I was playing around with motion detect scripting a bit and found that the pixel step parameter is pretty important, but it's effect is a bit hard to quantify, especially since I don't actually know what my camera's LCD resolution is for most of the time (I've read it somewhere but these things just don't stick
). As I understand it, pixel step of 1 will cause all LCD pixels to be monitored by md_detect_motion. Pixel step of 2 will use about 1/4 of pixels etc.
Which pixel does pixel step start selecting the active pixels? From top left corner?
If I set a grid denser than the pixel step (like a 30x20 grid and a hugely unpractical pixel step of 100), will some of my MD cells be left without a single actively monitored motion detection pixel? Or is the pixel step calculated from a corner of each cell guaranteeing at least one pixel for each cell?
If pixel step is not dependent on cells, does anyone think it should be, at least to guarantee there is at least one active pixel per cell (preferably one in the middle of the cell).
Also (or alternatively), it would be nice to have a some sort of way to make the active pixels visible. It could be a debug option which turns all active pixels red, or maybe the active pixels could be shortly blinked while drawing the grid if this doesn't slow things down.
