|
Quasimondo
|
I think it might be a nice idea to start some challenges with the goal to find out how to construct certain effects - ideally those that on first look seem to be impossible to build with the current toolset. Feel free to add new topics with other challenges - but please don't add a different challenge to this thread.
So my first challenge to you is to construct a circular gradient - not a radial one which we already have. Ideally it is black at 0 degrees, is 50% grey at 90 and 270 and white at 180 - but of course the perfect circular gradient that I am looking for can display any kind of color combination. |
|||||
| Posted: June 4, 2006 6:49 pm | ||||||
|
uberzev
|
Something like this?
I'll see what I can do. |
|||||
| Posted: June 5, 2006 3:47 am | ||||||
|
Quasimondo
|
Yes. Good job! Are you using refraction?
|
|||||
| Posted: June 5, 2006 4:56 am | ||||||
|
uberzev
|
I just wanted to make sure I knew what you wanted before trying to make it in FF. |
|||||
| Posted: June 5, 2006 5:00 am | ||||||
|
Quasimondo
|
Ah okay - but it's actually a good idea to use a photoshop prepared gradient as a reference - if you load this as an image and layer the calculated gradient with difference blend mode onto it you can see how good the implementation is.
|
|||||
| Posted: June 5, 2006 5:13 am | ||||||
|
uberzev
|
Here's what I've got so far. Not great but maybe something you can work from...
http://uberzev.googlepages.com/RadialGradient.ffxml |
|||||
| Posted: June 14, 2006 2:09 am | ||||||
|
Quasimondo
|
Oh - that's not so bad for a start. Good work!
|
|||||
| Posted: June 14, 2006 5:34 am | ||||||
|
CFandM
|
Nice so far. I was thinking of how to create this myself. I had a easy start then it turned out it was not going to work the way I wanted. I will give this a go. I will post results if I have any.
Stupid things happen to computers for stupid reasons at stupid times! |
|||||
| Posted: June 14, 2006 8:12 am | ||||||
|
uberzev
|
I think I nailed it...
http://uberzev.googlepages.com/CircularGradient.ffxml
Thanks to byRo for his brilliant 'TRUE Circles' from which this is based on. |
|||||
| Posted: June 14, 2006 6:05 pm | ||||||
|
uberzev
|
For comparison here's Photoshop's version using the angle gradient...
|
|||||
| Posted: June 14, 2006 6:30 pm | ||||||
|
uberzev
|
Here's a better version that use my tiles-to-circles method...
http://uberzev.googlepages.com/CircularGradient3.ffxml
photoshop mockup... |
|||||
| Posted: June 14, 2006 7:07 pm | ||||||
|
uberzev
|
Don't mind me, I'm just having a bit too much fun...
|
|||||
| Posted: June 14, 2006 9:23 pm | ||||||
|
uberzev
|
Here it is, the holy grail...
http://uberzev.googlepages.com/CircularGradient4.ffxml
Simply apply any elevation gradient you like and presto... |
|||||
| Posted: June 14, 2006 9:43 pm | ||||||
|
Quasimondo
|
Whoohoo - that is brilliant!
|
|||||
| Posted: June 15, 2006 6:46 am | ||||||
|
CFandM
|
Can I say DARN you beat me to it. I will also add HURRAY and a big congrats on working this out. Very clever indeed.
Stupid things happen to computers for stupid reasons at stupid times! |
|||||
| Posted: June 15, 2006 9:11 am | ||||||
|
Vladimir Golovin
Administrator |
Uberzev, this is brilliant. Looks like it's a proper circular gradient!
|
|||||
| Posted: June 15, 2006 9:51 am | ||||||
|
Kraellin
|
nicely done!
ok, for us non-technical types, what is the difference between a ciruclar gradient and a radial gradient and the also mentioned, angle gradient? craig If wishes were horses... there'd be a whole lot of horse crap to clean up!
Craig |
|||||
| Posted: June 15, 2006 2:43 pm | ||||||
|
Lucato
|
Congratulations GeniUberzev! You're a genius!
|
|||||
| Posted: June 15, 2006 2:59 pm | ||||||
|
byRo
|
Great work, uberzev!!!!
