I'm mixing textures in RGBA8888 format with textures in RGB565 format and I wanted to verify with you guys that I'm understanding what's going on correctly. Here is my scenario:
1) I set the default texture format to RGBA8888 in the AppDelegate using the following
[CCTexture2D setDefaultAlphaPixelFormat:kCCTexture2DPixelFormat_RGBA8888];
2) I have some 1024x1024 texture file stored in the PVR file format using RGB565 pixel format. Let's call this file 'MyTexture565.pvr'.
3) I have some other 1024x1024 texture files stored also in the PVR file format, but using the RGBA8888 pixel format. Let's call this file 'MyTexture8888.pvr'.
4) I load the two textures with the following code:
CCTexture2D* tex565 = [[CCTextureCache sharedTextureCache] addImage:@"MyTexture565.pvr"];
CCTexture2D* tex8888 = [[CCTextureCache sharedTextureCache] addImage:@"MyTexture8888.pvr"];Am I correct in the following assumptions?
1) even though I explicitly set the default texture format to RGBA8888, each of the PVR textures gets loaded using its native pixel format
2) tex565 sucks up 1024 * 1024 * 16 bits of device memory
3) tex8888 sucks up 1024 * 1024 * 32 bits of device memory
When I step through the code, the format_ variable of each of the CCTexture2D objects is set to "kCCTexture2DPixelFormat_Automatic" and I'm not sure whether that means what I think it means. Could someone in the know please verify my conclusions? Thank you!
