Set up an acceptable view for instant replay, use it for the stream and rewind the stream.
Why is it whenever we have a problem with human error (i.e. ball inflation & 25" inner width 2x4 boxes) we need whole new systems to solve it? We don't need people doing trig for camera placement, creating cost estimates, researching video camera tech.
My head elc mentor (and I suspect most of yours could too) can hook up an R-Pi with a 1080p wide-angle camera and an Ethernet cable to another with a keyboard and a monitor to record/play back full field video in less time than it takes to pump up a ball. I could figure it out in a day probably.
Why is everything so hard (and harsh)? We can build swerves, track vision targets, and launch a 24"-26" ball 50ish feet... but we can't record a video?
Where did all these come from?
http://www.youtube.com/playlist?list...joh1iAyo73NSrP
Sorry, this is a little indignant however, this is actually not hard... just different.