Audio sync and video delay with TBCs
 

I'm starting this as a new thread as there are some refs out there but they are old. For example -

Read more: http://www.digitalfaq.com/forum/vide...#ixzz2WoO35UrM

I believe that in pro video, audio is always routed through a delay when the video is processed as with TBC. A TBC with a full frame buffer will normally introduce a frame of delay (for PAL 4mS and NTSC 3.3mS) and this is considered significant, resulting in rubbery lip-sync. In my field (DVD and BD authoring), we would flag this in QC and correct it. I'm about to ingest a lot of VHS tapes, and will be using a TBC-100.. I'm going to delay the audio at ingest rather than later at file level. What do others in this forum think?

I'd do a test capture first and see how much of a delay the entire workflow introduces. If you are capturing from a capture card without integrated audio capture (like the ATI All-in-Wonder AGP cards), there will be some clock drift since the sound card and video capture card's internal clocks will not be synced. This is obviously not a problem if your capture card grabs audio as well. I haven't noticed any sync issues using the AVT-8710. My Video Toaster is much more noticeable in that regard even in straight analog mode (digital effects mode forces a 2 frame delay due to grouting video through frame buffers).

It would be a half frame at most, 1/60th of a second. You won't see lipsync errors until about 1/4th of a second. Most of the time, the lipsync is off anyway, either way, depending on what happened with it along the way.

It's really a non-issue.

Even an integrated chipset, audio+video, has clock sync/drift. It's not just external audio cards.

There's no issues using a AVT-8710, assuming it's not got that evil bad chipset issue.

Anything from 2005-2012 isn't old. What was true then is still true now, in almost all cases.

Hi NJRoadfan, thanks for your reply. This project is a weird mix of "professional" and "domestic", given the source (VHS). I'm capturing to MPEG-2 using one of those little USB A/D dongles.. it has Composite, S-Video, and Stereo RCA inputs for audio. It does a surprisingly good job, except for sources with bad timebase, such as VHS or (even worse it seems) Hi-8. As the audio is processed together with the video prior to the USB input, I would expect not to see any inherent A/V sync drift. I'll try some test signals with good A/V sync, to see how good the system is natively.

Hi admin, thanks also for your reply. As the TBC-100 is advertised as a "full frame memory" device, I would expect it to buffer a complete frame before output, hence 4mS delay for PAL. Is this incorrect?

In my experience, a trained observer will notice some degree of rubberiness for a one frame offset, more so if the audio is late. At 2 frames it starts to become an issue, and at 3 it would definitely need to be fixed.

The content that I have (lectures) is not critical.. but I am curious to see what people on this forum do regarding this, as in pro video it is always addressed. For example, a TBC with SDI input would delay embedded audio, if not the audio would go through a separate delay unit.

I started a new stream on this topic, as I noted a message advising that the quoted discussion was old.

I have read the blurbs on chipset problems and I don't think I have one of the "evil" ones!

what exact "USB A/D dongle" are you using?
if it is a easyCRAP or similar that is your problem -
get an ATI capture device either an AIW if running XP or a 600USB if running 7

1 frame (field) interlaced = half of an image at 25fps (29.97fps NTSC). Even cameras introduce delays. It's not 100% perfect. The goal is to stay within limits, not try for impossible imperfection. TBCs are used professionally. So I really don't see what the problem is here.

If there's a sync issue being observed, it's likely with the capture device intoducing the delays. What is it?

Hi8 is an excellent format to capture. If there are issues, that also points to the capture device being inferior.

Thanks for your reply folks, but please read the thread. There is no problem with the capture device, even though it is a low-cost USB dongle similar to EasyCap. I can capture SD video with complete reliability and no A/V sync issues, and even though this is just a domestic-level workstation that I threw together at home on an old XP box (as opposed to any number of high-end systems that I can access at work), it does the job just fine. Right now I'm looking to capture about 200 VHS tapes with good quality first generation content. I am seeing some time base issues which can be fixed with a TBC. Down the track I will also look at some Hi-8 captures, and my (admittedly limited) experience of the format suggests that tearing etc might be more prevalent here.

That's the background, but the title of this thread is: "Audio sync and video delay with TBCs" and the question is: do the users of this forum feel that a TBC which introduces a delay of one PAL or NTSC frame should be accompanied by a audio delay unit to compensate? This is standard procedure where TBC in a pro video industry (eg Broadcast), but I have not seen many references to it here, and those that I have seen suggest that at this level it is regarded as a non-issue.

What do others think? Personally I'm not that good at detecting A/V sync issues down to a frame (for me it becomes unacceptable at about 3 frames) but my QC guys can do it unfailingly.. so maybe this is nit picking to some but I would be interested to hear thoughts from others.

I'm also seeing statements such as "any video delay would only be 1/2 frame anyway" which does not sound likely to me, as units such as the TBC-100 are advertised as a "full frame memory" device, so I would expect a delay of pretty much exactly one frame (2 fields). Maybe I should set up a test rig and measure the video delay of the TBC-100.. I'll do that in the next few days.

But that's the thing ... timebase issues (real or artificial) can be caused by low-grade devices. The EasyCAP devices are known (infamous) for issues of this nature.

In all likelihood, this is the net result of several issues -- not great VCR, lack of TBC, inferior capture card, etc.

There's a reason I had to buy good gear many years ago. There's just no way around it if you want to do a good job.

the difference between an easyCRAP and an ATI card are NIGHT AND DAY!!

and the ATI cards can be found on ebay so ridiculously cheap there is no reason to settle for a inferior capture card.
i won the bid on a brand new sealed AIW9200 for 99 cents!!

Thanks LS, I gather that you are still unwell?
Thanks also to volksjager for his reply.

This is starting to sound like the Monty Python "argument" sketch:
M: I just paid!
A: No you didn't.
M: I DID!
A: No you didn't.
M: Look, I don't want to argue about that.

.. and I don't. Let me re-iterate: I DON'T have a capture sync problem. I'm sure that you are quite correct in that EasyCAP (and "DVD Maker" which this USB device is labelled) have been known to cause A/V sync problems, but this setup is just fine. Overall capture quality is OK too, and yes I have sent test signals into it and checked the result. If it's not broken, I won't be fixing or replacing it. I will add a TBC to the video chain because I am noticing some degradation of vertical lines etc which a TBC can fix.

What I DO want to talk about is: "do the users of this forum feel that a TBC which introduces a delay of one PAL or NTSC frame should be accompanied by a audio delay unit to compensate?"

Further thoughts, anyone?

The TBC-1000 gives the kind of delay you're looking for. :)

Thanks admin! I didn't know that.