I was a bit stuck on this one, and ended up "borrowing" a few of your ideas - but I believe that mine is still quite different enough to be interesting. Is there a second prize? Rô Circular Gradient byRo.ffxml _________________________________
My favourite question is "Why?". My second favourite is "Why not?" |
|||||
| Posted: June 15, 2006 3:27 pm | ||||||
|
byRo
|
||||||
| Posted: June 15, 2006 3:29 pm | ||||||
|
Quasimondo
|
That's no big deal with his method - just exchange the spectrum with a black to white gradient and then change its offset to rotate it. |
|||||
| Posted: June 15, 2006 3:50 pm | ||||||
| Ken |
Well Done Uberzev and Ro.
How about a spiral as the next challenge? Ken. |
|||||
| Posted: June 15, 2006 4:25 pm | ||||||
|
uberzev
|
It pains me to admit it but my method is just not 100% accurate. Note the difference between the 'ideal' version and my attempt...
Here's 'ideal' on its own...
Anyway here's another variation of the effect that uses byRo's circle technique. I can't seem to get it to work just right, but its simpler than my tiles-to-circles method. Maybe someone here can figure it out... http://uberzev.googlepages.com/CircularGradient5.ffxml |
|||||
| Posted: June 15, 2006 6:09 pm | ||||||
|
uberzev
|
That's easy.... http://uberzev.googlepages.com/SpiralGradient.ffxml My method isn't very efficient though. |
|||||
| Posted: June 15, 2006 6:23 pm | ||||||
|
uberzev
|
Some interesting distortions are possible as well...
http://uberzev.googlepages.com/CircleDistortion.ffxml As a perfectionist I hope to see the underlying accuracy issues be resolved before people start using these techniques in their 'official' filters. |
|||||
| Posted: June 15, 2006 6:33 pm | ||||||
|
uberzev
|
Anyway, I initially gave the square method a try but found it didn't have any advantages over the circular variation.
RADIAL...
ANGLE...
I suggest we use these more descriptive names instead of the ambiguity of 'circular'. |
|||||
| Posted: June 15, 2006 6:58 pm | ||||||
| Ken |
Uberzev.
Another great Job and a nice Flower. But I was thinking more of a Spiral Ken. |
|||||
| Posted: June 15, 2006 8:17 pm | ||||||
|
Kraellin
|
thank you, uberzev.
craig If wishes were horses... there'd be a whole lot of horse crap to clean up!
Craig |
|||||
| Posted: June 15, 2006 11:46 pm | ||||||
|
Quasimondo
|
Yeah great! That was my original intent of this challenge - to use the gradient as a base for circular distortions. |
|||||
| Posted: June 16, 2006 4:26 am | ||||||
|
uberzev
|
|
|||||
| Posted: June 16, 2006 5:30 am | ||||||
|
byRo
|
Like Polar Coordinates Rô _________________________________
My favourite question is "Why?". My second favourite is "Why not?" |
|||||
| Posted: June 16, 2006 10:55 pm | ||||||
|
uberzev
|
The 'official' version is now complete...
http://www.filterforge.com/filters/671.html Take a peak 'under the hood' if you like, it's much less complicated than my previous versions. |
|||||
| Posted: June 18, 2006 8:03 pm | ||||||
|
uberzev
|
Version 1.5 is done...
http://uberzev.googlepages.com/Circul...t1.5.ffxml Quality has been greatly improved. Photoshop Ideal...
CAG 1.5...
Difference... |
|||||
| Posted: June 20, 2006 7:35 pm | ||||||
|
Quasimondo
|
Excellent job! Thank you! Yes the improvement is clearly visible. I've already updated my filters that make use of the gradient with your new version.
|
|||||
| Posted: June 21, 2006 3:56 am | ||||||
|
uberzev
|
http://uberzev.googlepages.com/Circul...t1.6.ffxml
Minor update, nothing too exciting. Just simplified things a bit and made sure the 30° angles were accurate. |
|||||
| Posted: June 24, 2006 4:40 am | ||||||
Filter Forge has a thriving, vibrant, knowledgeable user community. Feel free to join us and have fun!
33,739 Registered Users
+5 new in 7 days!
153,586 Posts
+8 new in 7 days!
15,356 Topics
+6 new in 30 days!
14 unregistered users.