The Disquisitions of Betis

I thought it would just be jolly to start one of these...

Betis, stop feeling guilty and use tigerAspect's, unless you have more than just the shadow in mind. It's a free filter library; use it to your advantage. I could even give you a shadow arc filter since I already made my own.

Also, what's up with the planet names?

Dude, use it.
Credit it in the filter description if you really want to.
The only reason it's not a snippet is because I felt like slapping a moon-like texture on there.
It's MEANT to be used by other people.

You've made it way more complex than it has to be. I don't like tooting my own horn, but my method is precise, and extremely easy to tweak for whatever purpose you need.

I would be eager to hear some comments on my new "Gradual Blur" filter. It uses a Black and White map (like a bumpmap) where darker is more blurred.

No comments to the above? Oh well, I'm now getting into depth blurs (not focal blurs) So far I'm really close, but have a problem with switching focus distance. Here's the filter if you guys want a look at it. It's similar to Gradual Blur, but is... not, obviously.

Here are some Selections for Depth Maps and Images to test it on:

http://i400.photobucket.com/albums/pp..._depth.jpg
http://i400.photobucket.com/albums/pp...er_vis.jpg
http://i400.photobucket.com/albums/pp...square.jpg
http://i400.photobucket.com/albums/pp...blocks.jpg
http://i400.photobucket.com/albums/pp...mbered.jpg



Depth yNot Focals Blur.ffxml

Yeah, DOF.

And the Checkerboard is just for the test, not for realism.

And thanks for the heads up on the pool.

In reality, you have to have a full 3D environment and 3D Map for the light to bend around. With 2D images, you cant blur the foreground and see behind it, like you can in real life. If you have already downloaded my filter and the two images:

http://i400.photobucket.com/albums/pp..._depth.jpg
http://i400.photobucket.com/albums/pp...er_vis.jpg

You can see that it works fairly well, except when you get near focusing on infinity.

Hmm? Use the greyscale image as the selection, It's an actual depth map, unless you're already doing that. The greyscale works similarly to my gradual blur map. Don't you see the chain of blurs that are blended on a layer-like system? Like depth?

I'm not sure how useful this would be. Cameras already blur things that are out of focus, and you can already choose what distance the camera focuses on. Plus, the way lenses cope with distance, far away things (like the mountains in your example), are grouped together and so are all either out of focus or in focus.

Also, as Craig said, its pretty much impossible to create a depth map from a plain old photo anyway. Maybe you can with 3d environment generated in a computer program. And remember that if using a 3d environment, use the same rules as real lenses (the closer you are, the greater the chance two objects/points will be focused differently).

Based on the above, I have three questions:
1. How can you make a depth map in a 3d program?
2. Are there cameras that can focus multiple times for one image to get everything into focus?
3. Are there cameras that record depth in a file?

indigo,

#1. no idea, really.

#2. you dont need multiple shots or to focus multiple times. your depth of field is set by your f-stop. f-stop is just a fancy way of saying effective diameter of a lens (determined by the aperture setting) compared to the length of focusing (usually from the center of the furthest out lens to the film plane). the ratio of those two things determines how much and how far away is in focus. at least that's my understanding of it

#3. cameras dont record depth, per se. they record light. in film cameras light burns a chemical emulsion on a film plane to create the image. in digital, light 'burns' a sensor to record light data. actual depth perception comes from what we consider we see and from viewing with essentially two points, our eyes. it's said that folks with only one eye have very little depth perception. there are also certain factors that contribute to depth perception, like blurring, black and white/light and dark and even color. a blurred object usually says i'm further away when compared to a sharp object. black usually fades into the background where white tends to stand in the foreground. and different colors can be seen to be further and closer, depending on what they're compared to.

also, digital cameras all shoot in RAW. however, most low end cameras convert to something like .jpg before presenting them to you as a finished image. in better digitals, you can save the RAW and import it into photoshop or paint shop pro and work with the exact data you shot, rather than a transposition like .jpg. you always lose a little bit of data when the camera converts to .jpg or another format. and the reason i mention that at all, is your question about depth of field. RAW will show this better than .jpg just because you get more accurate data.

#2-Craig, assuming that a camera can tell how far away something is, would there be a way for the camera can focus on near objects, record an image for that selection, focus on far objects, record an image again, and blend them together to make both far and near objects in focus? FF can work by this same concept using thresholds and blends (take Betis' gradual blur filter or this depth one here). Sure, this method is not 100% accurate, but its enough to fool the eyes.

