Andrew B.
![]() |
Has anyone here figured out how to add sliders to a filter to control L, a, and b channels? The extract modules only change modes and invert.
BTW, I tried adding modules for each channel, making inverted duplicates of each, and adding a blend (with int slider) between each pair. That didn't work. |
|
Posted: December 26, 2011 10:42 am | ||
Morgantao
![]() |
Take a look at Indigo Ray's CMYK snippet. Maybe that would help (unless that what you mean by "adding modules for each channel, making inverted duplicates of each, and adding a blend (with int slider) between each pair", in which case forget about it).
![]() |
|
Posted: December 26, 2011 3:47 pm | ||
Andrew B.
![]() |
I looked at Ray's snippet. He is operating at a much higher level, that I can't even begin to comprehend. What I was doing is throwing rocks at the problem to see if I got lucky.
Anyway, thanks for pointing me to the CMYK snippet. I might be able to use that. |
|
Posted: December 26, 2011 5:02 pm | ||
Morgantao
![]() |
As far as I know it's much easier to convert RGB to CMYK and vice versa, since both are physical representation of light (They are actually the opposite of eachother).
LAB color however, is a mathematical representation of color. It only exists in computers. Therefore the conversion is purely mathematical. If you want to get geekier, read on. Otherwise run away ![]() Color is actually mixture of different light wavelengths. White light (Like we get from the sun) is a mix off all the colors of the rainbow - Red, orange, yellow, green, blue and violet (And a whole lot of other wavelength that we can't see, like Infra Red and Ultra Violet). You can also create white light by just mixing Red, Green and Blue lights, which are additive colors (Meaning they add light to eachother). That's how TV's and Screens work. They EMIT light. CMYK colors however, are substractive colors. That means they SWALLOW parts of the white light, and the colors we see are the "leftovers" that were able to bounce off from the surface to our eyes. White means no light was absorbed, Black means no light was able to escape, and all the light was absorbed. The CMYK colors (Cyan, Magenta, Yellow and blacK) are used in print, in which case you take a white paper that doesn't absorb light, and make it absorb differen wavelengths by printing on it. Each color space (RGB, CMYK, LAB) has different color properties. CMYK has the least amount of colors. RGB can show colors on screen that won't print on CMYK, because it's a wider color space. LAB color space has much more color than RGB and CMYK. In fact, so much color, that neither the human eye, nor screens and printers can interpret all of it. That means some colors in the LAB space are "theoretical" colors. They can't be recreated in the physical world. |
|
Posted: December 27, 2011 11:56 am | ||
Indigo Ray
![]() |
Well, the reason I made the CMYK snippet was because there was no "extract/assemble CMYK" components. But there already exists similar components for Lab color. If you want to change the amount of "a" for example, just extract all 3 channels, do whatever you want to the a-extraction (a tone-curve, blend with something else, etc.), then assemble the new "a" along with the old "L" and the "b". Or you can change all 3 channels. You can even plug in completely unrelated gray-scale maps into the assembler.
Morgan, I like the food analogy for the CMYK. ![]() |
|
Posted: December 27, 2011 1:48 pm | ||
Morgantao
![]() |
They must beairing too many coocking shows on TV....
![]() |
|
Posted: December 27, 2011 3:14 pm | ||
Andrew B.
![]() |
Thanks for the additional comments. Since I last posted I figured out the logical way to set it up. The a and b channels are represented with grays. In the a channel, lighter equals red, darker equals green. In b, lighter is yellow, darker is blue. And the L is Lightness. So, in theory, I should be able to control this with brightness sliders, assuming brightness functions a straight line way. Problem is, when I hook everything up, with all sliders set to 50 (out of 100), the image changes. It should be the same as the input. Also, when I move the sliders the colors change according to plan, but the a and b don't seem to be respecting the L channel as I think they should.
I should/could run more tests, but for now I'm just going to drop this. I bit off too much anyway. I'll worry about this in a future version. |
|
Posted: December 27, 2011 3:43 pm |
Filter Forge has a thriving, vibrant, knowledgeable user community. Feel free to join us and have fun!
33,711 Registered Users
+18 new in 30 days!
153,533 Posts
+38 new in 30 days!
15,348 Topics
+73 new in year!
35 unregistered users.