:

ThreeDee, Mon, 23 Jul 2012 12:03:03 -0400
Here's how a surface filter can be enhanced with addition of shadows. Left side plain, right side with shadows:

Sharandra, Mon, 23 Jul 2012 12:22:42 -0400
Wow, that´s totally cool and helpfull! Thx for sharing this :love: :-)

SpaceRay, Mon, 23 Jul 2012 16:16:16 -0400

Quote
Skybase wrote: I love you.

YES, I love you too ThreeDee for making this great and excellent shadow filter and that is very useful and more than you could think as I have changed your filter a little and have discovered that it has a much greater potential and much more useful and amazing than I could have thought looking at the example of the first post.

Quote
Sharandra wrote: Wow, that´s totally cool and helpfull! Thx for sharing this

I agree totally this is totally cool, helpful, awesome and well done and get a excellent effect and I thanks you very much for sharing this

As said I have changed a little your filter adding some more controls that change the usefulness and effects much and now you can do much more things with it IN FACT I HAVE DONE 50 PRESETS to show the new many possible ways this could be used.

See here below the list of the 50 PRESETS

SpaceRay, Mon, 23 Jul 2012 16:22:14 -0400
And here is the modified version of the filter for you to see for yourself those presets with your OWN images, as I have made them with the main image component, so you can load your own image and then try any of the 50 presets on your image to see how it looks like.

SpaceRay, Mon, 23 Jul 2012 16:26:26 -0400
Here is an enlarged version of the 50 Presets I have made to see them better than the above one

SpaceRay, Mon, 23 Jul 2012 17:11:51 -0400
Try and experiment with the settings and modify the presets to get even more possible effects

SpaceRay, Mon, 23 Jul 2012 17:15:25 -0400
:)

Skybase, Mon, 23 Jul 2012 22:49:11 -0400
Should go and try it with other surface filters SpaceRay :p

Morgantao, Tue, 24 Jul 2012 16:18:56 -0400
Looks awesome ThreeDee!
Why don't you add this to the library as a snippet? 8)

SpaceRay, Sat, 28 Jul 2012 23:30:33 -0400

Quote
Skybase wrote: Should go and try it with other surface filters SpaceRay :-p

;) :D Thanks, yes, I already know that the main reason for ThreeDee making this snippet is to help make shadows for surface filters as he has shown in the example well

But after looking at this and looking how it works I have discovered that this could be much more than a simple shadows snippet and that is why I made the above version.

SpaceRay, Sun, 09 Sep 2012 04:31:42 -0400
I really LOVE MUCH this filter in my own and special way as I use it for making tiled images with boxed shadows.

This is really beautiful to make this kind of squared collages

I have used the uploaded above filter with a slighty modification added to make this

What is amazing and great is that ALL this comes from a very simple and plain texture, all the color shades and shadows are added by the filter

SpaceRay, Sun, 09 Sep 2012 04:35:15 -0400
I have remembered about this filter after I have seen an image that was similar to this

SpaceRay, Sun, 09 Sep 2012 04:53:43 -0400
Here is the new updated and better filter version of this awesome Blast Shadows technique made by ThreeDee that I love much

Now you can control the tiles amount and have as few or as many as you want with the Tile Amount

Also have renamed some of the settings and have made a better order of them and cleaner (at least from my point of view)

Also I have added the option to choose the image source, unchecked means that it will take the main image source loaded AND it will also take the main image resolution.

If you activate and check the image source then it will change to the color control image source below that you load AND the main difference is that you will be able to put YOUR OWN FINAL RESOLUTION without having to use the image source resolution, so this way you can make the size much bigger than the image and design the square collage as you want.

Skybase, Sun, 09 Sep 2012 06:23:56 -0400
SpaceRay, one of these days you'll figure to use this on your snow filter. :p

SpaceRay, Sun, 09 Sep 2012 16:23:34 -0400

Quote
Skybase SpaceRay, one of these days you'll figure to use this on your snow filter.

I already know that this filter is mainly for giving shadows to simulate and give the illusion of 3D and make more realistic shadows.

I have ALREADY tried to use this Blast Shadows filter with my own Heavy Snow filter before releasing it to the FF Library to see if this could give better shadows than the Hard Shadows filter or could be used better, BUT as you have said right, I could not figure out how to use it right to give a good effect and I had a bad result, so I kept the Hard Shadows better.

SpaceRay, Mon, 10 Sep 2012 03:18:01 -0400
I want to put some the presets example I have done in my own personal filter to show more what I have done with it and what could be done.

SpaceRay, Mon, 10 Sep 2012 03:20:02 -0400
:) I like much how from a plain wall like the one from the FF window image you can get this tiled holes boxes and the nice shadows effect

SpaceRay, Mon, 10 Sep 2012 03:21:11 -0400
and then modifiying the setting can have much more different effects

Indigo Ray, Tue, 18 Sep 2012 21:26:21 -0400
I like this, ThreeDee. 8)

Indigo Ray, Tue, 18 Sep 2012 22:07:27 -0400
You can make hard shadows, too.

SpaceRay, Thu, 07 Mar 2013 05:43:12 -0500
Have made another different test with this great shadows filter, mixing two different images with stripes and applying this fast cast shadows filter and got this nice result

For this test you ONLY have to replace the two tiles components of the original filter above and put the source you want from any group of components

Burt, Thu, 07 Mar 2013 08:41:32 -0500
Wow that it awesome. Have to try it right now.

SpaceRay, Tue, 12 Mar 2013 16:27:37 -0400
Glad you like it and may be useful for you, Burt and if you want you could share any example that you could make

This filters can have some good uses

The DOs and DON'Ts of Filter Construction

Crapadilla — Tue, 12 Mar 2013 15:00:50 -0400

Crapadilla, Fri, 14 Dec 2007 10:13:44 -0500
I just set up a new wiki article (and provided the first entries). Comments and contributions are welcome! :D

The DOs and DON'Ts of Filter Construction

Vladimir Golovin, Fri, 14 Dec 2007 11:55:07 -0500
+1024 on the Perlin-into-Blur thing.

My only suggested improvement would be to type it in HUGE BLINKING RED LETTERS. Also, some irritating sound would be nice, something like the NAVY alarm :D

Sjeiti, Fri, 14 Dec 2007 12:47:59 -0500
Let's hope HTML 5 will have <blink /> again :-p

StevieJ, Fri, 14 Dec 2007 12:59:30 -0500
Dilla, that is excellent!!! Nice work!!! :)

Conniekat8, Fri, 14 Dec 2007 13:05:45 -0500
oooh, very handy information, especially for us noobs!
THANK YOU!

Crapadilla, Fri, 14 Dec 2007 13:14:37 -0500

Quote
Vladimir Golovin wrote: My only suggested improvement would be to type it in HUGE BLINKING RED LETTERS.

If anyone could assist in wiki markup editing to this end, I'd be grateful. ;)

Crapadilla, Fri, 14 Dec 2007 13:16:38 -0500

Quote
StevieJ wrote: Dilla, that is excellent!!! Nice work!!!

Quote
Conniekat8 wrote: oooh, very handy information, especially for us noobs! THANK YOU!

I'm glad you appreciate it, and I hope to add more items to the article as they cross my mind! ;)

As I stated above, feel free to contribute, all ye who feel like contributing!

StevieJ, Fri, 14 Dec 2007 13:40:04 -0500

Quote
Crapadilla wrote: feel free to contribute, all ye who feel like contributing!

Are you sure you want me involved in this??? You might have to do alot of editting..... :| ;) :D LOL....

I'm holding out for my first Editor's Pick before I go to the next level of contributing..... :devil: ......J/K ;) :D LOL....

Crapadilla, Fri, 14 Dec 2007 13:43:04 -0500

Quote
StevieJ wrote: Are you sure you want me involved in this??? You might have to do alot of editting....

If you have a meaningful item to contribute, then yes - by all means - please go ahead! :)

Sphinx., Fri, 14 Dec 2007 14:05:38 -0500
looks good - the channel trick is something I often use, helps alot not only to speed up the rendering, but also to keep the memory usage down (by limiting the number of components). For example if you are using bitmap based components to process grayscale sources, then you can process 3-4 sources seperately with one component (high pass ignores/bypasses alpha, so its only 3, RGB).

One thing.. in the stuff about Noise and Blur, it does look like a general statement - "never fed your noises to a blur" alike - but as I recently showed, it can speed up the rendering dramatically - specially when worleys are used on a height tree or get distributed out to many post processing branches that breaks with the single sample cache system (like offsets etc).

Ofcourse its a special case where the actual purpose not is smoothening, but rather buffering. But if the blur is there in first place to serve as buffer, and you later on figures that it actually looks good with a higher radius, well then you have a blur avail. at no expense since it already sits there for another purpose.. I wouldn't be so angry at those blurs :D (Just so you know.. I'm not opposing the motives here.. I fully agree on the overall point in it .. just adding in a little perspective here)

StevieJ, Fri, 14 Dec 2007 14:07:41 -0500
Well, I wouldn't on this subject matter because I'm just a novice compared to your knowledge of it.....but I will help out in other areas.....if I feel that I almost know what I'm talking about..... ;) :D

Right now, I'm finishing up a series of B&W photo filters with different types of soft lens effects......and then, ***I can't believe that I'm going to say this :cry: *** I'm actually going to start submitting some texture/theme filters (....and yes, simple results with no lighting :D ).....in a desperate effort to get that elusive first Editor's Pick..... ;) :D LOL.....

It's extremely nice of you to do this work for FF.....in addition to all the other things you have done here.....which I've certainly benefitted from......so thanks!!! :) :) :)

Mousewrites, Fri, 14 Dec 2007 14:10:56 -0500
Ooooo. Thank you muchly. This will be lots of help.

Mousewrites, Fri, 14 Dec 2007 14:15:29 -0500
::Reads the wiki::

::Reads it again::

::Feels like an idiot about 5 times::

Crapadilla, Fri, 14 Dec 2007 14:33:33 -0500

Quote
Sphinx. wrote: One thing.. in the stuff about Noise and Blur, it does look like a general statement - "never fed your noises to a blur" alike - but as I recently showed, it can speed up the rendering dramatically - specially when worleys are used on a height tree or get distributed out to many post processing branches that breaks with the single sample cache system (like offsets etc).

Sphinx,

would you be willing to supply a short paragraph about this 'caching' behaviour to complement the perspective on blur components that I gave?

Quote
Sphinx. wrote: I wouldn't be so angry at those blurs Big grin (Just so you know.. I'm not opposing the motives here.. I fully agree on the overall point in it .. just adding in a little perspective here)

Actually I like blurs very much, and I use them often.

But! With blur components (and the other the bitmap-based ones) you have to know exactly WHERE and WHEN to use them, or they do more harm than good. Consequently, it is also very beneficial to know when and how you can do WITHOUT them. ;)

Sjeiti, Fri, 14 Dec 2007 14:57:20 -0500
I was thinking of contributing something about the height input on surface result but I'm not sure if most people will agree with me on this:

When I assemble a gray scale image for height I make sure the darkest pixels approach black and the lightest pixels approach white, and afterward correct the height using surface height from the lighting tab.
Now of course you don't have to do this, you can make a height map that is almost even gray and still have a great looking output. But I can't help feeling these gray height maps are wasting information. When you have a 0-255 range why not use it? It is more accurate, especially when you intend to use the bump map output for 3D.

Kraellin, Fri, 14 Dec 2007 15:11:39 -0500
one might also mention that blurs can be used with perlins in certain circumstances, like feeding blurred results INTO perlin noise or backgrounds, AND feeding blurred results into the perlin controls themselves. when you take into account how, in your noise lab, you utilize a number of the perlin control slots with other things, those controls often cant blur as you state in the do's and donts. so, a separate blur is sometimes necessary there, also... i think.

sorry, everyone pokes their two cents in AFTER someone makes some excellent points and tutorial helps. and mention shld be made of the ORIGINAL work you've done here, dilla. excellent stuff! so, forgive my mods/two cents, please.

Crapadilla, Fri, 14 Dec 2007 15:26:26 -0500

Quote
Sjeiti wrote: But I can't help feeling these gray height maps are wasting information. When you have a 0-255 range why not use it? It is more accurate, especially when you intend to use the bump map output for 3D.

Yes, exactly! I perfectly agree. This would be a good one to add.

Sphinx., Fri, 14 Dec 2007 16:03:25 -0500

Quote
Crapadilla wrote: would you be willing to supply a short paragraph about this 'caching' behaviour to complement the perspective on blur components that I gave?

Well, I would like to hear an "official" statement from Vlad about the buffering method, because it essentially opposes the procedural sampling concept (in that it forces a rasterization of the blur input, and thereby prevents further real subsampling) - I can understand why that might not be considered good practice, and if that is the conception, it should not be promoted. Then again it could just be a "novel" unrelated method for which all the blur/bitmap-based component lynching seen here and there don't apply :D

However I think the performance gain shown in the snippet argues for it self, and what I really should do is to add a section to the optimization section of the wiki, which we then could add a small note about in the Noise and Blur text (e.g. "This ofcourse don't apply for cases where blurs are used for buffering mechanisms" or something). By the way.. today I played with the method for buffering large curve "systems".. and it works very well there too, and the side effects are perhaps even less visible here (some of the curve components are quite heavy performancewise, specially when controlled by map inputs)

Quote
Sjeiti wrote: But I can't help feeling these gray height maps are wasting information. When you have a 0-255 range why not use it?

I'm not sure I agree here. FF does range checking and clamping just for about every component output (e.g. if output > 1 then output = 1), so there is a good reason to stay in "safe" mid area ranges to prevent things from getting overblown (resulting in flattened heightmap areas). Since the internal precision level is way higher than 8 bits per channel, it is not a problem at this point. The problem may arise when the user choose to export to an 8 bit per channel format, and that is, I think, the area where the guiding effort should be put. If users want to ensure full exploitation of a given target bit range, they should render the maps to a HDR format, and do the appropiate adjustments afterwards in relation to the target bit depth. It would also make a good feature if FF had an option to maximize heightmaps for a given bitdepth (most relevant 8 bit).

Sjeiti, Fri, 14 Dec 2007 16:27:16 -0500
Well I always use the 'threshold trick' to prevent clamping. I'm not saying high and low should be within a three pixel range of white and black. I'm just saying that almost evenly gray height maps is a waste of ones and zeros. I'll just add the height map point to the page, refer to the danger of clamping, explain the 'threshold trick', and also add that you don't have to do this to create a good filter.

As for HDR, I know nothing about that, so feel free to add that.

(OT: hey wow I just learned Phong is actually a person, I always thought it had something to do with plastic)

Crapadilla, Fri, 14 Dec 2007 16:59:04 -0500
Well, personally, I prefer my height maps sprawling fully between the 0..1 range; call it an old 8-bit habit, if you will, but I think it feels somewhat clean and tidy this way. It is true though that - with the advent of HDR formats and floating point precision - this becomes somewhat less of an issue.

Sphinx., Sat, 15 Dec 2007 03:18:21 -0500
I just took a closer look on how the bump map and normal are rendered - I actually thought that the Surface Height slider also affected the rendered bump map (that flattening could occur even if the height input is valid with high Surface Height settings). It is not the case, only the normal map is affected by that slider.
With that knowledge I agree.. the heightmap input should be as close to full range as possible - how does that "threshold trick" work? Is it simply a fixed "safe" distance from the range limits, black and white?

