Hi Egon,
I would like to make a depth of field effect by using post proccessing. It requires rendering the object z buffer in 1st render pass, then combine with main scene with a custom dof shader set to the framebuffer for post processing. To render the z buffer, we need to render the scene on a texture match with the screen width and height. I'm thinking of using NPOTTexture, but you mentioned in doc saying NPOTTexture is preferred to be in power of 2 as it's not supported in some devices. How should we do in this case? I worry if we create a power of 2 NPOTTexture and it does not match to screen width and height, we might have some texture size ratio problem. Please advise, thanks.