VRVS Virtual Room Video conference System Testing Results

Geoff Pearce, Mayly Sanchez and I (Leon Mualem) have tested this video/audio conferencing system available from CERN at  www.vrvs.org .  The general impression has been good, with a few caveats:
 

Systems tested:

  • First, the setup has only been seriously tested by us on Windows NT computers, with some point to point tests between Unix (Linux/SGI) and NT.
  • Sound quirks:

  • Second, Sound can be picky.  The secret is to test, test, test.  Try out the microphone using whatever native recording facility available on the machine you want to use.  This should NOT be done 5 minutes before the conference.  Preferably a week before, to give time for others to help you.  Some microphones/sound cards can be picky.  I have had good luck with Sound Blaster cards, and labtec microphones, but your mileage WILL vary.
  • More sound quirks:

  • Third, Sound can be tricky.  There are many possible settings.  There can be a serious feedback problem if you have full duplex mode set.  The sound played through speakers gets picked up by the microphone and sent back out.  If this happens at more than one site, it can become positive feedback, very loud.  My recommendation is to try muting the microphone when you are not talking.  The settings are found in the RAT control panel accessed from the Options box on the audio tool panel.
  • Video tests:

  • Video tends to work, but you need to tell it to transmit when you start.  This is done from the Menu panel of the vic video panel.  There is a little transmit box at the top that needs to be checked.
  • Despite what you might see on the vrvs web site, just about any windows capable video camera will work just fine.  I use a parallel port b/w camera that I bought for $50 a few years ago.  Other people have used USB port cameras.  Bottom line, no capture card is necessary.  If you can see video in wondows, you can transmit video.
  • You don't need a video camera to see the video other people are transmitting.  Also, document sharing does NOT use the camera, so video is completely optional.
  • It appears that vidoe cameras for Win NT are not trivial to find.  Logitech makes one, the Quickcam Pro for parallel port.  It sells for about $150 US, and  supports NT.   You have to go to the software updates page and download the driver.  Try www.logitech.com though there may be others.
  • Winnov vedeum videconference Pro has been used successfully by Geoff Pearce of RAL.  It is a PCI capture card based system, #275 UK, $375 MSRP US.  www.winnov.com
  • Creative, the soundblaster people have a cheap WebcamII for about $60.  It is a parallel port device, so performance won't be great, but it is cheap, and has NT support on their web site.   http://www.soundblaster.com/video/webcam-2/
  • Whiteboard:

  • The most difficult system to use was the whiteboard.  It appears to be a bit picky about startup order, so it is best to wait until everyone has joined the conference.  The transmission from NT to NT seemed to be okay, but there seem to be unresolved issues between versions on Unix and NT.  The NT version is supposed to be compatible, but I was only able to get it to transmit Unix to NT, but not the other way.  I was able to import a postscript file in both Unix and NT.  The Unix system did not transmit the postscript to the NT machine though it did transmit other whiteboard drawings.  The windows version of the whiteboard can also import postscript, but the ghostscript executable and DLL files need to be in the windows system directory, usually WINNT/system32.  I was able to see the postscript on two NT windows this way.
  • Conclusions:

  • The system is useful, sound quality is good, except for the possibility of echoes.
  • Video is also possibly useful, and doesn't seem to impact performance on the tested machines.
  • The whiteboard may have some problems, but may not be an issue since everyone will be using computers anyway.  We may just  require everyone to post to hypernews, or on a public web server and the link posted to hypernews.

  •  

    Feedback:


    Other feedback is certainly welcome, you can contact me at mualem@hep.umn.edu