I really think that the histostretch ought to be handled by FF and not our guestimate of a "universal" stretch (which can only be safe if the heightmap contrast is not controlled by any user controls directly or indirectly). When FF is set to render a bump map, it should do this to an intermediate hdr map, search for min/max values and adjust the levels accordingly to those when saving the bumpmap to 8 bit formats.
Perhaps a good feature request?

Quote
The renderer has an optimization called "sample cache" that can save a significant amount of rendering time in situations when a single component has multiple outbound connections. Note that this speedup will not occur when using duplicates instead of connections.

I wonder how this works when some of the outbound connections end up in offsets, refractions or noise distortions at some point, and others don't. Consider the following case: we have a perlin noise with two outbound connections one that goes directly to the reflectivity input and one that goes to an offset which then goes to the surface input:

[perlin]----------------->[surface result (reflect.)]
. . . . . |
. . . . . |---->[offset]--->[surface result (surface)]

the offset alters the coordinate passed down to the perlin when rendered (i.e. it will fetch samples from an offset location).
Will this "thrash" the sample cache, because the last fetched sample constantly shifts because of the coordinate changing?

Sjeiti, Sat, 15 Dec 2007 04:20:41 -0500
The threshold trick...
In the image I use levels to alter the stones... I've attached two threshold, one 100, one 0, black and white. When you adjust the the levels both thresholds should remain perfectly black and white. So in this image I've set the white point too far.

Sphinx., Sat, 15 Dec 2007 04:41:54 -0500
ahh. I see.. its a guide for adjusting to maximum contrast. Nice!

I thought you had some sort of automatic solution (which is not possible without knowing the global min and max levels, thats why FF should handle it in a post render step).

Sjeiti, Sat, 15 Dec 2007 05:03:51 -0500
Damn Wiki... how do you apply breaks? <br/>

Sjeiti, Sat, 15 Dec 2007 05:22:29 -0500
Hey wow... I love that 'Channels trick'.

Crapadilla, Sat, 15 Dec 2007 05:56:20 -0500

Quote
Sphinx. wrote: I wonder how this works when some of the outbound connections end up in offsets, refractions or noise distortions at some point, and others don't.

As I understand it, the sample cache for a component with multiple outbound connections is 'located' at its output, and each component with multiple outputs has its very own sample cache. The component gets rendered once and the result is then cached for other components to quickly 'look up' as needed.

Sphinx., Sat, 15 Dec 2007 06:26:48 -0500

Quote
Crapadilla wrote: As I understand it, the sample cache for a component with multiple outbound connections is 'located' at its output, and each component with multiple outputs has its very own sample cache. The component gets rendered once and the result is then cached for other components to quickly 'look up' as needed.

Yeah, but doesn't that imply that the cache is invalid (and thus a new rendering is needed) if we have different offsets for the outbound connections?

Lets consider a simple hypothetic rendering case for that simple setup I mentioned (changed a bit here):

Quote
[perlin]----------------->[surface result (surface)] . . . . . \| . . . . . \|---->[offset]--->[surface result (reflect.)]

We are rendering a 2x2 pixel texture, and the result component starts rendering at position (0,0), meaning that it will request a sample from this position:

1. result[surface] passes (0,0) to perlin, which returns this sample and caches it for future (0,0) requests

2. result[reflect.] passes (0,0) to offset, which adds its offset, lets say (+1,+1), to the coordinate it recieved, resulting in (1, 1)

2.1 offset passes (1, 1) to the perlin to get a sample from this new offset coordinate

2.2 perlin checks if the request is in its sample cache (if current coordiate = previous coordinate then return cached sample), but since last coordinate was (0, 0) and new coordinate (due to the offset) is (1, 1), the cache is invalid and a new sample rendering is needed

See what I mean? If this is how the sample cache stuff works, hooking up many branches to the same leaf component will only be faster when the coordinates remain unaltered, so in there is a reason not to do so in many situations.

The question is which components changes the coordinates? I think that is Offset, Noise Distortion, Refraction, Components in Jumble Fill Mode.. is there others?

Crapadilla, Sat, 15 Dec 2007 06:34:28 -0500

Quote
Sphinx. wrote: Yeah, but doesn't that imply that the cache is invalid (and thus a new rendering is needed) if we have different offsets for the outbound connections?

I don't think so. If you connect a 'sample cached' perlin noise into three different offsets, the same sample cache is read by all three offsets, and SUBSEQUENTLY each offset calculates its own coordinate transformations using the cache as its source.

Sphinx., Sat, 15 Dec 2007 06:39:17 -0500

Quote
Crapadilla wrote: I don't think so. If you connect a 'sample cached' perlin noise into three different offsets, the same sample cache is read by all three offsets, and SUBSEQUENTLY each offset calculates its own coordinate transformations using the cache as its source.

Hmm, but that implies that there is more than one cached sample - thats not how I understood the concept - I'll do some testing an reading ;)

Crapadilla, Sat, 15 Dec 2007 06:44:37 -0500

Quote
Sphinx. wrote: Hmm, but that implies that there is more than one cached sample

Yes, the way I understand it, each component that has multiple outbound connections maintains its own sample cache.

Continuing the example of the perlin and three offsets in my post above, if you were to connect one of those offsets to three other components in the filter tree, this offset component would in turn be enabled for sample caching and build its very own cache. I presume that any given component's sample cache remains valid until the components' source input or its parameter values change.

This means that depending on where the sample cached component is located in the filter tree, tweaking filter controls may force a whole cascade of caches to be rebuilt.

Sphinx., Sat, 15 Dec 2007 07:06:18 -0500

Quote
Vladimir Golovin wrote: The sample cache is located at the output, not at the inputs of the component. Every component stores the coordinates and the value of the last calculated sample, and if the next sample has the same coordinates the component returns the cached value instead of evaluating the sample.

(source)

I read this is *one* single sample is cached in the output.

Quote
Crapadilla wrote: Yes, the way I understand it, each component that has multiple outbound connections maintains its own sample cache.

I don't think the fact that there are multiple outbound connections over just one affects wether sample caching is active. Seems like its a generic mechanism: last coordinate and sample is cached at the output, and if next sample request has the same coordinate, it will return the cached sample, and implicitly if the coordinate doesn't match, a new sample will be calculated and saved as the new cached sample..

Crapadilla, Sat, 15 Dec 2007 07:12:48 -0500

Quote
Sphinx. wrote: Seems like its a generic mechanism:

That is a question for Vlad or a programmer to answer. Sadly, there is not much info on FF's sample architecture available, so I'm just deducing how all this might work from the help file, Vlad's postings, etc. ... ;)

Sphinx., Sat, 15 Dec 2007 07:36:34 -0500

Quote
Crapadilla wrote: That is a question for Vlad or a programmer to answer. Sadly, there is not much info on FF's sample architecture available, so I'm just deducing how all this might work from the help file, Vlad's postings, etc. ...

LOL, yeah, same here :D

Ok check out this test filter.. its not a very large filter graph, but it still shows a significant difference. The filter has one option "Duplicated Perlins", which if checked uses two seperate perlin components on each "coordinate branch" - when the checkbox is changed I made it force a variation and color shift too so its easier to see the difference.. on my single core nutcracker, the duplicated chain is faster.. how does it perform on your CPU octopus? :D

(edit - updated file, this one shows more significant difference)

Crapadilla, Sat, 15 Dec 2007 08:16:44 -0500
I've added some general guidelines for contributors...

*** nudge nudge Sjeiti nudge *** ;) :D

And it seems the article is growing! :)

Crapadilla, Sat, 15 Dec 2007 08:34:24 -0500

Quote
Sphinx. wrote: how does it perform on your CPU octopus?

I'm currently not in the studio, so I only have an old athlon 1,0 GHz snail at my disposal. No performance testing until monday. ;)

Sjeiti, Sat, 15 Dec 2007 08:38:23 -0500
Ok... consider my contribution guided

Maybe a good idea to put these general 'guidelines for contributors' on the main page. Add to it a link to the Wiki help.
A lot of the people on the fora might feel a bit reluctant to contribute to the Wiki (the FF Wiki was my first too) so some small pointers on how a Wiki works might convince them it's not all that hard.

Sphinx., Sat, 15 Dec 2007 08:42:23 -0500

Quote
Crapadilla wrote: I'm currently not in the studio, so I only have an old athlon 1,0 GHz snail at my disposal. No performance testing until monday.

Ok.. You should check it out anyways, as it seems to verify that we're dealing with single sample cache - the example shows that using two seperate perlins with rather expensive settings is faster than using just one perlin with those settings!

It will be interesting to see if this holds true for multicore systems too - anyone out there with a dual/quad core is more than welcome to test if which of the modes render faster, "red" or "green" (duplicated perlins).

Crapadilla, Sat, 15 Dec 2007 08:49:49 -0500

Quote
Sjeiti wrote: Ok... consider my contribution guided

Great! And how about adding an image? ;) :D

Quote
Sjeiti wrote: Maybe a good idea to put these general 'guidelines for contributors' on the main page. Add to it a link to the Wiki help.

Well, people have different opinions on how to design and structure things, so currently I feel it is sufficient if the originator of an article provides some guidelines.

Sjeiti, Sat, 15 Dec 2007 16:30:16 -0500
mumble mumble adding image.. the things I do for humanity

One thing I never quite understood about FF's Perlin noise:
The Perlin noise I know has octaves and persistence. Octaves being the number of calculated noise maps, each of which has a doubled frequency (1,2,4,8...). And persistence (or falloff) is the amount in which these maps are blended with each other (persistence of .5 means 1*map1+.5*map2+.25*map3+.125*map4...).
If FF's roughness and detail relates to octaves and persistence, then how can detail increase render time?
They must have found some other exciting way to render Perlin noise, otherwise I really don't get it.

Sphinx., Sat, 15 Dec 2007 18:31:13 -0500
as I see it roughness defines how the different octaves are mixed, and details control how many octaves are in play, but moreover it controls the contribution of the last added octave (given some sort of fractional range). What I'd like to know is how many octaves are in play when Details is = 100...

Crapadilla, Sat, 15 Dec 2007 18:42:25 -0500

Quote
Sphinx. wrote: What I'd like to know is how many octaves are in play when Details is = 100...

It's eleven according to the help file... ;)

Crapadilla, Sat, 15 Dec 2007 20:25:30 -0500
The article has been updated significantly...

Time to get some close-eye! :)

Sjeiti, Sun, 16 Dec 2007 03:59:05 -0500
Ah 'help file' never thought of that.

So roughness is persistence and details is octaves. Oh wow, they even explain why roughness has can influence speed (which normally wouldn't happen).
And how in details floating points are used while octaves are integers.

Good help file.

Maybe a good thing to refer to it from the wiki more often. For instance: each time the Perlin noise component is mentioned link to it.

Crapadilla, Sun, 16 Dec 2007 05:28:41 -0500

Quote
Sjeiti wrote: Maybe a good thing to refer to it from the wiki more often. For instance: each time the Perlin noise component is mentioned link to it.

Yes, good idea!

Sphinx., Sun, 16 Dec 2007 10:28:26 -0500
Added some info on how to get Sharpen and High Passes from a blur (and more..).. its my first real wiki edit, is it allright?

One Blur does it all

Crapadilla, Sun, 16 Dec 2007 10:41:59 -0500
Very nice contribution, Sphinx. Thank you! :beer:

One never stops learning! :)

Crapadilla, Sun, 16 Dec 2007 10:44:21 -0500
I'm wondering whether we should start setting up broader categories for the entries. We've already got several entries on 'Blur' and 'Noise' for example...

Sphinx., Sun, 16 Dec 2007 10:54:58 -0500
Yeah, that might a good idea.. I have one more entry on blur, high passes and sharpens brewing..

Btw. how about merging in the "Optimizing Filters for Speed" section?

Crapadilla, Sun, 16 Dec 2007 11:00:27 -0500

Quote
Sphinx. wrote: Yeah, that might a good idea.. I have one more entry on blur, high passes and sharpens brewing..

Alright, I'll reorganize the entries once that second entry of yours has landed. ;)

Quote
Crapadilla wrote: Btw. how about merging in the "Optimizing Filters for Speed" section?

Yup. That might be a good idea, since its also about efficiency!

Sphinx., Sun, 16 Dec 2007 11:17:45 -0500

Quote
Crapadilla wrote: Alright, I'll reorganize the entries once that second entry of yours has landed.

Having second thoughts on that entry - its about using offsets to construct a low radius blur as an alternative to using the blur/highpass/sharpen component for low radius filtering (as shown, once we have blur, we also have High Pass and Sharpen).. but maybe I should just check that in as a snippet instead.. I think that is better.. so just go ahead with the reorganization :)

Building 'tall' objects out of polygons and clipping

Burt — Fri, 08 Mar 2013 18:23:15 -0500

Burt, Fri, 08 Mar 2013 14:12:06 -0500
I attached a piece of one of my filters that generates a feather duster worm. The 'head' part comes fr om a shape that is used to generate plants and other things. My problem is that when I feed the resultant worm into the Bomber it is tiny. I can do another scale but it is still too small unless I expand it to the point where the tube is partially clipped. May be a dumb noobish question but how do I increase the height without clipping? Do I have to cut it in half and blit both halves somehow or does one of the components allow stuff 'off screen' (the area represented by the Component window) to remain unclipped :?:

CFandM, Fri, 08 Mar 2013 15:58:50 -0500
Turn down the repeat to 1 or 2 on the bomber and use density instead.... ;) :)
That will give you more worms but won't reduce the size...

Burt, Fri, 08 Mar 2013 16:10:23 -0500
Thank you. :) The only issue then is that it is sized normally but all the other objects are now gigantic so I would have to use Scale to reduce all of them. :)

CFandM, Fri, 08 Mar 2013 17:48:48 -0500
Yep, sometimes you just gotta add a little bit of leverage to the bomber to make everything work and look good with it... ;) :)

Burt, Fri, 08 Mar 2013 18:23:15 -0500
It's ok. Maybe it would be nice in future version if each channel in Bomber allowed you to set a scale.

shapes to grid?

Burt — Wed, 06 Mar 2013 23:50:11 -0500

Burt, Tue, 05 Mar 2013 13:32:39 -0500
I was wondering if there was some 'secret' way to use something like the checker or tile component and load each grid square with a shape. I had been using multiple offsets to line up a few items and then feed the result to the Bomber - but it requires a lot of components, plus prodding and adjusting to get it right.

SpaceRay, Wed, 06 Mar 2013 03:10:50 -0500

Quote
Burt some 'secret' way to use something like the checker or tile component and load each grid square with a shape

:?: Sorry, that I do not understand well, what do you mean with "shapes" loaded into a grid square? do you mean to have FF shapes like the ones shown by CFandM in his awesome shapes tutorials and to have these shapes inside the little tiles or grid square of the checker?

Would be good and better if you could upload a example or screenshot of what you want to have or show what you have done until now

Burt, Wed, 06 Mar 2013 11:21:32 -0500
I am creating some coral polyps using the polygon. Rather than adding an additional Bomber with multiple polyps I was thinking of loading multiple into one component and using it in one remaining slot of my existing Bombers (already using 2 Bombers). I did this before using multiple offset like so

so I can load say 4, 6 or even 8 shapes on one component like MultiBlend and then 'blit' that from one channel of the Bomber and have very good coverage.

