Banging out some collision detection. Right now I can tell if the danger point (one pixel X/Y location where hits are checked) is within the 40 x (sprite height) area of an enemy. Next I'll need to write a small 'more precise' function to put on both banks with image data - this will check if the dangerous pixel does or does not collide with real image data.

Ignore the ugly colors below, that was part of my WIP color-decode.