Hi @Deevad, I’ve compared the results of the new spectral mix and the new spectral mix seems to keep the hue more intact when mixed with white and the red looks just a purer red without leaning towards magenta
Hello, this looks very interesting! Is it possible to find a test build of Krita with your work, but for windows users, or is it too early to think about it? If “not”, no worries or distractions, I just wanted to be curious.
Hi all, this is the windows package, the blur brush may not work properly at the moment, but the normal smudge brush should be sufficient, I think the best way to keep the current blending effect is probably to use XSIMD for code optimisation, which may require the help of the official developers to do this; the pixel brush engine brushes need to select the spectral blending mode, the smudge brush needs to be turned on in the brush settings to use the new version of the smudge brush engine, and the spectral blending needs to be under the linear profile to blend the correct colours: Krita
WOW, what an amazing addition to Krita you are making! THanks a lot for this!
Quick questions as I’m not the most knowledgeable in the coding of stuff, I was wondering if you absolutely have to use a 16 bit profile for the blending to work this way? I think It would be great to have it implemented for 8 bit as well, but I don’t want to push more work on you, it’s already sooo great that you are doing this!
Another thing, I was able to test out your settings thanks for the build you just posted and I was wondering why the colors are not reacting the way your first image shows:
From my point of view, this version is the closest to what traditional pigment would look like so I was wondering if there was a way to keep it like that? O.o
Thanks again for everything, you are really doing something great!
Thank you for your testing. The first image shows the pigment mixing effect in Artstudio Pro, which I feel is similar to Mixbox. Mixbox uses spectral data of pigments, while the version I built uses spectral data of RGB, which is the effect in the middle of the image. Therefore, it does not shift the color phase when mixed with white. Spectral mixing can work in 8 bits, but the most important thing now is to optimize performance, because there is a performance loss between converting from linear to nonlinear, I have basically done what I can because I am not a professional developer, and subsequent optimizations may require the official developers of krita to complete
However, the principle of Mixbox and SpectralMix is the same. As long as there is corresponding data, corresponding effects can be mixed. In the future, people can even choose the desired mixing effect from various mixtures, such as pigment mixing, spectral mixing, or other artificially made mixtures
@Wavenwater : the first version (top in the fig.) is more dramatic, sure, and probably appealing to the eye at a first glance, but also a caricature. Look closer how the orange has a value that’s … ‘glowing’. It’s even more dramatic on a color-calibrated display. No pigment I know (gouache/acrylic/oil) reacts this way. And we are speaking of pigment here flowing within a neutral mixture, not about how some random inks/watercolor or wash of oil that might have sometimes large hue variations with chemical reactions depending on the acidity of the paper/canvas, and chemical reaction due to mixing of water/spirits/petrol/bleach/oils/etc…
Thanks for the inspiration, I used KM finder to simulate the new spectral curve which contains K and S. If I find a way to convert K and S to spectra, I might be able to use them