What I wanted to see was if the Tile or Checker component could load images like below so I wouldn't have to use a whole bunch of Offsets to build this. Both offer grid-like output but it doesn't seem to be possible to put one shape per square -unless I missed something (which happens often) so that is what I am asking.

Burt, Wed, 06 Mar 2013 11:25:00 -0500
I know that I can use the Free Polygon to combine Offset and Shape into one and then chain them but I was wondering if what I asked is possible. There could be many situations where this grid would be useful.

[Note: ignore the 'jiggle' and other things in the component image above. This is fr om another post wh ere I suggested a grid component.]

Skybase, Wed, 06 Mar 2013 11:53:57 -0500
... well I thought I give this a shot but it turns out the way I did it didn't suit your needs. Either way here's a method combing elements on a grid without a bomber.

Burt, Wed, 06 Mar 2013 12:16:04 -0500
That looks very interesting and will try that as soon as I get back. I'm not sure if it requires more components than the offset method or if it uses less or more system resources but it is awfully cool and gives me ideas for other uses - so thanks!

With FF4.0 I guess we can roll something like this into a custom component so will be useful.

Burt, Wed, 06 Mar 2013 17:53:05 -0500
Check out the grid based on Skybase's wiring above with a little perspective and my fish.

SpaceRay, Wed, 06 Mar 2013 23:26:53 -0500
Nice Lionfish with those tiny fish :)

If what you want is to havve that kind of shapes in a checkerboard like chess game, I perhaps can recommend taking a look to this other filters that may be useful somehow

game board by voldemort

Checkerboard v3 by PixelDust

Burt, Wed, 06 Mar 2013 23:50:11 -0500
Hey Spaceray thanks for that link will check it out. :)

Some way to hide filters settings panel from excluding components?

SpaceRay — Wed, 06 Mar 2013 05:31:05 -0500

SpaceRay, Mon, 04 Mar 2013 21:43:25 -0500
I know that this maybe it is not possible, as I remember I have seen something somewhere else, but do not know where, but any way I ask it here for FF 3.0.

I want to mix three snippets that are excluding themselves, I mean that when using the switch ONLY one of the snippets would be used and not any part of the other two, BUT the settings of the three snippets appear all together in the same settings panel even when you do not need them all at once.

And there is no way to organize settings by groups or parts as all the settings are always mixed together in just one only list that you can of course organize and order as you want, but it can be a mess of settings.

And it can happen that you may modify settings that are not working because they are from the part of the snippets that is not used in that moment.

Does the Groups in FF 4.0 are able to be activated or desactivated in some way for this to happen?

Thanks

here below I show a possible example to help explain it more

SpaceRay, Tue, 05 Mar 2013 00:08:06 -0500
So the above example shows the setting that would be seen in the settings panel if you choose the blue part in the switch or checkbox provided in the filter

And here below are the five settings that would appear and be shown only when you choose the red part components in the switch or checkbox and the other 3 settings would be hidden.

CFandM, Tue, 05 Mar 2013 09:12:46 -0500
Well it looks like you will have to delete those controls and then adjust the settings on the component accordingly... ;) :)

SpaceRay, Wed, 06 Mar 2013 04:15:45 -0500

Quote
CFandM Well it looks like you will have to delete those controls and then adjust the settings on the component accordingly...

I think that from your answer, I have not explained it well (as usual from me)

I do not want to delete any controls as they must be there to configure the options, BUT the problem is that are being shown ALL in the same settings panel

As I think I have explained badly and shown a complex example above, I have made a new one more simple

SpaceRay, Wed, 06 Mar 2013 04:26:37 -0500
From the above example I want that with this filter that have two very different effects and that when going to the settings panel and when you have:

* the switch unchecked it would ONLY show in the settings panel the setttings applied to the RED group of components AND NOT SHOW the settings of the BLUE group

* and when you check on the switch it would ONLY show in the settings panel the settings from the BLUE group and do not show the ones from the RED group.

There is no point of having Mortar and Mortar width settings if you have choosen to use the first effect, as it will do nothing on the first effect and will not work at all, and would be confusing for customizing and knowing what can be used and will work.

And also the opposite, and if you only want the second effect, the others setting from the first group would not work or show any change.

FF 4.0 GROUPS SWITCH ON/OFF FOR SETTINGS

I think that this thread should have been much better in the "Features Wishlist" instead of here

I think that one good way and easy to make this would be to use the FF 4.0 Groups and that all the included settings from that groups could be switched ON/OFF if this group is selected and used by the filter, and become desactivated if it is not used. This could be with a simple switch component.

SEPARATING SETTINGS IN THE SETTINGS PANEL NOW IN FF 3.0

As far as I know the only way to separate and organize settings by groups inside the setting panel without hiding them is to use dummy controls that allows you to have some text on it to make a separation

SpaceRay, Wed, 06 Mar 2013 04:47:12 -0500
In the above example I have used ---------- to separate the settings from one group to another but it can be used also text to explain it also as shown here

lipebianc, Wed, 06 Mar 2013 04:51:04 -0500
SpaceRay, I understood perfecly what you wanted on the 1st post of this thread! ;)

Quote
SpaceRay wrote: I think that this thread should have been much better in the "Features Wishlist" instead of here

I was going to write exactly that!
*IF* this feature were implemented wisely, would be a huge game changer on organizing filter's GUI to the users for sure!
More options to customize filter's GUI is an old feature request, as I found some time ago...

SpaceRay, Wed, 06 Mar 2013 05:31:05 -0500

Quote
lipebianc SpaceRay, I understood perfecly what you wanted on the 1st post of this thread!

Good, that you have understood it

Quote
IF this feature were implemented wisely, would be a huge game changer on organizing filter's GUI to the users for sure!

Yes I agree with you that it would be a huge game changer, and this feature would open a NEW wider door to new kind of ilter that could have multiple different options and ways that are not easy to do now as it would be much confusing to put the controls ALL together inside the settings panel.

Quote

lipebianc

Quote
SpaceRay I think that this thread should have been much better in the "Features Wishlist" instead of here

I was going to write exactly that!

Yes, I thought it after thinking that there is no possible way to do this in FF 3.0 and so to solve this and explain it in a better way considering it from a better way I have made this other thread

Switch on/off or activate/desactivate settings inside Groups in FF 4.0

As I have suggested above adn in this new thread the possible solution would be to use the new FF 4.0 groups

The Gradient Thread

Burt — Mon, 04 Mar 2013 11:48:03 -0500

Burt, Mon, 04 Mar 2013 11:48:03 -0500
I used a gradient from http://filterforge.com/filters/8788.html Sjeiti "Scallops" filter as the basis of my own simpler shell. I don't really understand all the possibilities of gradients yet but am fascinated by the uses I have seen so far.

I looked through the forums and there are some good examples scattered about but several link to dead sites and the Wiki has mostly 'under construction' for the various gradients - so please take a moment to share your gradients here.

Just to illustrate using the gradient borrowed from Sjeiti's filter I created a simple shell, sea urchin-like critter, a plant and a 'polyp' in a hole.

Noise Lab

Crapadilla — Sun, 03 Mar 2013 18:31:44 -0500

Crapadilla, Sat, 01 Sep 2007 18:14:11 -0400
On this little excursion to my secret hideout, I'll be presenting - in brief - some of the results of my ongoing experiments with noises. This thread will be updated every now and then as new intersting results appear.

Hopefully, the following images will be proof enough that there is more to noises in FF than may be apparent at first glance. They key to simulating the complexity of real-world surfaces lies in combining the building blocks that FF provides into more complex noise-generators, yet - sadly - few filters on the library dare tread into this territory. This needs to change... ;)

Alright then, let's get this rolling. Here are few key 'ingredients' to my noise experiments that any noise chef worth his perlin should ponder:

- Map Inputs are key to complex noises!
- Map the 'Roughness' parameter with another noise to locally introduce surface chaos.
- Handle Roughness with care!
- Don't be afraid to stretch and rotate your noises. Many surfaces do have 'direction' to their chaos.
- Use noise at different scales to layer in different levels of chaos.
- Use Noise Distortion on your noises to add in chaos. If you're especially daring, use a Worley Distortion even!
- Map your Noise Distortion's parameters with other noises.
- Blend your noises. Yes, it's as easy as that!
- Experiment with blend modes and see what a difference it makes!
- Use profile curves! And did I mention: Use profile curves?!
- Map curve components' parameters with another noise (which could in turn be mapped by a profile curve, which... you get the idea)!
- Make your own profile curves!
- Feed a noise and a Noise Gradient into an Elevation Gradient.
- Offset your noises. Use other noises to map the Offset's parameters!
- Got two complex noise generators? Put them together and see what transpires!
- Experiment! Many interesting complex noises are happy accidents!
- Save interesting results (even if intermediary) in their own filter file. You never know when you might need them.

That should do for now. Besides, I've used up all my exclamation marks for the day. Let's see some images... :D

Crapadilla, Sat, 01 Sep 2007 18:14:45 -0400
Encrusted Debris

Crapadilla, Sat, 01 Sep 2007 18:15:10 -0400
Marbled Veins

Crapadilla, Sat, 01 Sep 2007 18:15:38 -0400
Tar Pit

Crapadilla, Sat, 01 Sep 2007 18:16:04 -0400
Weathered Wood

Crapadilla, Sat, 01 Sep 2007 18:16:34 -0400
Pumice

Crapadilla, Sat, 01 Sep 2007 18:17:04 -0400
Limestone Cave

Crapadilla, Sat, 01 Sep 2007 18:17:54 -0400
Mossy Growth

Crapadilla, Sat, 01 Sep 2007 18:18:41 -0400
Crack Melt

Crapadilla, Sat, 01 Sep 2007 18:19:25 -0400
Rough Plaster

Crapadilla, Sat, 01 Sep 2007 18:21:28 -0400
Variant A. I just exchanging a perlin with a worley noise here.

Crapadilla, Sat, 01 Sep 2007 18:22:23 -0400
Another Variant. Just changed a blend mode... ;)

Crapadilla, Sat, 01 Sep 2007 18:23:45 -0400
Just using a different blend mode again. Notice the drastic change in the look of the noise.

Crapadilla, Sat, 01 Sep 2007 18:24:49 -0400
More coming tomorrow... :D

Kraellin, Sat, 01 Sep 2007 20:36:40 -0400
what do you mean by 'map inputs'?

capnsparrow, Sat, 01 Sep 2007 20:51:26 -0400
I think he is referring to any of a components inputs which can be mapped with an image or curve based component, as opposed to a numerical value.

For example: You can plug a 'perlin' noise into the 'roughness' input on a 'cells' noise, and it will create a roughness pattern in the cells, based on the perlin map.

:)

jffe, Sat, 01 Sep 2007 22:01:46 -0400
I am not super familiar with the others, and lacking the coloration they are tough to comment on, but that *pumice* example looks great. :D

jffe

onnetz, Sat, 01 Sep 2007 23:22:39 -0400
anyone hungry?
perlin noise, cel noise and a wave curve.
The cel noise isn't really necessary for this but it does add a little to it.

onnetz, Sat, 01 Sep 2007 23:24:46 -0400
one thing I've noticed and was said in another thread is that most noises with a curve as a profile really slow things down. with perlin somewhat being the exception.

this one could be flames? its perlin with a wave curve as the profile.

onnetz, Sat, 01 Sep 2007 23:25:35 -0400
steps are really cool too.. this was a cel noise and a step curve as the profile.

onnetz, Sat, 01 Sep 2007 23:28:06 -0400
same thing just a different scale.. I could see this easily being water..

garbanzo, Sun, 02 Sep 2007 02:21:45 -0400

thanks for the wonderful tips crapadilla, and for your input too, onnetz. good stuff, i'm going to bookmark this page right now and get experimenting this week!

Kraellin, Sun, 02 Sep 2007 02:29:28 -0400

Quote
I think he is referring to any of a components inputs which can be mapped with an image or curve based component, as opposed to a numerical value.

hmmm, well, wouldnt just anything be a map input then? i mean, aside from controls, assemble and extract and those max and min components, just everything else would have map inputs, yes, no? basically, wouldnt it just be the green nodes then?

onnetz, Sun, 02 Sep 2007 03:19:57 -0400
One thing to note about mapping say one noise to anothers background input, you lose the individual control of parameters like contrast. Its now controlled by the parent in the chain.

Crapadilla, Sun, 02 Sep 2007 03:50:01 -0400

Quote
jffe wrote: lacking the coloration they are tough to comment on

The examples are all surface-filters with just a height map plugged in and diffuse set to white. I'd like to keep it that way to focus on surface (i.e. height) detail in this excursion.

Quote
onnetz wrote: perlin noise, cel noise and a wave curve.

Another tip:
- High 'Contrast' values lead to 'clipping' (i.e. flat plateaus or valleys on the heightmap), so it is best to keep this parameter as low as possible.

Btw, let's aim at keeping our noise examples as non-synthetic looking as possible. We're trying to imitate real-world surfaces here as best as we can. Many of the library filters just use one standard noise with a few parameters changed, and it kind of gets boring to look at after a while. In contrast, this excursion is about the 'noises seldom seen'. ;)

Quote
Kraellin wrote: hmmm, well, wouldnt just anything be a map input then? i mean, aside from controls, assemble and extract and those max and min components, just everything else would have map inputs, yes, no? basically, wouldnt it just be the green nodes then?

Correct. But the key is 'mapping' these 'green nodes' in a meaningful way. This is where the experimentation comes into play...

onnetz, Sun, 02 Sep 2007 05:57:09 -0400
no problem.... :-)

Crapadilla, Sun, 02 Sep 2007 10:45:41 -0400
So, what if I just ripped the innards of that Crapdillo dude's 'Old Parchment' filter and turned them into a surface? Take a look.

Turbulent Sediments?

Crapadilla, Sun, 02 Sep 2007 10:46:59 -0400
Even more turbulent...

Crapadilla, Sun, 02 Sep 2007 10:49:34 -0400
Now let's change the blend mode to... Darken!

Crapadilla, Sun, 02 Sep 2007 10:50:01 -0400
Or maybe set it to Difference...

Crapadilla, Sun, 02 Sep 2007 11:08:57 -0400
Seaweed Reefs? (Height map also used as diffuse map)

Crapadilla, Sun, 02 Sep 2007 11:43:43 -0400
For starters, try and experiment with two stone noises (use different variations!) fed into a blend. Feed the blend into the height input of the result component (Don't forget to get rid of that darned default surface color!). See how by using different blend modes, you can drastically change the look of the resulting surface.

Especially usable blend modes are:

- Normal 50% (100% would be a bit boring, no? ;) )
- Darken & Lighten (Extremely useful for carving out and smoothly 'glueing on' detail, respectively. Try it out on some zero detail stone noises to see the effect)
- Difference (can give interesting results)

- Multiply, Screen, Overlay (might be worth a try for the experiments sake)

Note that the various 'burn', 'dodge' and 'light' blend modes usually introduce 'clamping/clipping' to the height map...

That's bad! :puke:

Crapadilla, Sun, 02 Sep 2007 12:28:14 -0400
Mossy Rock

Kraellin, Sun, 02 Sep 2007 14:51:56 -0400
oh, that last one is excellent!

ok, here's a variation of what dilla is doing. i didnt use a blend mode, though i can see that might work better. what i did was use 3 perlins and plugged one into the noise node of the third and the 2nd plugged into the background node of the third. you lose the blend modes that way, but can still get some interesting results.

this first one is with NO surface height. it's just set to 0 to show what a bland thing this is.

