So this texture is basically just noise on 3 channels. Here’s the red, for example
So, we can sort of apply the color output in an overall sense to the scene texture, like this
which gives
But you can see, the pink ( red ) is dominant. And we’re already ‘faking’ the lerp input from the 3 black and white channels together, so we might as well fake the whole thing, and have control over how much of each color we have, like this
You see, I am just taking the b&w source from each channel. I have to multiply it by various amounts, because the colors have different strengths in the original texture, then lerp it. The only advantage I get from using the different channels is the fact that the noise is in different positions. Now we get
It’s a bit much, maybe. We can reduce it like this
Finally, I ended up with this sort of MI configuration
I dropped the blue because the noise for that channel was almost in the same position as the red.
