[opendtv] Re: ATSC and Lip Sync
- From: "John Shutt" <shuttj@xxxxxxxxx>
- To: <opendtv@xxxxxxxxxxxxx>
- Date: Sat, 30 May 2009 07:25:41 -0400
----- Original Message -----
From: "Cliff Benham" <flyback1@xxxxxxxxxxx>
So as I originally said, not many are doing their jobs to make receivers
that actually work properly overall.
http://sportsvideo.org/blogs/hpa/2009/02/18/atsc-cea-working-to-make-tv-receivers-less-prone-to-lip-sync-woes/
ATSC, CEA Working to Make TV Receivers Less Prone to Lip Sync Woes
February 18th, 2009
Lip sync errors coming out of broadcast and cable plants is one thing.but
how about lip sync errors caused by poor receiver and set-top box software?
The ATSC and CEA are nearing completion on some work that will take steps
towards solving those issues.
ATSC TSG/S6-4 and CEA R4 are working on recommendations of the proper tools
to be used for receivers and decoders to properly sync audio and video.
At the HPA Conference in Palm Springs industry consultant Adam Goldberg
presented an overview of the work being done. "One of the pitfalls for naïve
decoder implementations is to process a time stamp once and never look at it
again," said Goldberg. That can create problems as small sync problems can
compound over time and, eventually, become big errors that are noticeable.
The good thing for receiver makers is that the recommendations won't lead to
increased manufacturing costs as 90% of the hardware is built-in to existing
silicon. "One can build a car without a non-slip brake pedal but that would
be a bad idea," said Goldberg. "And building a receiver without the pieces
we know are required is doing it wrong."
Among the hardware recommendations will be Program Clock Reference (PCR)
latching/loading hardware that captures incoming PCR information coupled
with a software controllable variable rate oscillator.
"Once you have the clock data you have to use it and process each
Presentation Time Stamp (PTS) for each audio and video unit and then adjust
each audio and video unit so that it is as close as possible to each PTS."
Historically some manufacturers have run into the classic problem of the
theoretical meeting reality when it comes to decoder design and sync issues
when processing MPEG-2 signals. "The physical decoder is not the same as the
theoretical described in the standard," explained Goldberg. "The theoretical
assumes instantaneous decode time and instantaneously fast memory. But that
isn't possible in a real implementation."
----------------------------------------------------------------------
You can UNSUBSCRIBE from the OpenDTV list in two ways:
- Using the UNSUBSCRIBE command in your user configuration settings at FreeLists.org
- By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word
unsubscribe in the subject line.
Other related posts:
