What is more expensive at run time: one large texture or several smaller textures? e.g. would a 4096x4096 be favored over sixteen 1024x1024 textures, since the totals of each are the same?
What does it mean to “bake” a texture, and is that somehow beneficial?
Do colors with a lower numeric value (like solid black and the pure red colors) or grayscale colors necessarily render faster than greens, blues and light colors?
If I have a mesh that uses, say, a 2048x2048 texture and it’s UV map pretty much consumes that entire area, and then I have another, smaller mesh that CAN use that same texture but needs only a very small part of it, would it be better for them to both use the same texture/material or should I make a smaller texture map and separate material for the smaller mesh?
Can I use different size maps for color, normal, occlusion, specularity, etc. for a material or do they all have to be the same size. e.g. if my color map is 512x256 do the other maps also have to be 512x256? And if they can be smaller/larger, do their dimensions have to be the same scale (512x256 and 1024x512 or 128x64) or can one be, say, 512x256 and another be 256x256?
If textures are concerned, using 16 1024x1024 textures will be slower than using 1 4096x4096. That is situationally true when textures are used in one material on a single mesh, but is subjected to changes under certain factors.
Wide term, generally used to describe process of collecting data and storing it in a texture.
No.
Typically, that can be avoided, and should be done only in desperate attempts to save last bits of texture memory.
As long as dimensions are power of two, any combination will work.