#3-Correct me if I'm wrong, but don't cameras have to measure distance in order to focus? Unlike people with two eyes, cameras only have one lens. Being a person with only one fully-functional eye, I find that there are mental tricks like comparing scaling and shadows to create depth perception. However, cameras aren't very intelligent, so you can't rely on tricks alone. Don't cameras use infrared sensors to measure distance more accurately? And wouldn't it be possible for the camera to save this distance info it measured? Or is this included in RAW format? I looked up RAW on the web a bit and found nothing about distance.

indigo,

what i was trying to say in that first post about cameras and depth of field is, you dont need two different focuses and then a blend. any half way decent camera can already do that with just one shot. it's done by the lenses. if you look at even small, inexpensive digital camera, you shld have both a close up range (very narrow depth of field) and a longer range focus (infinite depth of field). if you want everything in focus, just set the camera to 'landscape' and that will set the lenses and aperture to keep everything from about 18 inches to infinity in focus. if you want to take a picture of a flower, though, and fairly close, set the camera to 'closeup' and focus on the flower. the background will automatically be out of focus and blurry. that's all a function of the lenses and aperture of the camera. you dont need more than one shot.

cameras dont 'measure' anything really. they are simply receptacles for the recording of light. focusing is done by the lens or lenses (some have multiple pieces of glass within). light comes into the tube of the lens and is focused by the lens, once or even several times till it reaches the destination, the film plane or sensor, where it is then recorded. so, all focusing is done by the lens. it is simply curving the light in such a manner as to make something visible and coherent on the film/sensor. the depth of field is captured by how the lenses are focused and by the aperture opening. those two things determine how the final image is focused and recorded.

camera RAW is basically the data straight out of the sensor. it has no alterations and no header to the file. the header is what tells a viewing program how to arrange the data to be viewed. so, a .jpg file is a file that has a header and the data. the header data may contain things like image size, how many pixels by how many pixels and so on.

and yes, there are mental tricks that can compensate for depth of field. i had an accident years ago that left a patch over one eye, temporarily. i still had to drive a car and so on and because i already knew distances and relative sizes of things and so on, i didnt have too much trouble compensating. the mind can almost completely substitute there. but, it also shows how optical illusions work, to some extent. the mind is so used to certain relationships and agreements, that you can trick it with optical illusions.

and no, cameras do not use infrared to measure distances, at least none that i know of. RAW is basically just like film. the data is just the data. it's you and me that interpret that data to show depth or consider that something has depth. in actual fact, that data is strictly 2d, (at least until we start making true 3d cameras ) and therefore there is NO depth measurement or special recording. it's just raw data coming in from the lenses and captured by film or sensors. that's it.

think of it this way, maybe; you've got all these colored photons flying around. they enter a lens and get arranged a certain way and hit the sensor or film and are recorded. there is no depth to the photon, per se. there is no method of truly capturing depth other than those photons flying around and they dont give a damn about depth, nor does your sensor. the sensor just records those photons hitting the sensor and that's it. that's all that is recorded

now, you might could build a more intelligent camera. i'm not saying that your ideas along these lines wouldnt make for a better camera, but currently, a camera is only a photon catcher and that's it

So, Indigo, when will I get that picture?

Anyway, I just submitted a filter about 1 minute ago called Paper People, it's pretty awesome and all, but that's not the point, the point is is that it's totally unprepared for presentation in Recent Filters. One of the sliders had all remapping backwards, there were Cul-de-sac node streets, and other trivial imperfections. I think you've all had this happen before, so you know what it's like. But finally, to get my point across for posting this whole message is that wait for an update before going, "Betis, look at all this messy work you just tossed in the smoldering heap of a filter library he have here..."

I'm also just rambling because I feel like it though, so I'm actually not to stressed about anything... there I go again, well I guess I rambled (and am rambling) about rambling, so I'll just say good day, I'm off on my way.

that's a pretty thick atmosphere. cut it back a bit, maybe?

and the nebulae are competing too strongly for attention. tone em down a bit. attention shld either go mostly to the planet or mostly to the nebulae, but doing both strongly doesnt quite work for me.

I have not forgotten about Galaxy Explorer, it's just hard to work on with such a high expectation.

But I have a new filter that I'm also working on. Beams of Electromagnetic Radiation! I made this before seeing ThreeDee's Light Rays filter, and was not inspired by it (although it is inspirational). I have compared the two and deemed them very different, with dissimilar results as well. My filter also instead uses the selection as a depth map, to see where the object is placed in the scene, as well as which objects occlude the light. The guts may also seem too much like Uber's Zoom Blur filter, but I came up with it on my own as well. Anyway, here are some samples.

This is just a generic cool one where the light is behind the grate. Here you get to see how the Glow Effect control works.


I made a custom depth-map using the picture of Half Dome, and in this picture, the light is a giant glowing ball behind the trees, in front of Half Dome, and near on the same plane as that side cliff. This is a good representation of the 3D capacity. This also shows how the Glow Effect control has an effect.


And here you see mostly how the filter works, although it still looks cool.

Why, yes of course, with another sample image

very nice on the beam/glow filter!

Quote
Betis wrote: Beams of Electromagnetic Radiation!

Those rock!

lookin good, but why post on photobucket when you can post as good here? i mean, maybe you can get larger file sizes and such on photobucket, but they've started to use popups and your image size there is fairly small. you could have posted a larger image size here at 250k and made it all easier.

Ok, i'm just going to say it. You need to plug that glow effect into a nebula filter ASAP, back-lit nebulas are like some kind of holy grail.

I whipped this up in a few minutes, thanks to bombers.

Reminds me of Dilla's Cheese, but this is hopefully more recognizable as a sponge.

It's weird, that's at maximum darkness and opacity. I'm using the Soft Light blend mode, but I think I can do some more passes.