Correct texture map usage

I usually perform better when I understand how something truly works. I have been confused lately about how texture maps are supposed to be used. :spin:

Ok, so this is what I have gathered:

The amount of memory in a graphics card is the amount (in MB) of texturing that you can have displayed in the game.

Keeping your textures in powers of two isn’t required, but decreases processing time required to display the texture.

A 128MB graphics card does not have a limit of a 128x128 texture.

Is this correct so far? I have three other questions:

Will sections of one 256x256 texture multiple times throughout different models be more efficient than separate 32x32 or 64x64 textures?

Will a graphics card need to be capable of multitexturing if your environment model uses two or three different textures?

If you are developing for a certain grade of graphics card, how do you decide on the correct size of your textures?

Sorry about the long post, these things have just been taxing my mind for a while.:smiley:

Someplace Somewhere, maybe in the wiki that is explained. But, alas i cannot remember where…

Everytime you use a different texture it has to be loaded… Most will try to fit the most used graphics on one picture.

Because it is faster.

If your using a machine made in the last 5 years I wouldn’t worry too much. textures are better if they are in the 32, 64, 128, 256, 512 order but I tend to use a smaller one if I can get away with it and a larger one where I can’t. And then test the game and see how its running… 256Mb graphics cards are cheap and you can have a lot of textures in there.