When taking a pair of images for a stereo photograph you get ‘keystone’ distortion if the cameras are ‘toed-in’ .
This is easily corrected by rendering each image to a quad and rotating the quad into the screen and in the plane of the screen.
The corrected image can be saved by copying from the frame-buffer but is limited to the window size (and ultimately the texture size).
How can I manipulate and save a larger image ?
If I draw on a mosaic of quads how would you seamlessly create the single, final image ?
No problem !
The answer was quickly found ‘surfing the Net’ .