Im using surface objects to store images in memory, and and a hardware surface
for the application window (The application plays a sequence of images back
from disk). If i make the images hardware surfaces also, will SDL fall back
on making them software surfaces when a machine runs out of vram, or will
they just fail to allocate?
Is it even worth making the images hardware surfaces as i have to decode them
from disk into memory anyway, which for the formats im working with tends to
require about a dozen or so operations per pixel, and the library
documentation says pixel operations are often quicker on software surfaces
Would I benefit from using a YUV overlay (and having to do RGB->YUV colour
space conversion) than just using a plain old SDL_HWSURFACE?
^^ | Tom Badran
(oo)____ | Imperial College
(__)\ )/\ | Department of Computing
||----w | | -----------------------
|| || | Using Debian SID