Is there any difference in terms of rendering overhead, if we simply use a color (e.g., green) or load a texture image file (e.g., green.png file) for a 3D object?
Shouldn’t the OpenGL ES finally create a texture even for a colored object?
I am using Android API v8 along with the emulator and the goal is an actual Android phone.
Why would it finally create a texture for a colored object?
I can’t speak to every conceivable implementation, but it seems to me that the uniform would be a much better solution. The uniform value is likely cached very locally in the datapath with quick access, versus a texture which has to interpolate the texture coordinates, retrieve the texture from vram, and sample it. I’m not sure how that could be as fast as just reading a couple floats from the uniform.