Kraellin, Sun, 02 Sep 2007 14:52:57 -0400
this one is the same at the last but i kicked the surface height slider up to 83. this was in the 'day entrance' lighting environ.

Kraellin, Sun, 02 Sep 2007 14:53:35 -0400
and this one is the same as the last one, but i changed the environ to 'church'.

Kraellin, Sun, 02 Sep 2007 15:45:21 -0400
ok, i went back into mine and re-arranged things to add the blend. i also mapped both noises (the first two) into each other's roughness per dilla's instructions on roughness. yes, you can do recursive mapping on a 'faked' basis. you just have to use a third and duplicate of one of the others.

granite walls, anyone?

Crapadilla, Sun, 02 Sep 2007 16:38:37 -0400

Quote
Kraellin wrote: granite walls, anyone?

Now we're talkin' Noise Lab! :D

Crapadilla, Sun, 02 Sep 2007 17:05:03 -0400
Here there be spiders...

Kraellin, Sun, 02 Sep 2007 17:11:24 -0400
thanks, dilla. you've given some excellent tips in all this. i simply borrowed some. the 'granite' showed up when i hit the randomizer. i was very pleasantly surprised. i've used perlin and other noises quite a bit, but those extra tips really added some pizazz.

and you've got some great examples here. that spider piece i would call asbestos fiber. very cool :)

and the diversity of these is really quite amazing. i didnt even use any distortions to get get mine. but, i'll be adding some in. i did try one with an elevation gradient, but didnt like that particular line of thought on that one, but i've found in the past that the elevation gradient is a great way to take any of these noise filters and add color to them. you simply plug the noise into the 'elevation' and the gradient components into the 'gradient' node, and voila', intstant colored noises.

good thread :)

Kraellin, Sun, 02 Sep 2007 17:28:55 -0400
this one isnt quite as spectacular as some, but it reminded me of driftwood, so i'm throwing it in. i also added in two noise distortions, one after each perlin. what's nice about the noise distortions is that, naturally, you get all the distortion, but you can also zero it out and have just whatever effect the perlins produce without the distortion. and, they are fast, so you lose almost nothing by adding them.

Crapadilla, Sun, 02 Sep 2007 17:29:47 -0400
Thanks, Kraellin.

I was hoping that this thread would spark some interest in noise experimentation, so that we are all spared the "drag-slap-dash-done" noise filters in the future (of which there are too many, IMO).

Also, creating interesting and realistic looking noise-generators isn't hard, and doesn't take wiring hundreds of components. In fact, it takes just a handful of them, carefully selected and tweaked, and I've given all the hints as to which ones I am talking about. ;)

Crapadilla, Sun, 02 Sep 2007 17:31:14 -0400
Thick Plaster? Rough chiseled stone walls?

Kraellin, Sun, 02 Sep 2007 17:47:03 -0400
ah, good rough plaster, though you've got a bit of clipping in this one. still, that's covered by the fact it's plaster and the trowel simply smoothed that part :)

ok, here's my work environment via a screenshot. notice the custom lighting environment and how it affects the effect. this is another aspect of these noise generators that shld be stressed; they are a great way to look at and understand how surface filters work and how lighting environments work. you really get a good look at the process with these.

Kraellin, Sun, 02 Sep 2007 17:50:27 -0400

Quote
Also, creating interesting and realistic looking noise-generators isn't hard, and doesn't take wiring hundreds of components. In fact, it takes just a handful of them, carefully selected and tweaked, and I've given all the hints as to which ones I am talking about.

yup. i made that granite one with just four components and the results component, 3 perlins and a blend.

Crapadilla, Sun, 02 Sep 2007 18:02:11 -0400
When blending three or more noises, working with different blend modes on each single blend helps give the surface much more variety, and may yield surprising effects.

In this example I used three noises and blended with lighten and difference blend modes:

Chiseled

Kraellin, Sun, 02 Sep 2007 18:11:01 -0400
cool. where are you putting your blends? you take 2 noises and blend those 2 together and then blend that with another or some other way? and yeah, the variety is near infinite.

Kraellin, Sun, 02 Sep 2007 18:11:51 -0400
ok, here's one that reminds me of fossils. i had to turn the anti-aliasing back on for this one. was getting too much noise.

Kraellin, Sun, 02 Sep 2007 18:19:09 -0400
allright, these are ridiculously addictive! must...put...down...filter forge...

how bout some brain coral? :)

Crapadilla, Sun, 02 Sep 2007 18:20:23 -0400

Quote
Kraellin wrote: where are you putting your blends? you take 2 noises and blend those 2 together and then blend that with another or some other way?

Exactly! :)

Mini-tutorial for sci fi tiles in FF

Mike Blackney — Fri, 01 Mar 2013 07:19:50 -0500

Mike Blackney, Fri, 01 Mar 2013 07:19:50 -0500
I made a mini tutorial for some of the guys on Polycount on creating science-fiction-y textures for games in Filter Forge.

I thought I'd share it here, though a lot of you guys won't need it :)

http://mikeblackney.com/?p=337

Need help designing a filter

Ouroboros — Wed, 27 Feb 2013 12:46:56 -0500

Ouroboros, Wed, 23 Jan 2013 10:59:12 -0500
Hi Folks,

I am looking for someone to help me with a painting stippling/pointillism filter. The stipples would not touch each other and be semi-random painstrokes. I can post examples if need be, thanks

Paul

CFandM, Wed, 23 Jan 2013 12:13:00 -0500
Welcome Ouroboros and yep example images will help out more then just a description...People would have a better understanding of the look you are going for with example images...

Ouroboros, Wed, 23 Jan 2013 14:54:37 -0500
Oops here it is

Ouroboros, Wed, 23 Jan 2013 14:57:07 -0500
The stipples bring out the highlights and shadows of a flatter image. Ideally their direction could be controlled by the user so they can wrap around objects.

CFandM, Wed, 23 Jan 2013 16:03:15 -0500
You can study this one and or modify it.. :) It will give you a real good start on making the filter....
http://www.filterforge.com/filters/9817.html

Heres what the mod looks like of the above filter....

Ouroboros, Thu, 24 Jan 2013 00:24:54 -0500
Amazing! thanks a lot FS, I created my first filter its pretty easy. Still trying to figure out how to add a bit of twisty randomness to the 5 particles in bomber...

CFandM, Thu, 24 Jan 2013 08:25:14 -0500
Heres the mod of the filter that I did for the above image....

Ouroboros, Thu, 24 Jan 2013 08:57:09 -0500
Thanks! Any idea idea how to randomize slightly the shape of each particle?

CFandM, Thu, 24 Jan 2013 09:40:42 -0500
There are a couple of ways that I can think of at the moment but they are not efficient and will slow the filter down to a crawl. You can control many things in the bomber by connecting other components as seen in the filter. Each individual particles shape is not one of the easiest things to do. However you can control the shape before you plug it into the bomber itself. Then use the "Chance" and "Randomize" of what shape shows up...Maybe someone else would weigh in on this subject.
Heres the mod again with variations in the shape.

CFandM, Thu, 24 Jan 2013 10:11:58 -0500
Here is one method that is more of a distortion for each shape after the bomber. Change the color on the gradient for more distortion....Although its not really a control for each shape and it does distort the image a bit it gives a little more variations in the shapes....

Ouroboros, Thu, 24 Jan 2013 19:36:48 -0500
I just tried implementing what you have on the pic, but the image gets garbled. Do you think you could post the file for this method? thanks

CFandM, Thu, 24 Jan 2013 22:43:31 -0500
Yep here it is....btw one needs to have the percentage of the offset at 2 or 3 so the image will not distort as much..

Ouroboros, Fri, 25 Jan 2013 09:16:27 -0500
thanks, I have a lot of experimenting to do now

CFandM, Fri, 25 Jan 2013 11:18:40 -0500
You are welcome Ouroboros...Thats one of the funs bits of FF all the experimenting.. ;) :)

SpaceRay, Sat, 26 Jan 2013 11:51:22 -0500
This looks good, will see how this develops, good idea.

Ouroboros, Mon, 11 Feb 2013 07:18:13 -0500
I finally found time to start experimenting with this. Do you guys know if its possible to have the particles on a transparent background? sorry for the newbie question

Skybase, Mon, 11 Feb 2013 08:32:34 -0500
Totally possible :) FilterForge does come with a ton of freedom so I suggest giving a shot. Most questions you got will be answered! :)

CFandM, Mon, 11 Feb 2013 08:38:11 -0500
Yep here is that first filter again with a transparent background...There are a couple of ways to do it....First connecting a get alpha component to the opacity on the bomber...Or having a threshold in the mix after the bomber...Or as this version have those both for more control....Use the threshold adjustment on the threshold component to adjust how much of the image to make transparent.....
It also helps to load an image with a transparent background... ;)

Ouroboros, Wed, 27 Feb 2013 12:36:09 -0500
Thanks a lot I have been having fun with this. So far I am working on something for a silk screen with 3 colors. I am trying to replace halftone gradients with something that looks more handmade (see below). I used the "size" function in bomber to do this.
Question: is there a way (i.e. with scripting) to recreate gradients, not by changing the particle size, but by varying its density (as with stippling or voronoi stippling), thanks!

Ouroboros, Wed, 27 Feb 2013 12:37:10 -0500
Image

Ouroboros, Wed, 27 Feb 2013 12:46:56 -0500
Stippling sample, same size, different density:

mapping red areas larger than minimum size

Burt — Wed, 27 Feb 2013 10:27:22 -0500

Burt, Tue, 26 Feb 2013 16:26:41 -0500
I am working on a coral filter. I am trying to isolate the red coral over a certain size to attach polyps with Bomber. If you look at the white polyps circled in blue you will see that aside from 'floating' they aren't only near red areas. I have tried using Extract RGB, Threshold and the Median, Maximize and Minimum filters to create a map of just the bigger areas of red but you can see that the polyps are over areas that seem mostly green or blue but in reality may have tiny bits of red -or maybe suring rendering it 'sees' the red layer beneath the blue? Any suggestions on how I can get a more exact placement on the red coral.

Note: The red circles show another 'polyp' placed more exactly after a Smudge thanks to Sphinx and some extra demystifying for my poor brain by Skybase here http://filterforge.com/forum/read.php?FID=18&TID=11034 I am looking for a color based approach to place the white ones in the same way.

ThreeDee, Wed, 27 Feb 2013 04:08:42 -0500
I suggest using blend with difference mode and a threshold. This will give you good control over isolating a specific color.

Burt, Wed, 27 Feb 2013 10:27:22 -0500
That does look more accurate than my efforts. Thanks.

How do you know what components to choose to do what you want in FF ?

SpaceRay — Sun, 24 Feb 2013 03:42:55 -0500

SpaceRay, Mon, 18 Feb 2013 03:31:29 -0500
How do you know what components to choose to do what you want in FF ?

How have you learnt what combination of components will give the result that you are searching for?

How do you figure out and find a way to make a filter with the idea you want based on the FF components?

How could I learn what you already know and how to do it?

Sorry if this questions may sound silly, basic or perhaps dumb, but I want to learn HOW to be able to make FF filters in a better way

I know how to make simple and easy things, but many other things I can´t know and do no know HOW to learn to make them.

Quote

I have already done the following to try to learn:

- read the manual 3 times to know what each component does and so be able to use it, BUT do not know how to mix them to get the final result.

- I have also seen most of the video tutorials available and watched them carefully

- I have also seen many forum threads to learn more about making filters,

- and have also opened and looked inside lots of filters to see how they have been built, BUT do not know WHY they have been built this way, well at least some special parts of the filter.

I have seen that many of the experienced and skilled FF masters that knows how to make filters very well, knows of course how to choose all the components needed to be able to make the filter they want according to a final idea they may have, or they know how to solve a problem that there may be to add something to another filter.

I have had some experience making filters BUT they are simple ones, and most of the ones are based on bits and parts of others filters and when I wanted to make something different and add an idea I could have it has been really difficult and hard to make it and could not figure out and know which combination of components would be needed to reach the desired result.

I have had the great help of Skybase, uberzev, ThreeDee, Morgantao, CFandM, Inujima, Sphinx, Indigo Ray and Sharandra to solve some things I could not make and it seems that for you is very easy or at least not hard to find a solution if it can be done.

I think It would be a great help for me and probably for many others that are beginning and want to learn more on how to make FF filters to know how have you learnt to make them and how you know how to combine the different component to solve and make the things you want.

I really admire to all of you that have been able to learn so much and so well how to convert your ideas into FF Filters.

Thanks very much for your help

Sharandra, Mon, 18 Feb 2013 07:55:55 -0500
When I started out, I took other ppls filters apart to see how they achieved things.
Then I used the techniques I had learned from that in my own filters.
Often it was just trial and error. I had lots of fun experimenting with plugging components into each other and see what happens. The Noise Lab thread was a great source of inspiration.
Oh and ofc I read the articles about filter construction and optimization.

About how I know what components to choose - experience. Or simply trial and error.
When I know what I want, I also have a look at existing filters. If there is something similar to what I need then I might use that as a starting point.

Crapadilla, Mon, 18 Feb 2013 10:29:15 -0500
SpaceRay,

did you read this thread? It might provide a good starting point for your own explorations.

And then there's also the Beginner's Cookbook, which focuses on very basic "visual recipes".

Skybase, Mon, 18 Feb 2013 10:34:21 -0500
Yeup, what Shar said. Adding to that I think it's also important to have a goal making something. It's better being able to see where to aim rather than wandering around without much thought. You can get pretty cool results doing experiments, but for starters, I think having a goal is important than wondering what combos work.

As an example let's say we want a mossy cobblestone wall texture. I know you're going for effects but this was an easier example. Well first instead of just diving straight into FilterForge, we look for references that depict what we want. A bit of google search helps. So after picking some references we think works in our favor, we can deconstruct that texture to its elements.

A mossy cobblestone wall" consists of the of following elements: 1. The cobblestone wall. 2. The moss. 3. Extra foliage. We can then break this into more pieces. The cobblestone wall consists of the cobblestones themselves which are rough, chipped, cracked, weathered, and grayish. The motor that binds the cobblestones are light gray, concrete, and may not be present. See the image below for the full deconstruction

By doing this we start to get a clearer idea of what components we need to shop for. For the cobblestones we need the bricks component. The stone texture for the cobblestones will need to be produced by combining several perlin noise patches together. The mortar should also be created separately by similarly blending elements together. The moss is an overlay of more noise combos. The extra foliage might be taken care of with other types of noise patterns.

The point here is that by separating each element into smaller pieces we start getting a better sense of what components work in favor.

MOST of filterforge is blending together chunks of elements utilizing masks, and otherwise. We start getting into advanced techniques to save us time, but for the most part we put together elements for a final result. This applies similarly for effect filters, though effect filters tend to be simpler than texture filters.

The point is this. There is no right or wrong, but there are good practices and you can read those on the wiki. You'll have to kinda learn that on your own by observing what you think is a good example of a filter. But the idea is still the same. There are elements and there's the final output. There's a tendency for people to solve the problem in one shot, but what you really need to do is to solve the problem in chunks, not as a whole.

Skybase, Mon, 18 Feb 2013 10:43:43 -0500
[Unrelated note]
Crapadilla, I kinda take it that cookbook needs a serious upgrade. I kinda wish I knew how to edit wikis :( I can provide examples at some point.

