03-16-2010 03:32 PM
03-17-2010 07:58 AM
It sounds like a combination of parts of option 2 and option 3 would be best. I've got the imaging source video to 1394 box and it works great, I imagine that the video to USB has the same chip, just with USB communications. Using it and the black box s-video switcher, with RS232 control should make it possible to control switching from the program you are writing. There should be no need to put three video to 1394 or video to USB devices on the computer, just one will do. But, since the price of the switcher + 1 video to USB is equal to the price of 3 video to USB boxes, I guess it depends if you think you or your customers will ever need a fourth channel of s-video. Warning, the video to USB from Imaging source will ship with an ActiveX control that will work nicely with LabVIEW (as long as you don't need to record audio), but NI-IMAQdx 2009 will also read directly from the device. Cost will play a factor here too, the ActiveX control is free with the Imaging Source hardware, NI-IMAQdx 2009 will cost ~$400, and require run-time liscenses if you install it on more than one machine (beyond the machine you develop on).
03-17-2010 02:32 PM
Well, it’s good to know that at least the analog to 1394 converter works well.
Yes, it would be slightly less money to use the switching box in front of only one converter but I think I’d probably rather reduce the equipment complexity by converting each of the three analog streams to 1394 and do the switching in software by selecting the correct cam #. The number of streams will be limited to only three for other design constraints of the product.
Reducing the equipment complexity is the reason I’m looking for a frame grabber with multiple S-Video inputs(that’ll work with LabVIEW). That way I wouldn’t need any external hardware.
Thanks for your input. Anyone else have any?
Doug
03-18-2010 08:24 AM
You may want to consider the Video to USB, since it is cheaper. The only issue you would have is that it would require NI-IMAQdx 2009, if you want to use NI-IMAQdx to bring your video into LabVIEW. You may also want to consider one of the PCI or PCIe framegrabbers that the ImagingSource sells. If you had a desktop PC, or rackmount PC, there would be no software changes to having the machine switch between 3 USB framegrabbers and 3 PCI or PCIe framegrabbers. The ImagingSource also sells a couple of 4 input cards. I'm using the DFG/MC4/PCIe ($690) and it works great. I can stream all eight video inputs (2 cards in one machine) to disk without any problems. The card has four inputs that have BNC connections, but those could be converted to S-video with off-the-shelf converters. All of the cheaper internal framegrabbers that Imaging Source sells have at least 1 s-video input. So, unless you have a machine without enough PCI or PCIe slots, I would choose an internal framegrabber. Remember, with all of these, if you want to use a LabVIEW native method to pull in the video, you will have to use NI-IMAQdx 2009, unless you choose the Video to 1394 converters that Imaging Source sells. If you are willing to work with ImagingSource's ActiveX control (which is a very good ActiveX control), then you can choose any of thier framegrabbers.
03-18-2010 12:11 PM
03-18-2010 12:41 PM
I have had better luck with the ActiveX controls I've used in the past saving AVI files, than I have had with NI-IMAQ's AVI file writing routines. One thing that may work for you is to save the video you want as an uncompressed AVI file with the NI-IMAQ routines, then use FFMPEG (a command line image and movie conversion utility) to convert the movie into one of many formats that you may need. The NI-IMAQ routines have a known problem with some codecs, especially Divx and other codecs that aren't as widely used.
01-04-2011 05:58 PM
What do you mean by the phrases, "pick brains" and "not sure if this" ?