[BC] Need audio delay buffer

[James Potter]

Robert:
Thanks for the insights. 

This is an HD video recording facility I set up for church.  We use reasonably good pro or sub-pro components, including cameras, projectors, etc., plus a beefy PC with 4 processors.  I don't notice any other performance latency problems on the PC.  However, the signals from the cameras come over Cat-5 cables into Camera Control Units (CCU) where they are converted to SDI (BNC cables). The SDI outputs are routed to a Vaddio video production switcher.  The SDI switcher output is converted from SDI to HDMI through an in-line converter box.  That HDMI Program bus signal is routed to a Kramer video crossbar matrix to feed the various diaplays (plasma TVs on the walls, and three projectors).  One of the crossbar outputs is fed to another Kramer switch that performs source selection and signal conversions.  It serves to select between composite video DVD outputs for opener, closer, and commercials, then to regular Program content (church service) input from the cameras and a PC showing slides.  The output of this Kramer switch is routed to the input of the Black Magic Express recording software in a PC.  All of those intermediate devices are active (plugged into AC) and I assume contribute a conversion latency.  The audio is straight analog out of the audio console record bus.  When you watch the video display on the recording PC and listen to the audio, there is at least 1- or 2-millisecond delay between spoken words and mouth movements.  The delay is similar to some I see on network cable TV, such as O'Reilly or Hannity or others. I heard years ago that for commercial TV, the video and audio signal paths are different -- video via satellite, and audio via T1 landline, wherein propagation delay explains at least part of the latency.  Not a problem in my control both, however -- just a kluge of series-connected video DSPs, each one slowing down the parade a little bit. 

I figure for around $100 if I can retard the audio variably, I can match it up with the lip movement to make a satisfactory experience.  Another post said he used a piece of gear at his TV station during their NTSC days in the manner I described, and it worked fine for years. 

As to your suggestion to measure the latency -- good idea in concept, but I have no idea how to do that in my booth.  We have ears and eyes, and that will have to suffice.

Thanks/Regards/J