LabVIEW Vision is complicated. I had a colleague who was recording rodent behavior. He had a sensor that said "Behavior happening now" and he wanted a 10-second video made at that time. His (original) technique was to take a 10" video, close it, take another 10" video, close it, etc., and if no event occurred during the n-th video, he would (in a parallel process) delete that video. ["Events" happened every few minutes, so more than 95% of the videos were deleted during the run of the program].
He found out I knew LabVIEW, but I'd never done any Vision. So he taught me some basic IMAQdx, I learned more from reading the manuals and learning from thie LabvEIW Forum, and I learned about Image Buffers. I also taught him the virtue of writing little Test routines to "figure out how LabVIEW works".
We were recording videos at 30 fps (I forget the resolution, maybe 640x489). We allocated a bunch of buffers, and if the "Go" event wasn't detected, we let newer frames overwrite the old ones. But when we got a "go" event, we noted the Buffer, opened an AVI file, and started streaming from the "start buffer" and continuing until we got sufficient video (5-10 seconds). As I recall, we acturally started the "saving of images" a second or two before the "Go" event. Great fun, seeing what happens before the trigger event.
So learn about allocating Buffers for smoother recording.
Bob Schor
