Testing Astro Video Capture in Linux

I recently picked up an older Thinkpad T400 that had Windows 7 installed. After a bit of angst, I went ahead with my plan to scrub the disk and install Fedora 21. Why the angst? Well, I had originally thought of using Windows 7 32-bit to control my older Canon Digital Rebel XT cameras. They are no longer supported by Canon and not at all on 64-bit platforms. But the T400 came with the 64-bit version of Windows 7 thought they seller did provide a 32-bit install disk. However, I already knew I should be able to get basic functionality out of the cameras with gphoto. Even there, the older camera are only partially supported, but I have a Perl script that can trigger long exposures from the camera, so something should be possible and I can't be any worse off than if I stayed with Windows.

But my real interest was in whether or not I would be able to do video capture. The latest thing I've been playing with is setting up for occultation timings. Have a look at the IOTA web site, occultations.org. I've purchased all the equipment: a GPS video timer, a Supercircuits PC164EX low-light camera (which they just recently stopped selling), and already have telescopes. So the question was, could I do video capture under Linux. I'll worry about analyzing video after I have capture working.

The answer is yes. But its more complicated. I have multiple USB capture devices not all of which work equally well. The first one I had picked up had drivers/software that didn't work particularly well under Window 7 64-bit. The video screen would blank out when I started capture which was highly annoying. After posting on the IOTA Yahoo group, I received some recommendation and went out and bought one of each. As it turns out, the first one I bought I already had.

I purchased the  Diamond VC500 One-Touch. I actually already had one. That was was the one that was working strangely with the Window 7 64-bit software. I'm not sure why it was working strangely, but it was do bad I had already deleted the old software and my camera had been sitting idle. When I received the new one, I loaded the software and it worked fine. I dug the old VC500 out of a box in the garage and found it worked fine, too (remember, we're still talking Windows 7) here. 

So the next test was for Linux. It started out not so good. At this point I didn't have a normal lens for the camera so I couldn't be sure, but the image didn't look simply out of focus, it looked downright scrambled. So I bought a C-mount lens, waited a few days for delivery, then tried again.

As it turns out, not all VC500s are created equal. I have two now and there is no external, visible difference. Howeve, lsusb shows one tiny difference: the bcdDevice number for one is 40.01, for the other it is 40.02. The 40.01 works consistently and without fail. Disconnecting and reconnecting causes no problems. Stopping video, killing vlc while it is connected, nothing causes any problems. The 40.02 is another story. Kill vlc while it is connected and you'll have to disconnect the VC500 and wait a few seconds before reconnecting. In fact, just about anything, including just stopping video from vlc, will cause the VC500 to start scrambling the video. That's what I had seen before I got the lens. 

So, yes video capture works. But not all VC500s are created equal.