How would enabling triple buffering make other cores slower? They would be slower only if they render at more than 60fps (so in the case where you emulate too fast).I thought using triple buffering added some overhead as opposed to "render as many frames as possible"...
The cores i was talking about are not even running fullspeed by the way.
SDL_SWSURFACE is the slowest,
SDL_HWSURFACE is fast but has tearing issues and no VSYNC,
SDL_HWSURFACE | SDL_DOUBLEBUF is fast, gets rid of the tearing issues and provides VSYNC, but the games are limited to 60/X fps (so 60fps, 30fps if it can't reach 60fps, 15fps if it can't reach 30fps etc.)
SDL_HWSURFACE | SDL_TRIPLEBUF is fast, has no tearing issues, has VSYNC, and without this problem.
So the general idea is:
- never use SDL_SWSURFACE,
- never use just SDL_HWSURFACE,
- if the game/emu will always run at 60fps, use double buffering (it has a slightly smaller latency),
- if the FPS of the game/emu will fluctuate, use triple buffering.