Crapadilla, Mon, 18 Feb 2013 10:47:16 -0500
Well, the cookbook - or rather the wiki in general - needs contributors!

Regarding the editing of wikis: Clicky!

By the way, those FF1 techniques that are in the Cookbook still hold true, but people other than me could add some FF2+ techniques. I'd be very happy to see folks on the forums pick up some wiki editing duty. After all, this should be a collaborative effort!

( I've kinda given up on the "collaborative effort" ideal though, considering the fact that since the wikis inception there have literally only been a handful of contributors. Apparently, people are just not interested in creating FF learning material in wiki format. )

Sharandra, Mon, 18 Feb 2013 11:27:50 -0500
Dilla, I´d contribute, but I´m just not so great at explaining stuff. I don´t think I could add anything really helpful.

Skybase, Mon, 18 Feb 2013 12:37:27 -0500
I mean I thought about contributing to the wiki several times. What seriously hinders me is the fact that I'm just not used to its editing system. Also kinda don't have much courage to add anything. I fear I may be wrong about stuff sometimes and I seriously hate to misinform. heh.

Crapadilla, Mon, 18 Feb 2013 14:21:13 -0500
Skybase, wiki markup editing is really easy to pick up. Furthermore, with your proven track record of filter authoring excellence, why should you not contribute?

ddaydreams, Mon, 18 Feb 2013 14:38:17 -0500
I start with a goal. Then see who did something close to that already. Open it up in the filter editor and see how they did it. Then I find the other missing attributes of my dream filter in many other filters and see how they did it. Then combine parts of all. In doing so, you learn useful things.

Sharandra, Mon, 18 Feb 2013 14:43:42 -0500
Skybase, you have a great way of explaining things, so please don´t be shy and contribute!

See, your mossy cobblestone example above, it´s almost exactly what I do, but I didn´t even think of it.

SpaceRay, Sun, 24 Feb 2013 03:42:55 -0500
I am sorry that I forgot and missed to answer on this thread :blush:

WOW!! Thanks very much for all the answers and comments and showing how you would do it.

Specially to Skybase that you have taken the time to explain it all very well and in a detailed way and even made that good graphic to show it better.

Also to Sharandra that you have also learnt very well how to use FF for your own purposes and know how to get the things done well with interesting and very good and amazing realistic results.

SpaceRay,

did you read this thread? It might provide a good starting point for your own explorations.

And then there's also the Beginner's Cookbook, which focuses on very basic "visual recipes".

Thanks for the links, there are interesting and useful, I have not seen before the first one and like it, although I have found the cookbook as it is on the wiki.

-----------------------------------------
------------------------------------------
------------------------------------------

Quote
SKybase Crapadilla, I kinda take it that cookbook needs a serious upgrade

Quote
Crapadilla Well, the cookbook - or rather the wiki in general - needs contributors!

YES! I agree that it would be very good to upgrade it and have more intersting "recipes" in the FF cookbook specially from all of you that know much more from FF and could have the time and would want to help the FF community with their skills, experience, and ideas.

Quote
Sharandra Skybase, you have a great way of explaining things, so please don´t be shy and contribute

I totally agree, after having seen your FF video tutorials, you know very well how to shown and teach other how to get things done and is easy to understand and follow.

It would be a very good help, when you find a hole in your alreaby busy working day, to contribute making more tutorials or updating the wiki, and/or making more snippets or whatever you may think to help learning more about FF, of course always if you want to do it.

Would be possible to make this poligonal artistic effect with FF?

SpaceRay — Fri, 22 Feb 2013 13:36:24 -0500

SpaceRay, Fri, 15 Feb 2013 01:10:35 -0500
I have seen an advertisement of a telephone company that have done some cool and interesting artistic effect with lots of poligonal shapes, lines and colors, and I was wondering if this nice effect could be possible to be made in some way with Filter Forge, I have been looking in the filter library but there is nothing tha can be compared to this, also have searched in the forum for unpublished filters, but found nothing useful either.

found this crossed lines

The only possible similar reference, although is NOT the same, could be only this one hereA "Stones"-like noise with triangular forms?

I am refering to this examples here

I do not know if with FF would be possible to convert a source image into these examples shown above, if not possible perhaps would be possible in some other way with another software or plugin.

Thanks very much for any help you could give, will be much appreciated

SpaceRay, Fri, 15 Feb 2013 02:42:16 -0500
I forgot to say that I have already tried to make something but the result is far away from this and totally random and without any control.

This is suppossed to be the FF built-in kingfisher bird, but is only possible with transparency and without any black border edges on the poligons as I do not know how to do it.

SpaceRay, Fri, 15 Feb 2013 03:41:01 -0500
and this is the FF lifesaver that also has transparent poligonal shapes because I can´t make them solid and keep looking good.

Burt, Fri, 15 Feb 2013 10:17:50 -0500
I saw a couple filters that might get you started on this

http://filterforge.com/filters/223.html

http://filterforge.com/filters/10983.html

maybe something like those and also run a pass through the edge filter to get the black lines?

SpaceRay, Fri, 15 Feb 2013 17:43:55 -0500
Thanks Burt, the first one is not what like the ones above, BUT the second one has very good the fine black or color border edges as the ones above and this could be useful and interesting.

Quote
Burt also run a pass through the edge filter to get the black lines?

Oh, yes, I have also thought about this and I am going to try it.

ThreeDee, Sun, 17 Feb 2013 14:27:59 -0500
You can get somewhere reasonably close.

ThreeDee, Sun, 17 Feb 2013 14:46:10 -0500
Kingfisher.

Skybase, Sun, 17 Feb 2013 15:04:17 -0500
+1 for filter. Now I really want to know how it was done.

SpaceRay, Mon, 18 Feb 2013 02:00:50 -0500
WOW ThreeDee !!!! :eek: :eek: :eek: :eek: :eek: :ff: :ff:

Your examples are really well done and like them much, and are really very similar to the examples I have put on the first post.

You are really a master of FF !!!!

Quote
Skybase Now I really want to know how it was done.

YES! I also really want to know how have you being able to do this and also very fast.

Thanks very much

SpaceRay, Mon, 18 Feb 2013 03:42:17 -0500
This shown here in this thread is an example of what I am asking in this other thread

How do you know what components to choose to do what you want in FF ?

I mean, how have ThreeDee been able to know what components he should use to be able to make such good examples as shown above?

ThreeDee, Mon, 18 Feb 2013 11:11:51 -0500
I improved upon the method some.

The slowest part if the outlines, for I am doing them with "Detect outlines", not the fastest component around. Without outlines, the filter renders 5x faster.

But here are some samples from the reworked version. The previous version had 6 selectable colors, this one has 9.

The results are interesting at lower "vector" frequencies:

ThreeDee, Mon, 18 Feb 2013 11:12:17 -0500
Same without lines.

ThreeDee, Mon, 18 Feb 2013 11:12:55 -0500
With mask:

ThreeDee, Mon, 18 Feb 2013 11:14:06 -0500
And with mask, without lines.

ThreeDee, Mon, 18 Feb 2013 11:16:00 -0500
I'll make a dropshadow option and a couple of tweaks, then it will be ready for the queue...

ThreeDee, Mon, 18 Feb 2013 15:01:59 -0500
And here's the shadowed version.

ThreeDee, Mon, 18 Feb 2013 15:48:10 -0500
The filter should arrive soon to the library under the name of Polygon Fill.

This is what I chose for the default preset.

ThreeDee, Mon, 18 Feb 2013 15:57:20 -0500
Here's one of the things that happened to Mr. Lionfish.

ThreeDee, Mon, 18 Feb 2013 16:16:44 -0500
In fact, you can make patterns without an input image.

Skybase, Mon, 18 Feb 2013 16:44:07 -0500
This is just beautiful!!!

Sharandra, Mon, 18 Feb 2013 16:49:41 -0500
Ooh nice one, Threedee! :D

Crapadilla, Mon, 18 Feb 2013 17:06:15 -0500
ThreeDee, this is awesome! :beer:

SpaceRay, Fri, 22 Feb 2013 13:36:24 -0500
For anyone that want to try it, it is already available in the FF Library

Polygon Fill by ThreeDee

Thanks really very much for making it and sharing it for us here, you have done an awesome work, and even better than what I suggested in the first post above.

Impressionist Plugin

Chazzy — Thu, 21 Feb 2013 07:56:04 -0500

Chazzy, Tue, 19 Feb 2013 14:28:41 -0500
Would it be possible to recreate the Impressionist Plugin in Filter Forge?

It is an ancient plugin but still regarded as the best for making pictures look like paintings and pencil drawings. But because of it's age, it has a really small preview screen and 64 bit programs don't recognize it anymore.

It seems to use textures to achieve its look. Played around with offset and rotate but just can't seem to get it right. Can anyone help?

The actual photoshop plugin is here; http://asnailpace.com/blog/impressionist-plug-in/

lipebianc, Tue, 19 Feb 2013 14:52:06 -0500
I found about this great plugin 3 years ago (by the same time I've discovered FF :D )

Quote
Would it be possible to recreate the Impressionist Plugin in Filter Forge?

Using the bomber (and some techniques) it's not hard to come up with something similar. Probably, the great difference would be, as you mentioned, the use of external textures, that allows faster performance and a lot of variety.

Getting results you want requires some trial and error, as the preview window is VERY tiny (that was some of the resaons I've abandoned this for a while) :(

At time I got this, I rendered some presest for reference..it may help:

lipebianc, Tue, 19 Feb 2013 14:54:32 -0500
Another bunch of reference presets:

Chazzy, Wed, 20 Feb 2013 05:56:34 -0500
Thanks! Not a big fan of bomber but I'll play around with it and see what I can come up with.

lipebianc, Wed, 20 Feb 2013 11:56:54 -0500

Quote
Chazzy wrote: Thanks! Not a big fan of bomber but I'll play around with it and see what I can come up with.

A great starting point: http://www.filterforge.com/filters/9079.html

Chazzy, Wed, 20 Feb 2013 13:37:01 -0500
Excellent - thanks!

Skybase, Wed, 20 Feb 2013 14:33:16 -0500
Those pictures are inspiring for a new filter. Definitely will build at least one when I have the time!!

Ozmandias, Wed, 20 Feb 2013 16:29:40 -0500
Skybase... hope you haven't given up on this one...

Painting Effect

Skybase, Wed, 20 Feb 2013 21:23:38 -0500
Ozmandias, I haven't! :D Just that my pacing for developing filters comes after working on "normal-everyday-life-things" which typically takes the entire day so I just have to suffer and find time making stuff these days.

Ozmandias, Thu, 21 Feb 2013 07:56:04 -0500

Quote
Ozmandias, I haven't!

Super! I very much look forward to it!

Quote
Just that my pacing for developing filters comes after working on "normal-everyday-life-things" which typically takes the entire day so I just have to suffer and find time making stuff these days.

Darn those "everyday-life-things"!... They sure get in the way sometimes!

Bombing the Bomber

Burt — Tue, 19 Feb 2013 13:36:52 -0500

Burt, Tue, 19 Feb 2013 01:44:36 -0500
This is a random test I did today - it may be common knowledge and you may scratch your head and wonder what the heck I'm going on about but hey I made all these screen shots so gotta post them.

I was adding lots of details to a filter via multiple Bombers and it was crawling so I had an idea I wanted to test.

I set up the Bomber two different ways. The first 'regular' way I just fed 3 ellipses to the 5 particle slots, and used a Perlin noise for the chance.

and on the second try I fed the 3 ellipses into 3 offset components, then used a Multiblend to feed the result combined images into 1 particle slot (using multiple slots with option 2 renders the exact same image so I imagine they are one of top of the other (hmm... another idea there).

I attached the .ffxml in case you want to verify.

I set the parameters of the bomber exactly the same. The result is a rather large image but you can see them side by side.

Not only does the second option, with 4 additional components render substantially faster but to my eye, even the results are 'fuller' and appear to be more random - though logic tells me that shouldn't be the case. (OT: does anyone have a script or some way to compare render times?).

Again this may be me excited about nothing but I wanted to add a lot of critters to a coral filter and this second way seems to provide a way without making everything crawl.

ffxml file

The test was with FF 2.014 Pro.

Skybase, Tue, 19 Feb 2013 01:59:57 -0500
Yeup, you can compare render times by enabling "Show elapsed render time" in the options > interface tab.

http://www.filterforge.com/more/help/...tions.html

I'm not super technical with this part but there may be caching going on behind the scenes and those do improve render times. *correct me if I'm wrong*

Burt, Tue, 19 Feb 2013 02:02:17 -0500
Great - thanks! I guess that is a version 3 option have to try it on my laptop tomorrow but just eyeballing it and hitting "Next Variant" as fast as I can I can see a substantial difference.

Burt, Tue, 19 Feb 2013 13:36:52 -0500
Well I should have guessed it. I tried this on FF 3.0 on my newer laptop (my desktop is 10 years old) and the rendering time was essentially the same, and very fast, either way. I still like the result for method 2 better - seems more random and 'fuller' but that may just be me trying to hang on to something. :blush:

Method 1 on the left, method 2 on the right.

OT: Are there any other cool node based programs out there?

uberzev — Tue, 19 Feb 2013 00:31:16 -0500

uberzev, Thu, 05 Apr 2007 20:42:06 -0400
Perhaps for video or sound production?

Sjeiti, Fri, 06 Apr 2007 01:17:25 -0400
I thought Virtools (3D engine) had a node based 'programming' environment. I never used it myself though.

Torley, Fri, 06 Apr 2007 02:37:22 -0400
Sound-wise, there's names like Reaktor, Max/MSP, and Kyma — imagine if Filter Forge could be sped up with DSP boxes?

jffe, Fri, 06 Apr 2007 02:48:14 -0400
Filter Forge UAD ha-ha :D

jffe

Crapadilla, Fri, 06 Apr 2007 05:27:33 -0400
VVVV

James, Fri, 06 Apr 2007 05:57:31 -0400
FF's competition i guess -

http://www.darksim.com/

Doesn't look as cool as FF though IMO and i have never tried it so i dunno if it's any good :ff:

uberzev, Fri, 06 Apr 2007 06:02:29 -0400
FF has really nailed things as far as UI design. Most of those apps are downright ugly.

Quote
James wrote: http://www.darksim.com/

They don't provide a proper example of a texture generated with the software, all I see are final 3d renders.

BLUEFROG, Fri, 06 Apr 2007 09:42:17 -0400
DarkTree does make nice textures but the program is outdated and nowhere near as intuitive as FF. (What the screenshots don't show you is the options dialog attached to each node that looks like a 747's cockpit!) It is VERY powerful for texturemaking though.

Their new product (Tactile Ink) does look interesting but no word on a release (or even a beta).

-Jim

Crapadilla, Fri, 06 Apr 2007 10:51:37 -0400
uberzev,

since you asked for 'video production', there are a few node-based compositing systems you could look into, namely eyeon Software's Fusion (my favorite), D2 Software/The Foundry's Nuke and good olde Apple Shake. All of these are excellent visual tools that sport a fully node-based workflow.

Torley, Sun, 08 Apr 2007 03:32:05 -0400
I had a look at DarkTree awhile back; one thing I learned from their website I'd like to see here, is a list of notable creations where Filter Forge has been used. On this page, they laud DarkTree as having been used to make DOOM III's textures, for example.

