Talking with John Brooks on how he did Rastan; he compiled his sprites into code that instead of a simple rectangular blit, he had code that would put each pixel in the correct place, skipping over any transparent pixels.
I think this might be mostly beneficial for large sprites because there will be areas where you're writing entire bytes rather than masking off the edges, so I'm not clear that it would be a benefit in the general case for small sprites. But it's something to think about if I ever get a sprite editor.
Talking with John Brooks on how he did Rastan; he compiled his sprites into code that instead of a simple rectangular blit, he had code that would put each pixel in the correct place, skipping over any transparent pixels.
I think this might be mostly beneficial for large sprites because there will be areas where you're writing entire bytes rather than masking off the edges, so I'm not clear that it would be a benefit in the general case for small sprites. But it's something to think about if I ever get a sprite editor.
See http://www.brutaldeluxe.fr/products/crossdevtools/mrspritetech/index.html