In order to select d3d9, compatibility, or xp with no,
No.5 video RGB32 format the memory data real-time, according to the superposition of every road is a 25 FPS,
Unripe d3d9, access to 2 d images show, the first is the image texture way to load, and then to texture mapping as a quadrilateral can; The second is to use 2 d point, want to consult a familiar friend, which way is more efficient, thank you, or any other way, trouble you tell me, thank humbly.

CodePudding user response:

Don't use the elves... Which in some graphics card there will be a strange question, but efficiency is not going well
Decisive quadrilateral start to yourself, and simple, want to group all can

CodePudding user response:

1, the use of orthogonal
Stick your rest 2, quadrilateral another texture
3, need image overlay can be to write a simple shader