On Tue, Jul 31, 2001 at 10:56:32PM +0800, Mathias Koerber wrote: > 2. What options would I have for multi-input cards? I'd like to > capture continuously off both (or more) inputs and not have to > switch between inputs. Can v4l2 do that with one card? None that I know off. You would need to put 4 cards in one machine, however, I have read that there can be bandwidth problems depending on the size of your capture image. > 3. I found the motion package to detect/mark motion in captured video. > Does something like that exists in which certain areas of > perpetual/expected motion can be excluded from detection? If you were to read the documentation what comes with motion you would find it has a mask feature that allows you to mask out areas of motion you do not want triggering an event. > 4. Just as a rough guideline, how much space does captured video > take per second/minute etc (assuming PAL and full motion)? It depends on the resolution, capure mode, and whether you compress the data. Assuming PAL 704x580, RGB32 format, no compression, and 25 frames per second: 704 x 580 x 32 (bits) x 25 (frames per second) = 326,656,000 bits/second. (about 40 mega bytes per second) (By the way, I'm in NTSC land, and couldn't remember the vertical resolutiom number for PAL, so I guessed). -- Brian Litzinger <brian@xxxxxxxxxxxxxxxx> Copyright (c) 2000 By Brian Litzinger, All Rights Reserved