Remember tho, a YUGE strength of Filter Forge is that it has such an emphasis on user-created content, which can come up with more excellence from diverse creative minds than any one company could hope to achieve (especially considering they must focus on bettering their product). It's lovely we have a community here to discuss things.

I wonder how Tactile Ink will be able to simulate more subtle effects, because from this page, it actually feels to me like they went all mid-90s-style hyperexaggerated with the bump mapping, lighting, beveling, etc. — in other words, looks like a grotesque gauntlet of Kai's Power Tools! Hehe. But the descriptions below sound compelling; I can think of many reasons why I'd want to do that in realtime and smoothly.

Currently, I use PhotoSEAM to help me get seamless textures, but it hasn't been updated in 6+ years. One of its big strengths is how easy it is to use with the brushes, which makes me crave being able to "paint on" effects and filters with Filter Forge selectively!

uberzev, Sun, 08 Apr 2007 03:41:18 -0400

Quote
Torley wrote: Currently, I use PhotoSEAM to help me get seamless textures, but it hasn't been updated in 6+ years.

Have you looked at imageSynth? it has some neat photo tiling features...

http://www.luxology.com/whatismodo/imageSynth.aspx

*Apologies to the FF peeps if this thread is out of line. FilterForge will always be #1 in my heart. :ff:

Crapadilla, Sun, 08 Apr 2007 04:25:56 -0400
I can only second that. ImageSynth is an excellent weapon in every texture artist's arsenal.

Carl, Sun, 08 Apr 2007 06:21:33 -0400
You asked about sound programs - Scope SDK [Software Development Kit ]enables design and creation of Sharc DSP algorithms and GUIs i.e. DSP plug ins. A wrapper gui that assembles C++ code in the background as you create the algorithms in a user friendly enviroment [ Drag and drop ] - analogue breadboard.
The software is based on the Analogue Device SDK that is used to create algorithms for many application such as medical imaging, geological survey, and the fun one missile guidance systems.......... Carl

Torley, Sun, 08 Apr 2007 12:12:05 -0400
@uberzev: I tried imageSynth, but was not impressed with how it assembled stuff together. In my experience, it had a very "chunky" look, and didn't crossfade rough edges of the little snippet squares (granted, they <i>are</i> called chunks!), giving them more of a "cut up a paper and collage it together" than melded blends. It looks good for obviously homogeneous stuff like sand or grass, but not so good when I tried really crumpled tin foil, or even wood with warping and knots. So it was interesting, but not that controllable.

What I'm currently doing with textures is taking 768x768 source files, loading them into PhotoSEAM, and using 512x512 so here's some excess material to blend between. I use PhotoSEAM's brushes to paint cross-border to get precisely the variance and seamlessness I'm looking for without having to chop the original into many little pieces. It's such a simple, straightforward way to go about things. :)

Vladimir Golovin, Mon, 09 Apr 2007 07:43:50 -0400

Quote
FF's competition i guess - http://www.darksim.com/

DarkTree is excellent. Comparing it with Filter Forge is not exactly correct, since DT is a three-dimensional procedural texture generator -- their procedurals don't require mapping coordinates when applied to objects. The downside is that DT can't do blur, sharpen and other similar bitmap-based operations because the underlying bitmaps would be huge.

Vladimir Golovin, Mon, 09 Apr 2007 07:47:19 -0400

Quote
uberzev wrote: Perhaps for video or sound production?

Conduit, a plugin for Final Cut:
http://www.dvgarage.com/prod/prod.php?prod=conduit15

SpaceRay, Fri, 15 Feb 2013 03:47:41 -0500
Please, is there any other NEW ones already in 2013 ?

As this is from 2007, I think that there should be new and interesting ones in this five years.

Skybase, Fri, 15 Feb 2013 05:33:46 -0500
This thread is seriously out of date. Ancient bump! Guess I'll write some new ones anyway because its been mentioned.

Here are some of the newer releases I know of:
PixelConduit is now free and available in the appstore! You can expand the functionality but most of what you need is usable. PixelConduit works with as well with FinalCut as a plugin so you can make your own little video plugins. Very handy! Works just like FilterForge (with less functions)
http://pixelconduit.com/

Vuo looks like a really promising as a nodey editor. Looks very simple and seems to have a lot of flexible complexity. They seemed to have aimed it for a variety of artsy productions (video, interaction, art... stuff...) but not available to the public yet.
http://vuo.org/

Blender has seriously vamped up its node-based editor since this thread has been created. There's a complete compositing suite, image editing capability, shader creation, and a ton more. The program has really gone fr om being hard to use to something a bit accessible. Definitely worth checking: http://www.blender.org/

MAX/MSP/Jitter had a major overhaul since 2007 has really helped with everything. http://cycling74.com/products/max/

Derivative Touch Designer 088 is soon to be released, the current beta is available but much of the software, for non-commercial purposes is free for use. The software behaves much like QuartzComposer and is geared towards visual experimentation and interactive design. There are a lot of technical capabilities! Lots of good stuff! http://www.derivative.ca/Events/2012/088BetaRelease/

Most of these get quite technical but the simple-surface stuff is really really easy and fun. Just when people want specifics it does get very complicated quickly and most of these unfortunately aren't as easy as FilterForge since there's a lot more data handling to do. But there's a lot of good stuff that's possible. I encourage to sit down and just learn one thing at a time. If you learn one, most others will feel the same, just naming conventions and functions become mildly different. :)

Either way 2013 looks promising for more visual programming projects but the well-known softwares are keeping things well updated. Vuo is honestly the most promising for a really well-rounded tool. Most node-based editors become biased towards a subject.

Vladimir Golovin, Fri, 15 Feb 2013 07:58:48 -0500
I'm surprised nobody mentioned Houdini:
http://www.sidefx.com/

It's a paid product, but the learning edition is free and I believe it has full functionality except for distributed rendering and high resolutions.

Vladimir Golovin, Fri, 15 Feb 2013 08:00:33 -0500
Also, Unreal Kismet:
http://www.unrealengine.com/features/kismet/

Skybase, Fri, 15 Feb 2013 08:28:52 -0500
There are so many haha. Some of them are tucked away behind the software like Xpresso in Cinema4D, the shader editor in Maya (has more capabilities than most people understand).... etc etc....

http://en.wikipedia.org/wiki/Visual_p...g_language

Skybase, Sun, 17 Feb 2013 15:41:16 -0500
Oh yeah Peacock returned as Nodewerk if you folks didn't know. It's now an Adobe Air based software: http://nodewerk.com/

*ahem* Quasimondo and friends doing more werk.

Crapadilla, Sun, 17 Feb 2013 16:41:51 -0500
I'm surprised no one mentioned Substance Designer... :dgrin: :devil:

SpaceRay, Mon, 18 Feb 2013 01:30:26 -0500
Thanks for all the links and all the answers, specially to Skybase for the long post.

By the way, it seems that you have forgot one that has been told already in some other thread, NODEBOX and GENETICA

Even having so many node based programs, I think that Filter Forge is really great and very user friendly and easy to use and that has a nice and good editor and is well done and like it much more than the one on Substance Designer and Genetica (from my particular experience) .

I really can´t tell more about others software as I have not tried them other than the two already said before here

Skybase, Mon, 18 Feb 2013 02:43:29 -0500
Nodebox is really handy for quick-shape development really. I used it a lot to generate a full set of particles in no time. Plus it's nice because it's all Python which to me is easier to understand than Lua :p

Lucato, Mon, 18 Feb 2013 17:05:52 -0500
Other would be...

- Werkkzeug3 texture
- MapZone 2.6 Editor
- ...

Skybase, Tue, 19 Feb 2013 00:31:16 -0500
Oooo those are classic. MapZone 2.6 became SubstanceDesigner actually. :)

Where's Waldo?

Indigo Ray — Sat, 16 Feb 2013 22:17:25 -0500

Indigo Ray, Sat, 16 Feb 2013 22:17:25 -0500
:D :D :D :D Oh, he's there...

3D

Ramlyn — Fri, 15 Feb 2013 15:46:21 -0500

Ramlyn, Fri, 08 Feb 2013 14:04:39 -0500
I don't know if this has to be a request for FF4 or a suggestion for FF in general.
Anyway..
Let's suppose we have a picture to work on. Just a normal picture, with people and a landscape. The face of the people is nearer. The landscape is not on the same distance as the face of the people. How to apply the same effect to the whole photo keeping the different feeling depending on the distance?
We could use Photoshop. Divide it in layers and after using FF on every layer, but the result would be so so.
Any other idea?
Could a new FF4 work on different layers as they were very far ? How to say.... If we have a person as main subject of the photo, and a background as something very far... Can FF works on them and set a unique effect? I tried the effects of FF3, but, even helping with Photoshop, the result is not satisfying.
Yes, doing a lot of job with Photoshop, we can clearly get a better result, but that's not my idea. I just wonder if the new FF could do a smart work on the distances ( for example, selecting two layers, one as main and a second as far layer )

Ramlyn, Fri, 08 Feb 2013 14:15:27 -0500
I'm sorry if I cannot explain very well. I hope everybody got an idea of the meaning.

For example, let's think about creative filters. If we could simply choose two areas, one for the front and another for the background. Then choosing the distance between them. This would allow the filter to set the effect. Depending on the distance, the filter could set the perspective lines in different way

Ramlyn, Fri, 08 Feb 2013 14:18:10 -0500
Using FF as Photoshop plug-in, it could be surely done, but I don't know if it is possible as stand alone program.

ThreeDee, Sat, 09 Feb 2013 03:53:15 -0500
Interesting question. I wonder if there is a way to create a script that can use a z-depth channel to vary the amount of effect. Similar to a variable blur (lens blur in Photoshop) but which can be plugged in as a component to any existing filter...

My thought, although it will probably not work as I expect, would be to have a script that, from the viewpoint of the FF rendered, enlarges the image more the further away it is (the deeper it is in z-depth map) and thus creates an effect with a smaller radius for further-away objects. Don't know if that can actually be done.

Sphinx., Sat, 09 Feb 2013 06:06:19 -0500
I've been thinking about something similar for years, i.e. how to "detect" the depth of field in a natural photo - with relative post processing in mind of course.

So I'd say the first problem here is how to determine the depth - I thought about somehow using correlation of frequency changes in different color spaces. This is not easy, and my attempts so far did not produce anything useful (too much noise/errors).

Skybase, Sat, 09 Feb 2013 08:13:14 -0500
Here's something interesting that was at NAB a while back. http://www.yuvsoft.com/stereo-3d-tech...h-effects/

They also seem to have figured a method extracting depth from focus, which is really nice.

I did some compositing work in the past and I ran into frequent issues manually dealing with depth maps. During some research that came up. I don't have that plugin myself, nor do I really know how it works, but it's definitely what's being discussed. Quite difficult I'd say!

Sphinx., Sun, 10 Feb 2013 05:25:16 -0500
Skybase - that is exactly what I meant, i.e. frequency ranges change over depth due to focal change. Lets get a challenge going :-D

ThreeDee, Mon, 11 Feb 2013 13:03:19 -0500
That certainly seems error-prone. For instance, how do you tell the difference between something that is out of focus versus something that is rounded with a similar radius? Even image noise can affect the result.

I would imagine it would be "easier" fr om a stereo pair of images where you can compare the locations rather than the focus.

As a first step I would try addressing a situation wh ere you already have a z-depth map in existence. Would there be a way to create a script that used that to affect the amount of effect from a filter? Ideally you could plug it in as the last component before Result (or first after External Image?) in any existing filter tree.

ThreeDee, Mon, 11 Feb 2013 14:53:40 -0500
No, hold on a second: Does it even matter whether the area is out of focus or just lacking sharp edges?

If the effect is dependent on how sharp the image area is, it might not matter whether the area is out of focus or just without much high-contrast detail, in many cases the results would likely be visually similar.

Which makes me think that some High-pass-related solution might be adequate. In other words, where the image result run through the High-pass component is close to mid gray the effect would have a larger radius, and the further from mid gray the High-pass result the smaller the radius.

Sounds simple enough.

Sort of like unsharp mask or smart blur.

ThreeDee, Mon, 11 Feb 2013 15:03:01 -0500
Well, it didn't take long to figure out the first problem with that simple approach: You wouldn't be able to scale the filter output that way. Thus patterns would appear the same size no matter where they were in the image.

ThreeDee, Mon, 11 Feb 2013 15:13:27 -0500
Okay, so here's another thought:

Most images have a certain type of perspective, the most common being a ground plane with a sky (or back wall). In this scene there can be individual objects.

You could have a few simple inputs for the basic layout, such as horizon height and back wall distance, perhaps focal distance. This is the background set.

And then, wherever there are areas that have sharper details than a given threshold (different enough from expected background focal values) you assume there to objects. All the rest is background.

Perhaps this method could be used to construct a basic depth channel with just a couple of inputs?

ThreeDee, Tue, 12 Feb 2013 11:39:45 -0500
I did some quick tests on this idea. I seem to get a "focal map" rather than a depth map as a result. One problem is that the objects in focus are not necessarily the closest objects in the scene. When they are in the middle ground of depth with out-of-focus objects in the front and in the back, there is no way to distinguish between foreground and background. An initial mask for foreground vs background, even rough, would help the matter considerably.

I only now realized to check out the YUVsoft link. Very nice results. They also (can) use a rough mask and improve upon it.

Skybase, Tue, 12 Feb 2013 22:38:46 -0500
This thread has become ThreeDee talking to himself thread. D:

YUVsoft seems to just basically put all what's being discussed here into a package. I really want to try their products since it really adds to what I do. :( But they probably cost a few thousand bucks.

Sphinx., Thu, 14 Feb 2013 05:24:21 -0500
Well, I'm reading ;-)

I got the experiments going too, but so far I've not been able to produce anything useful (though some quite interesting outputs were produced). What I'm trying to do is this: For each sample, look at the relative difference fr om different low pass steps. The idea is to find the lowest acceptable (by some threshold) frequency relation.

Frequency variations in image data is clearly highly related to the actual objects and patterns etc., but the changes in focus is to be considered as a frequency sweeping low pass filter, where the given low pass frequency is given by the distance of the objects in the scene - lets call this focal point F.

Now all the variations in the actual scene will at any point be restricted to 0.. F in possible frequency. This fact is what I'm trying to make use of..

Of course there will be issues when the focal point is far away from the lens, but even so the output is useful, and the problem of correcting that error seem simpler than the current.

Edit: Here is a page with some general depth of field introduction:
http://www.cambridgeincolour.com/tuto...-field.htm

Sphinx., Fri, 15 Feb 2013 04:53:55 -0500
Here is one of the strange bi products of the experimentation. The input image is the Ladybug image :-D

Sphinx., Fri, 15 Feb 2013 05:26:42 -0500
Here is the best variant so far - it pretty much uses the approach described above, only it searches for the lowest acceptable low pass frequency within each sample.

Skybase, Fri, 15 Feb 2013 05:35:34 -0500
Ooohhhh ho hoho..... these results are look seriously promising. So I've been wondering... how long does it take to process the image?

ThreeDee, Fri, 15 Feb 2013 05:58:30 -0500
Sphinx,

That is certainly more scientific an approach than mine. For my first tests I used a down-and dirty approach of edge detection plus percentile to smooth out the results.

