Hello,
First of all, I did some shallow prior search, but none had satisfied me fully.
If there is a thread already answering my questions, please redirect me to it.
I am trying to make my own implementation of a dimetric world.
I plan to represent each cell with three sprites (top, front, right side), so I need three textures for each type of sprite.
(As it is a dimetric projection, I could flip the front or right side to represent the other, but I want to differentiate them).
I ended up with a 60x30 image for the top side and two 30x60 image for the other two sides.
Here comes my first question,
Should I enlarge the file to a power of two (i.e. 64x32 and 32x64) manually or just let the computer do what it needs to?
And to top that,
Should I make each texture 64x64 on its own?
I might be misunderstanding but I think I glimpsed over a notion that texture with same width and height is better for graphics card to work with.
But then again, the computer could do it for me even if this is true.
Right now I'm packing all three of them in a single 60x75.
If I load this file as a single texture and divide it up when creating sprites with texrect, will it be padded to 64x128 (or 128x128 if worse)?
Basically, when is the padding done? At the level of texture object or the sprite object?
There are a lot more questions I still have in mind, but they are mostly on the similar topic with "else" cases.
So for now, I would appreciate some thoughtful answers or other suggestions on my implementations.
Thank you.