I’d like it if the TextureMap input had it’s own input, a vector, that would allow you to define a sample at some place OTHER than the current mapping vector.
A scalar for time would also be cool, but not essential to what I need currently.
Why would I want this? So we can sample derivatives and such from a map. Make blurs, get average colors in a region, other map sampling stuff.
You may argue that if it has inputs, it can’t be called an Input itself. Ok, so maybe it’s more like a Function, but basically I want to sample something, per-particle, on something that isn’t intrinsically a per-particle value, any more than DNoise is.
That’s an idea we plan to tackle, since it would be awesome to evaluate Texmaps without having to build a whole bunch of support KCMs just to use a procedural texture coordinate.
Evaluating Max maps at different times is probably not easy or a good idea. I’d love to have the time derivative also, but the entire map tree needs to get updated when you want to evaluate at a different time so this would have to happen for every particle. I can think of several maps offhand that would cause my head to explode if I tried to do this.
For my current task, doing central differences on a texture’s mono gradient to generate normals, I ended up just using Warp Texture. But yeah, in cases where I needed to modulate the lookup on a per-particle basis (not per texture) I would be stuck without a vector input to the In:Map.