Here is what I get with the kingfisher image:

ThreeDee, Fri, 15 Feb 2013 06:14:25 -0500
With less smoothing.

Sphinx., Fri, 15 Feb 2013 08:57:42 -0500
Found a bug in my algorithm.. here is the new raw result - it could probably use a median filtering or something. The processing time is about 30 secs, but it depends on settings (number of low pass searches).

Sphinx., Fri, 15 Feb 2013 09:20:51 -0500
ThreeDee - that looks good! We both seem to have a problem with the "fill" though. Here is the same algo and image with some other settings.

ThreeDee, Fri, 15 Feb 2013 10:19:11 -0500
Sphinx,

I suggest we include a optional mask input. The purpose of the mask is to distinquish between foreground and background objects (White for foreground and black for background). As objects in the background as well as objects in the foreground can be out of focus, there needs to be a way to separate them out from each other.

I believe this will greatly improve the resulting depth channel output, and can likely also be used in estimating whether a fill is needed at some location.

For instance, if we used a mask for foreground vs. background on the kingfisher image, you could easily determine by sampling the mask that there is a fill needed in the missing areas of the bird, as that area is all foreground.

It is a pity that you can't access the masks of the built-in preview images, for that would work here.

ThreeDee, Fri, 15 Feb 2013 10:22:08 -0500
You could then force the z-channel for foreground objects to be lighter than (for instance) 50% and background objects darker than 50% gray value. The actual value could be adjusted based on focal depth.

CFandM, Fri, 15 Feb 2013 12:35:31 -0500
You could always do this for masks....It the way I had to do for some filters..
Or is this what you are thinking :?:

ThreeDee, Fri, 15 Feb 2013 14:29:48 -0500

Quote
CFandM said: You could always do this for masks....It the way I had to do for some filters..

I'll be darned. Exactly that -- thanks! Didn't think to try that method. External selection never worked, so I had given up on it.

It was that well-hidden "Use transparency" option that I missed :)

ThreeDee, Fri, 15 Feb 2013 14:37:05 -0500
I guess you could consider it cheatin', but adding the mask gives me this:

ThreeDee, Fri, 15 Feb 2013 15:15:04 -0500
Faking it even further by blending some of the brightness channel information and blurring the mask for additional roundness:

CFandM, Fri, 15 Feb 2013 15:46:21 -0500

Quote
ThreeDee Wrote: I'll be darned. Exactly that -- thanks! Didn't think to try that method. External selection never worked, so I had given up on it. It was that well-hidden "Use transparency" option that I missed smile:)

Glad I could help in some fashion with this one.. ;)
Yep that option in hidden well... :D

Fractal Tessellation Challenge

uberzev — Fri, 15 Feb 2013 01:41:04 -0500

uberzev, Fri, 17 Aug 2012 00:38:24 -0400
Fancy Example...

Simple first 3 steps...

The scaling is 1/sqrt(3) = 0.5773502692 for subsequent steps.

More info...
http://www.mathartfun.com/shopsite_sc...pedia.html

SpaceRay, Fri, 17 Aug 2012 00:48:47 -0400
Cool and nice

and very interesting and useful link

Also in this same website there is this other great math art examples

http://mathartfun.com/shopsite_sc/sto...index.html

uberzev, Fri, 17 Aug 2012 00:54:24 -0400
8)

ThreeDee, Fri, 17 Aug 2012 04:52:22 -0400
Hehe. I haven't recovered from the last challenge (Hyperbolic Tiling) yet...

This is actually easier that the previous one. And with the new Components in V3 it should be considerably faster and easier than in earlier versions.

Challenge nonetheless.

Ghislaine, Fri, 17 Aug 2012 10:25:32 -0400
And this futur filter is for when ??? :)

ThreeDee, Fri, 17 Aug 2012 13:19:02 -0400
Today?

Ghislaine, Fri, 17 Aug 2012 13:35:09 -0400
Cool ! :)

ThreeDee, Fri, 17 Aug 2012 14:06:03 -0400
Ripping off the unit tile from the first sample...

ThreeDee, Fri, 17 Aug 2012 14:07:04 -0400
And adding inverse to every second tile (slightly more complicated), we get:

Morgantao, Fri, 17 Aug 2012 16:07:04 -0400
You da man!

SpaceRay, Fri, 17 Aug 2012 22:28:30 -0400
Nice and well done ThreeDee !!!! you have been able to do it !!!

uberzev, Sat, 18 Aug 2012 02:35:31 -0400
Nice job 3d!

Don't post your filter yet. I've nearly got this figured out and don't want to be accused of copying! :blush:

uberzev, Sat, 18 Aug 2012 02:56:42 -0400
Not quite there yet but this is an interesting variation...

uberzev, Sat, 18 Aug 2012 03:12:48 -0400
Can you make an image like this ThreeDee?

(might be good as one of the defaults for your filter)

SpaceRay, Sat, 18 Aug 2012 03:28:53 -0400
Good one and well done, I like the black lines that defines the polygonal shapes

I wonder if it would be possible to have MORE different shapes than this one using the same technique or algorithm?

Perhaps like the star shapes or a triangular shape ?

uberzev, Sat, 18 Aug 2012 03:32:45 -0400
The technique would be similar but it would be a different algorithm. As the scaling/rotation/offset would change.

ThreeDee, Sat, 18 Aug 2012 03:43:24 -0400

Quote
Uberzev wrote: Can you make an image like this ThreeDee?

Similar. Since I scale the input, the lines shrink down as well (although that could be dealt with if needed). Here I mapped the spectral gradient into it.

uberzev, Sat, 18 Aug 2012 03:56:00 -0400
Looks very good! (but the formula is a bit different that the one I was referring to)

I realize now that you've re-engineered the "fancy" spiral version as apposed to the the more homogenious one. That's the one I've been working on so you haven't completely stole my thunder yet. :D

Your version seems to have a "sharper falloff" (for lack of a better term). The next levels only have 18 copies vs 24 for the rounder version.

ThreeDee, Sat, 18 Aug 2012 04:03:13 -0400
Oh, I see -- I missed the obvious difference! :blush: Well, it should be just a different offset and/or scale value.

uberzev, Sat, 18 Aug 2012 04:04:40 -0400
Guess you want all the glory. ;)

SpaceRay, Sat, 18 Aug 2012 04:13:22 -0400
Looking at the color examples from uberzev and Threedee, I prefer much better the uberzev result that is cleaner and looks more elegant and perfectly done, although I think that ThreeDee now what is happening with his version and could perhaps fix it to be more like the nicer uberzev´s version

Let´s see who wins ;) :D

uberzev, Sat, 18 Aug 2012 04:20:27 -0400
My Version is "borrowed" from the Fractal Tilings website so ThreeDee wins by default. ;)

As far his ugly edges that's a pretty basic fix but not important right now.

SpaceRay, Sat, 18 Aug 2012 04:39:34 -0400

Quote
uberzev wrote My Version is "borrowed" from the Fractal Tilings website so ThreeDee wins by default.

I see, so this explains why this is perfect and beautifully done ;) :D I thought that you have done it yourself in FF

ThreeDee, Sat, 18 Aug 2012 05:14:42 -0400
This is more like it.

uberzev, Sat, 18 Aug 2012 05:22:25 -0400
Damn you! :evil: :D

I've got everything working for the first four iteration levels. The fifth gets wonky though. :hammer:

uberzev, Sat, 18 Aug 2012 07:49:33 -0400
No lines yet, but I finally got all my hexagons under control. :D

ThreeDee, Sat, 18 Aug 2012 10:17:54 -0400
Looking good. I imagine we have rather similar constructions -- there ain't that many ways of doing this (discounting scripting).

I doubt I will do a version with unified line weight. Which is really a matter of different thresholds at different levels. I like the fractal style where everything scales down.

Here's a test from the "fancy" version with a different unit tile.

uberzev, Sat, 18 Aug 2012 15:07:09 -0400
That looks fabulous ThreeDee. 8)

Morgantao, Sat, 18 Aug 2012 15:43:25 -0400
Very good, guys! :)

Sharandra, Sat, 18 Aug 2012 16:00:59 -0400
Wow, sweet! :D

uberzev, Sat, 18 Aug 2012 17:27:26 -0400
Still need to add image support but that shouldn't be too hard.

Check out how simple I've got it. 8)

uberzev, Sat, 18 Aug 2012 17:53:36 -0400
:D

uberzev, Sat, 18 Aug 2012 17:57:52 -0400
Detail...

uberzev, Sat, 18 Aug 2012 18:05:00 -0400
:!:

uberzev, Sat, 18 Aug 2012 18:21:45 -0400
Still a but rough but you can see the potential...

uberzev, Sat, 18 Aug 2012 18:40:24 -0400
Tool's next album cover? :D

SpaceRay, Sat, 18 Aug 2012 21:11:55 -0400
WOW !!! :eek: :eek: :eek: :eek: I like very much the last 4 examples shown and look really beautiful and very well done, and the tesselation and tiling is excellent.

Congratulations uberzev!

Skybase, Sat, 18 Aug 2012 23:37:37 -0400
... so when I saw the way it was done I was like "oh" :| hahaha the hell have I been doing for the last couple days. It's just a simple for loop gosh D:

uberzev, Sat, 18 Aug 2012 23:54:16 -0400

Quote
... so when I saw the way it was done I was like "oh" hahaha the hell have I been doing for the last couple days. It's just a simple for loop gosh D:

lol, I spent about 8 hours doing it wrong before I figured it out. :evil: :hammer:

:banana:

Skybase, Sun, 19 Aug 2012 00:01:02 -0400
I mean the "getting right" part IS HARDCORE. But the way you did the for loop never occurred to me. :| It was this loud sigh of "oh" and then me yelling "... the f that's better than what I'm doing." :(

Pretty jelly here

Morgantao, Sun, 19 Aug 2012 01:21:48 -0400
I like those what you've dome here uberzev!
This looks really cool :)

ThreeDee, Sun, 19 Aug 2012 03:15:19 -0400

Quote
Uberzev wrote: Check out how simple I've got it.

Whoa, no kidding! I was gonna post a "check out how complex I've got it" but I leave it to your imagination ;)

uberzev, Sun, 19 Aug 2012 04:21:22 -0400
Accidentally deleted something and this came out...

SpaceRay, Sun, 19 Aug 2012 09:20:41 -0400
I have found this beautiful Fractal tesselation here in this artists website

And also this other one here

here

Skybase, Sun, 19 Aug 2012 09:49:28 -0400
Oh Apophysis, you show up anywhere you hear "fractal"

SpaceRay, Sun, 19 Aug 2012 12:45:56 -0400

Quote
Skybase wrote: Oh Apophysis, you show up anywhere you hear "fractal"

Are this examples I have put made with Apophysis software?

I have seen before it although never used it yet.

SpaceRay, Fri, 15 Feb 2013 01:41:04 -0500
What happened with this?

Please uberzev and ThreeDee, has this been a private and personal project that it is not going to be a filter available in the library or the forum and only for your personal use?

Of course that I understand and agree if the answer is yes, but if not I would like to know how you have done this awesome examples shown above and if the filter will be available in the library sometime.

Thanks very much

Billboard images and placing bomber elelemnts 'on top'

Burt — Thu, 14 Feb 2013 20:58:07 -0500

Burt, Thu, 14 Feb 2013 20:58:07 -0500
DELETE

Help with boat wake filter

Cannos — Thu, 14 Feb 2013 11:51:24 -0500

Cannos, Wed, 13 Feb 2013 19:24:54 -0500
I just started using Filter Forge and I'm currently trying to make a boat wake filter. I'm making progress, but not quite getting what I'm looking for yet. Below is an image comparing the desired effect versus my current filter.

Right now I just have a single layer of noise that's been mirrored and blends out on the edge. I think I need a couple more layers of noise, but the main thing I haven't figured out is how best to make it not so straight on the edges. I want the edges to scale in and out a bit and have a harder edge like in the desired image.

I'll take any advice or tips I can get. Thanks!

Sharandra, Thu, 14 Feb 2013 05:24:35 -0500
You could use another noise to control the profile gradient´s curve.

Cannos, Thu, 14 Feb 2013 11:51:24 -0500
Thanks Sharandra, that looks promising. I'll try that out.

Cell color from cell area

Zoltan Erdokovy — Fri, 08 Feb 2013 13:18:11 -0500

Zoltan Erdokovy, Thu, 07 Feb 2013 05:50:47 -0500
I have this plain solid filled cell noise like on
the attached image. What I'd need is gray values
of those cells corresponding to cell area: small
ones would be dark, big ones light. Any pointers?

ThreeDee, Thu, 07 Feb 2013 10:42:36 -0500
You need them output in what fashion? As a list of values? Do you need to know where in the image the values are? Or just a list of every value that is different? Do they need to be in sequence from small to big or some such thing?

Zoltan Erdokovy, Thu, 07 Feb 2013 12:14:09 -0500
I mean I need the cells to be colored according to their size.
(And not random like what the default noise node produces.)
The bigger the cell is the brighter it should become.

Formally:

[CODE]Cell.Color = float3(Cell.Area/BaseAreaConstant);[/CODE]

Of course FForge doesn't work like this but this shows the idea.

Sphinx., Thu, 07 Feb 2013 15:42:36 -0500
Some time ago I figured out a simple (but not perfect) way of finding some sort of area proportional output. It is not precisely what you ask for.. but it might be what you're looking for :-)

Note that this does not work for 'flat' mode.

Zoltan Erdokovy, Thu, 07 Feb 2013 16:34:23 -0500
Thanks Sphinx, is that for FF4? I'm kinda reluctant to upgrade, not fond of beta testing.

Sphinx., Fri, 08 Feb 2013 13:08:25 -0500
Here is an FF3 version with two methods. Note that none of them are perfect, but they're roughly expressing cell size..might be good enough (the method is dead simple and fast.. gotta count for something :-D)

Zoltan Erdokovy, Fri, 08 Feb 2013 13:18:11 -0500
Wow, very clever, I'd never though of this... Cheers, this should
keep me rolling. :)

int control question

ronviers — Fri, 08 Feb 2013 02:50:20 -0500

ronviers, Wed, 06 Feb 2013 22:44:38 -0500
I am having a difficult time getting the troughs of the sin node to stay matched up with the transitions of the step node over a range. Has anyone else worked something like this out?
I would like the number of waves to be from 1 to 300.

thanks:)

ThreeDee, Thu, 07 Feb 2013 10:55:12 -0500
Hi Ronviers,

The main problem seems to be that the frequency of the Sine node only goes up to 100. Here's the filter with range 1-100.

I put an additional Math node there so that the first circle is not black. You can take that out if you want.

If you want 300 waves, have to do this a bit differently, with 3x repeat and elevation gradient mapping.

TD

ronviers, Fri, 08 Feb 2013 02:50:20 -0500
hi ThreeDee. Yes, 300 ripples complicates it. neat trick to use the divide node that way but i do not understand what it's doing, didn't realize we could use hdr outputs that way.
Thank you very much:)

Census of Active FF 1.0 Filter Authors

GMM — Wed, 06 Feb 2013 17:06:28 -0500

GMM, Fri, 01 Feb 2013 08:08:04 -0500
Hi guys and gals,

With FF 4.0 in active development, we don't have much resources for older versions. In particular, we are somewhat reluctant to update FF 1.0 so that it would conform to new filter submission requirements that are to be introduced in 4.0.

