Sorry im quite new to compositing, so maybe it’s a stupid question. Sorry if I wasted your time :-(.
When I want to render an image with a predetermined color, I use emission modifier with a high strenth value, that alows me to match external colour scheme with correct rgb values. Combined with freestyle it looks like the emission wraps around the line, and remove all alpha values. I thought the new freestyle pass would fix this, but the problem still persist.
When I try compositing in an external program it works correctly and lines look the same. It helps if you increase resolution, because the lines are composed of more pixels, but I think that shouldn’t be the solution to my problems.
Is there a different options for reaching exact rgb values? I’m quite new with compositor, so maybe I use wrong node setup, there is no difference if I use alpha ower or mix node.
This would realy help me in my blender workflow. Thank you all, i love you an the blender community.
After searching around and playing with the holdout shader, I figured out a half solution.
So i used holdout shader for the emission material and the composit the right rgb value in compositor. I noticed that my white background is not white but is a light gray color. I then used the color picker and clicked the brightest part of my image and noticed that the rgb values were above 1, they were 20. Then i changed the alpha over layer to have rgb values 20, however the freestyle lines disappeared,I figure because program treats higher values (over 1), so they overwright alpha channel respectively on the strenght over value 1.
After looking into this promblem i found a reason on a reddit forum, It looks like the problem is that the filmic colour management has a wider range to represent light wich increases dinamic range and therefore makes scenes more photorealistic. When I changed the setings to standard, the emission values respect the 0 - 1 range, and the problem is avoided.
Thankfully I don’t need filmic colour so this solved my problems, however the problem is still a valid one, I am interested if there is a workaround, maybe with some math nodes to decrease the initial light values.
