[opendtv] Re: Lip sync problem resurfaces

The -1/+2 frame rule-of-thumb only works at the end of the process,=0D=0Awh=
ere the signals meet the human=2E  Upstream processes should have a much=0D=
=0Atighter tolerance, or be correctable to a tighter tolerance=2E  Identify=
=0D=0Awhere the needed sync points are, break the system up into manageable=
=0D=0Achunks, and make those subsystems correct=2E =0D=0A=0D=0AAs far as th=
e great unknown of the home viewer, what about using some=0D=0Asort of "hom=
e networking" to communicate what the processing delays are?=0D=0AThese are=
 usually video delay, as previously mentioned=2E  Then the audio=0D=0Adelay=
 in my processor could be automatically set to match, rather than=0D=0Ame h=
aving to do it on a per-input basis, as I have to do now=2E=0D=0A=0D=0AKen =
Hunold=0D=0A=0D=0A-----Original Message-----=0D=0AFrom: opendtv-bounce@free=
lists=2Eorg [mailto:opendtv-bounce@freelists=2Eorg]=0D=0AOn Behalf Of Gary =
Hughes=0D=0ASent: Sunday, April 25, 2004 2:14 PM=0D=0ATo: opendtv@freelists=
=2Eorg=0D=0ASubject: [opendtv] Re: Lip sync problem resurfaces=0D=0A=0D=0A=
=0D=0A> From: Eory Frank-p22212 [mailto:Frank=2EEory@motorola=2Ecom]=3D20=
=0D=0A> I think the negative delay (audio leading video) is even=3D20 tight=
er =0D=0A> than 30 ms=2E Due to the slow speed of sound relative=3D20 to th=
e speed of=0D=0A=0D=0A> light, it seems that our brains are=3D20 comfortabl=
e with a small amount=0D=0A=0D=0A> of audio lag, since this=3D20 occurs in =
the natural environment=2E But =0D=0A> when audio leads=3D20 video, even by=
 a small amount, it is very =0D=0A> unnatural and disturbing=2E=0D=0A=0D=0A=
When I was working on transport splicing the goal was no worse than=0D=0A-1=
5ms/+30ms for lip synch=2E Much more than that and it is detectable=2E=3D20=
=0D=0A=0D=0A=3D20=0D=0A> Require audio loop-through, so you can delay audio=
 appropriately=2E =3D20=0D=0AI rahter like that idea=2E=2E=2E dial in the a=
udio delay=2E But wouldn't you=0D=0Aalso need an option to delay video by a=
 frame period for cases where=0D=0Aaudio is already too late=2E Or is the a=
ssumption that the synch problem=0D=0Ais always caused by delays in the vid=
eo chain?=0D=0A=0D=0Agary=0D=0A=3D20=0D=0A---------------------------------=
-----------------------=0D=0A=0D=0A=3D20=0D=0AThis email message and any fi=
les transmitted with it contain =3D=0D=0Aconfidential information intended =
only for the person(s) to whom this =3D=0D=0Aemail message is addressed=2E =
 If you have received this email message in=0D=0A=3D error, please notify t=
he sender immediately by telephone or email and=0D=0A=3D destroy the origin=
al message without making a copy=2E  Thank you=2E=3D20 =3D20=0D=0A =0D=0A =
=0D=0A---------------------------------------------------------------------=
-=0D=0AYou can UNSUBSCRIBE from the OpenDTV list in two ways:=0D=0A=0D=0A- =
Using the UNSUBSCRIBE command in your user configuration settings at=0D=0AF=
reeLists=2Eorg =0D=0A=0D=0A- By sending a message to: opendtv-request@freel=
ists=2Eorg with the word=0D=0Aunsubscribe in the subject line=2E=0D=0A=0D=
=0A=0D=0A-----------------------------------------=0D=0AThis message (inclu=
ding any attachments) may contain confidential information intended for a s=
pecific individual and purpose=2E If you are not the intended recipient, de=
lete this message=2E If you are not the intended recipient, disclosing, cop=
ying, distributing, or taking any action based on this message is strictly =
prohibited=2E=0D=0A
 
 
----------------------------------------------------------------------
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: