Quote Originally Posted by Darkhog View Post
Problem is that when checking for collision with one sprite (it's pixel-perfect by the way) I can't tell if part of sprite player is colliding with is spike or harmless object, hence two sprites per chunk - one for harmful things and other for harmless. Supposedly I could do like you are saying, but I'm too much of a n00b yet and I'm basically learning Allegro with this project so...
collision sprite could have one color for marking harmful area and second for non-harmful, other colors could be also used if needed for other properties like bumpiness. This could be broken down to using one color channel for harmufulness and other for bumpiness to allow both on same surface
I think this info could be also encoded directly into single sprite image with some work.