To my (personal) surprise ~~Jffe~~ certain filter authors still submit filters from the first version. So, my questions are:

- how many conservatives are there in the active community?
- would it be troublesome for you to switch to 2.0, 3.0 or 4.0?
- would you be willing to do a little bit of extra editing in 1.0 because certain component combinations will no longer be accepted by the server?

FYI: users with FF 1.0 constitute a single-digit percentage of our library server traffic.

Crapadilla, Fri, 01 Feb 2013 09:06:44 -0500

Quote
[...] new filter submission requirements that are to be introduced in 4.0.

Quote
[...] certain component combinations will no longer be accepted by the server?

Could you elaborate on this?

Vladimir Golovin, Sat, 02 Feb 2013 06:56:47 -0500
GMM, is this about the situation where the server-side Problemator rejects filters submitted from versions prior to 4.0 due to the IntSlider range warning that was introduced with the new IntSlider?

In any case, a better way to figure this out would be via an SQL query to the filter database, something along the lines of: SHOW ME ALL FILTERS WHERE SUBMISSION DATE IS WITHIN RECENT 3 MONTHS AND VERSION IS LESS THAN 4.0.

(Or, if the version data isn't available as a field, and it should be, use a full-text search in filter's XML code: "... AND FFXML CONTAINS VERSION TAGS FOR FILTERFORGE V1 OR V2 OR V3.")

Sphinx., Sat, 02 Feb 2013 07:15:50 -0500
(I guess that must be FFSQL :-D)

I'd say as long as you offer a free upgrade path then this sort of change is ok.

GMM, Mon, 04 Feb 2013 03:07:45 -0500

Quote
is this about the situation where the server-side Problemator rejects filters

Yes.

Quote
In any case, a better way to figure this out would be via an SQL query

Done that, I know for sure there are recent filters submitted from 1.0 — and I'd like to know the community response for that.

SpaceRay, Mon, 04 Feb 2013 17:56:43 -0500
Is there really many others using FF 1.0? FF 2.0 and FF 3.0 adds SO MUCH MORE to the program that is a great pity to keep using FF 1.0, I understand that some people maybe can´t pay the upgrade, but really is not expensive and even more considering the high upgrade discounts available, and it adds really MUCH more AND of course you can use MANY MORE THOUSANDS OF FILTERS.

I wish and want that the owners would be able to upgrade to a higher version to enjoy more about FF and do not keep using the old FF 1.0 and send them my best wishes

GMM, Tue, 05 Feb 2013 03:24:57 -0500

Quote
that is a great pity to keep using FF 1.0, I understand that some people maybe can´t pay the upgrade

The strange thing is that certain FF 1.0 authors have enough HU filters to buy themselves lifetime upgrades for the next several lives.

GMM, Wed, 06 Feb 2013 09:15:05 -0500

Quote
Could you elaborate on this?

Plugging Intslider to an input that would limit its range no longer works: [1], [2]. We will surely fix this in 3.0 but would prefer leaving 1.0 unchanged.

As I don't see much whining about that, conservative filter authors will probably need to upgrade to a later version or avoid plugging Intsliders into narrow inputs.

SpaceRay, Wed, 06 Feb 2013 17:06:28 -0500

Quote
GMM The strange thing is that certain FF 1.0 authors have enough HU filters to buy themselves lifetime upgrades for the next several lives.

:?: :?: :?: :eek: then if it is NOT a problem of paying the upgrade in this cases that you mention, HOW is possible that they do not upgrade and KEEP using FF 1.0? :eek: :?: :?: :?:

Perhaps instead of asking it here in the forum, if you know who they are, FF customer support could contact them as politely ask about their reasons and know more about this mistery.

Starfish generator

ddaydreams — Mon, 28 Jan 2013 15:59:07 -0500

ddaydreams, Mon, 28 Jan 2013 15:59:07 -0500
Need Delete option for this message.

Selection Example?

Burt — Thu, 24 Jan 2013 20:19:30 -0500

Burt, Thu, 24 Jan 2013 17:22:29 -0500
Is it possible to create a selection in FF? For example say I use the Brick component with black and white, could I somehow select the white, then apply that brick shaped selection to a string of components that is giving me a rock look,so that I can then create effects that are constrained by the selection area? I know I could use it as a mask but that would still be applied to the whole 'rock' image rather than the individual bricks.

If possible then can someone provide a simple screenshot or ffxml so I can see (visual guy).

Hope that is clear. At work so sneaking a moment or I'd do a diagram.

Sharandra, Thu, 24 Jan 2013 19:14:07 -0500
You can use the Threshold Component for that :-)

I made Low transparent in the example.

Burt, Thu, 24 Jan 2013 19:34:00 -0500
Right but this still functions as a mask correct? What I am thinking is for example the 'bump' some people are creating using the Smudge component, when you mask it I assume the displacement would occur on the whole image rather than each brick seperately - is that correct? Hope that makes sense.

Anyway I wanted to play a little with 3.0 and had that thought buzzing in my head.

Sharandra, Thu, 24 Jan 2013 19:53:19 -0500
You can just feed the smudge result into High, then only the white part will have the effect on it. You can then feed something else into Low for a different result on the black part.

Burt, Thu, 24 Jan 2013 20:18:48 -0500
I'm not always good at explaining what I mean so here is an example using Photoshop.

I created a simple cloud texture. I make some random selections and apply the Displace filter 10 times using the same image as a displacment map. Here is the result

Next I start again with the cloud texture, but using the same random selections I create a mask and use the Displace filter in the same fashion the same 10 times. Here is the result

Not only are the borders different but the Displace filter with the mask is applied to the Cloud texture as a hole, it just hides the parts from you. You can see the difference here using the same selection (and the cloud layer as backdrop for the deformed ones for better comparison)

Might not seem like much but might be fun to test.

Burt, Thu, 24 Jan 2013 20:19:30 -0500
Thanks for the idea. Saw it after my post. Hope it explains better. [EDIT: of course I applied the displace to the cloud texture and the mask -thump on the head... will mplay more this weekend. :blush: will try bevel ttoo see if it is applied the same - just a random idea. Thanks for taking the time to respond.]

Make a Color Control image 1-to-1 with the main image?

fisholith — Fri, 18 Jan 2013 07:07:19 -0500

fisholith, Wed, 16 Jan 2013 03:52:33 -0500
Assuming I already have an image loaded into the "Image" component (I'll call it the "main" image), is there a way to load a second image using a "color control" component, such that the second image's scale is 1-to-1 relative to the main image?

That is, the pixels in the main image and the second image are the same size.

The "color control" component has an option menu called "Placement", which has the options "Fit", "Stretch", and "Fill", but no "1:1" or "Don't scale" option.

e.g. If the main image is 400 x 400, and the second image is 200 x 200 and it's at 1-to-1 scale, then the second image should be half the width and height of the main image.

The "color control" component only gives me options that stretch the second image in various ways, but unfortunately I don't want to stretch it.

Is it possible to load a second image at 1-to-1 scale using a script? It would certainly be nicer to use an open dialogue, but if the "color control" component doesn't give you a 1-to-1 option, and doesn't tell you what the second image's dimensions are, I'm thinking there might not be a way to automatically rescale the image with further nodes.

The only thing I can think of at the moment is to open the properties of the second image in windows explore, read the dimensions, type them into value sliders in filter forge, and then use some components to do the appropriate scaling math to make the second image 1-to-1 with the main image. And that might work if I only have to do it for one image, but doing it hundreds of times seems like it could take a while
:)

Sharandra, Wed, 16 Jan 2013 08:34:54 -0500
You could use the scale component (under transforms) to scale the second image up or down.

Vladimir Golovin, Thu, 17 Jan 2013 03:18:16 -0500

Quote
The "color control" component has an option menu called "Placement", which has the options "Fit", "Stretch", and "Fill", but no "1:1" or "Don't scale" option.

This is a good suggestion. We'll look into it.

(Note to self: also implement wrapping beyond the image boundaries similar to how it's done in the Image component).

Betis, Thu, 17 Jan 2013 13:30:56 -0500
wow you actually managed to convince vlad to consider your suggestion with just one post. :dgrin: :beer:

fisholith, Fri, 18 Jan 2013 07:07:19 -0500
Thanks Sharandra, Vladimir, and Betis for the replies. :)

I've included an explanation of the current workaround I've settled on below, along with an example filter, and a walkthrough for building one of the scaling methods I used, (since some of the steps might not be immediately obvious fr om simply looking at the wires in the editor). Hopefully if anyone comes across this post with the same question I had, it will help them out. :D

A workaround:
For situations wh ere the second image is perfectly square, the workaround I'm using for now is to use a combination of the "Scale" component and a script to output the "SIZE" value of the main image, since I need to know what the main image's dimensions are if I want to resize the second image to be 1-to-1 with it. (I know there are likely faster ways to get the SIZE value without a script but I went with it for simplicity.)

For situations wh ere the second image is not perfectly square, I used a script to output the width and height of the main image, and then used a "Lookup" component and a horizontal and vertical gradient to scale the UV space correctly for the secondary image. (I could also have used the Scale component here, but it was a bit easier for me to think about the math in terms of UV space at the time.)

On the plus side, I was able to get the second image to be scaled 1-to-1 relative to the main image, and thankfully no aliasing occurs in the second image. Unfortunately, you still have to manually type the pixel dimensions of the second image into a pair of value sliders to get it to scale correctly. All in all, it's a good solution for the time being. :)

Example filter snippet:
https://dl.dropbox.com/u/38240401/Publ...0118.ffxml
(Note: You may need to right-click the link to download the filter.)

This filter demonstrates a method for scaling an image loaded in a "Color Control" component, to be at a 1-to-1 pixel scale relative to the primary source image's dimensions.

Two different methods are presented in this demo. One method specifically for square images, and another more general method for rectangular images. You can switch between these scaling methods using the slider control labeled "Square, Rect, None".

Controls:

"Square, Rect, None":
Select the scaling method. 1 = Square. 2 = Rectangle, 3 = Show the primary image only.

"Square Image":
This is the image to be used when the Square scaling method is selected.

"Square SizeXY px":
This is the size (in pixels) you want the square image to be scaled to. If you enter the pixel size of the image you loaded into the "Square Image" Color Control, then it will be at 1-to-1 pixel scale with the primary image.

"Rectangle Image":
This is the image to be used when the Rectangle scaling method is selected.

"Rect SizeX px":
This is the horizontal size (in pixels) you want the rectangle image to be scaled to.

"Rect SizeY px":
This is the vertical size (in pixels) you want the rectangle image to be scaled to.

"Blend":
This is the opacity of the "Square Image" or "Rectangle Image", as overlaid on the source image.

(Note that all the "Built in" images in Filter Forge are 600 x 600 pixel squares. To better test out this demo, you may want to try loading some images of your own into the "Square Image" and "Rectangle Image" controls. You'll want to take note of the pixel dimensions of any custom images you load in so that you'll know what to type into the size controls.)

Walkthrough:
Here's a brief overview of the setup I used for square images. It can undoubtedly be optimized, but it works decently for the time being.

1. Create a "Color Control" component and load an image into it.
In the "Image placement" dialogue, set
"Horizontal Anchor" to "Left"
"Vertical Anchor" to "Right"
"Placement" to "Fit"

2. Create a "Map Script" component and make it output the global "SIZE" value on the R, G, and B channels.
e.g.
return SIZE , SIZE , SIZE , 1 ;

3. Create a "Value Control" component and set the value to the pixel size of the second image you loaded into the "Color Control" component earlier. (e.g. If the second image is 1024 x 1024, then enter the value "1024".)

4. Create a "Divide" component.
Link the "Value Control" component to "Dividend".
Link the "Script" component to "Divisor".

5. Create a "Scale" component.
Link the "Color Control" component to "Source"
Link the "Divide" component to "Scale"

6. That should be it. :)

Im New, Any help?

Bubba J Gump — Wed, 09 Jan 2013 13:21:20 -0500

Bubba J Gump, Fri, 04 Jan 2013 03:24:44 -0500
Hey, guys, I just recently joined the site, and I was wondering if someone could tell me how to get started. :o

Skybase, Fri, 04 Jan 2013 06:49:57 -0500
http://www.filterforge.com/more/tutorials/ <--- these will guide you! :D

CorvusCroax, Wed, 09 Jan 2013 13:21:20 -0500
Hi Bubba J Gump

Welcome to the FF community!

The best way to learn is to find some filters you think are cool and start tinkering with them / takinging them apart. Start simple and start putting your own things together.

DrEvil's Belgium Ripple filter Improved

ddaydreams — Sat, 05 Jan 2013 20:24:28 -0500

ddaydreams, Sat, 05 Jan 2013 14:28:39 -0500
I've completed the Frankification process on DrEvil's Belgium Ripple filter. So I no longer need help on this.
Here is my final version, I'll call the extreme version. Not to be confused with my earlier Frankifications of this.
I'll message DrEvil and ask him to post this filter to the Gallery so all can find it. I would post it there but he's the creator of the original.
Enjoy.

Indigo Ray, Sat, 05 Jan 2013 14:41:20 -0500
Cool! Just be sure to put your name as the filter author. ;)

ddaydreams, Sat, 05 Jan 2013 19:19:53 -0500

Quote
Cool! Just be sure to put your name as the filter author.

Ohh I did not notice that author comes up as you because of your contribution.
Sorry. I'll change that and I'll change the Filter name as well it's now
"Ripple Explorer".
Has me as Author now. I'm not sure what the etiquette is here. Credit is given to DrEvil in the description. And I asked him to post it in filter gallery.
I just asked him Today. I'm thinking I should wait a week for his reply and then if I don't hear from him. I'll post "Ripple Explorer" in the gallery myself.

Is this right/wrong way to go about this?
I don't want to upset or ripoff the original Author DrEvil
But I've made significant changes and I want to get this available to people.

ddaydreams, Sat, 05 Jan 2013 20:24:28 -0500
Here is a quote of what I asked DrEvil in a message today.
"After thinking more about it. I've made so many changes that you may not want to have anything to do with it.
I've renamed it Ripple Explorer and I mention DrEvil in the description. So check it out and with your permission I'll post it to the filter gallery with me as the Author.

I'll wait a week to hear back from you before posting.
If you don't want me to post it or need me to change something, let me know.
I'm not sure what the etiquette is here for derivative work."

Filter request

Happy Mistakes — Tue, 01 Jan 2013 10:37:43 -0500

Happy Mistakes, Mon, 31 Dec 2012 15:19:32 -0500
Just ordered my copy and fiddling around... I like to create a filter that is a combination of these two:

http://www.filterforge.com/filters/2305.html

http://www.filterforge.com/filters/8054.html

Basicaly I'd like the glossyness and bevel of the first one to be a mother of pearl kinda color/finish, can someone please help me get started or maybe even create one? Thanks and happy new year!

Betis, Tue, 01 Jan 2013 00:56:58 -0500
Easy! Take the first filter, and put in white as the color input. BAM! 8) :D

Happy Mistakes, Tue, 01 Jan 2013 04:27:20 -0500
Thanks but that does not give me the mother of pearl finish I am looking for.

Indigo Ray, Tue, 01 Jan 2013 10:37:43 -0500
Happy new year to you! :)

It seems it's the lighting settings (the lighting tab) and surface settings (reflectivity, etc.) that make the nice pearl finish in the 2nd filter. Maybe you could replace the circle shape with the image bevel found in the 1st filter?