It seems the Python Imaging library can make a new image by using a stencil image and 2 other images. I used to do terrain texture splatting with Panda3d, with texture combine modes and shaderGenerator. However this will allow me to have shaders on the terrain and ground (shadows), without loosing the texture splatting effect as PIL will create a new image which i will use as a single texture.
My question is: is this a bad idea? I can’t think of any drawbacks.
Obviously, do shadow shaders also effect the casters? Then I should forget about this either way.
Texture resolution/size can be a problem as the limit (at least for me) is at 8k x 8k which is not sufficient for large terrain chunks.
True, but I have had great success using a 2048X2048 texture, and applying a detail map over it.
Thats the limit of PIL or what?
Where you talking about my post?
No, I have not tried PIL for this (though I think its a great idea).
I use 2048X2048 because I find no reason to use a bigger texture, and use up so much memory.
A 2048X2048 with a wrapping detail map looks great.
A four-channel uncompressed 2048x2048 texture takes up 17 MB of texture memory. For 4096x4096, this is 67 MB. For many GPUs out there, that is more than half of the available texture memory. It’s simply not feasible to go bigger than 2048x2048.
My GPU also limits it to 8192, but I can imagine that older GPUs would have lower limits.
If you use mipmapping, then you’d even multiply those numbers by 1.333, so it’d be even worse.
This wasnt a problem for me anyway.
Im just concerened if all types of shader shadows will break the shader generator on the ‘casters’ as well.
Allow me to put the latest question in another way:
Are there any kind of shadow rendering techniques that will not break the shaderGenerator on everything in the scene?