Hi,
i am working with Opengl ES (opengl for embedded systems) and there is
a problem with
bitmaps. I cannot use directly bitmaps in order to use one of them as
the background image.
There are two way to do this:
1] use two triangles that are perpendicular to the camera and cover
the screen, disable lighting and perspective correction, use
orthonormal projection and use the bitmap you want as a texture,
render the two triangles, render the scene and swap the buffers....
2] do all the rendering on a surface, copy the surface on a bitmap,
manually copy every non black pixel of that bitmap on the background
bitmap and finally visualize the background bitmap
Consider the fact that the background bitmap change at every frame and
that the frame per seconds cannot be > 15. The system does not have
hardware acceleration (is a phone). I use the method 2], wich is slow
(14/15 fps) but not so slow as the method 1] (7/8 fps) (but i am not
sure that i have coded the method 1] correctly).
Furthermore, every frame i have to scan the background bitmap pixel by
pixel for other types of calculations.
The question is: can the method 1] be faster than the method 2] and
how can i implement the method 1]? I would like to be sure that the
correspondence is 1:1: a pixel of the background bitmap (the texture)
is mapped to a pixel of the backbuffer.
Thank you,
Luca


|