Is a full frame TBC necessary or VCR's builtin TBC enough?

[TCmullet]

I think you should try an entirely different capture device, and see how it goes. Dazzle isn't a popular brand on these forums AFAICT, so with something else you might not need an external TBC.

That's odd to think Dazzle isn't popular, at least for the DVC-II. Yes, it was a bear at first, fraught with lots of problems. But we, the user community, solved all of them (www.dazzlegeek.com) in those years 2001-2004 or so, and it has seemed to be a high-quality and reliable workhorse for many of us. Of course, I don't know what Dazzle has been doing SINCE the DVC-II. Also, I can't afford to buy any new card in time for this project to be completed. (Rats.)

This thread is interesting...
https://forum.videohelp.com/threads/224314-Does-TBC-affect-Audio-Sync
also this...
https://forum.videohelp.com/threads/269590-Datavideo-TBC-1000-audio-inputs-question

Very helpful. Thank you.

If you are encoding a DVD, you can flag the audio delay - no need to change the actual audio (the DVD player will make the change during playback).

Not actually authoring DVDs, so not applicable. But whether I did that or stick w/Mpeg2 program streams, either way I have to know how much delay to introduce, whether DVD flagging, or remuxing with a delay.

You could check the actual delay at a given point (accepting it will change) using an audio sync test card (record it to VHS and play it back).

Again, no money for that. But good idea. But when I googled that, I found an interesting quote from a BBC engineer at this link:
http://www.bbc.co.uk/blogs/bbcinternet/2008/12/a_christmas_present_from_the_h.html
"The effect of AV sync has been measured and tested quite extensively by the international broadcast standards bodies and we usually work to a tolerance of +20ms to -40ms (+half to - 1 frame) for a programme delivered to the BBC." Keeping in mind that they are PAL and we are NTSC (+half to -1 frame for us is +16 to -33ms), it's also interesting what he said later, speaking of where audio is 1 or 2 frames ahead of picture, "In nature this is not normal and we can detect sound ahead of vision very quickly and it is 'just not right'!" Very interesting that they allow up to 1 full frame audio delay in material coming to them. Maybe I shouldn't be so picky. Maybe.

I'd forget about S-VHS vs VHS tapes - do the most important tapes on the best machine. First.

That's an interesting philosophy, definitely worth thinking about. However, I find it hard to determine what "most important" is. I consider all of them equally important. (Pardon my arrogance.) However, I wouldn't want to waste (for example) 120 hours on the S-VHS deck for the 120-hr project (which is not the current urgent project) that is originally in VHS-EP. And as I'm wanting to try to get everything done eventually with no head or deck replacing, I still think doing the bulk of the collection (VHS) on this good VHS deck and reserving the S-VHS deck for the small percentage of S-VHS tapes is the best use of resources. But all other factors being equal, yes, I'll do the "most important" things before the lesser ones.

I read the threads you gave. Defintely helpful in clarifying some things, particularly that the delay is 0-1, not 1-2 frames. But one thing seems a bit in error, at least based on my own understanding. It was Davideck's comment (2nd thread):

The video delay through a TBC/Frame Synchronizer ranges from 0 to 1 Frame, and it is constantly changing as the video output reference timing drifts with respect to the video input. The delay can also step abruptly between 0 and 1 Frame whenever an input frame is repeated or dropped in order to maintain frame synchronization.

I think he's referring to a TBC with sync lock to an external sync source. I think that if the sync is locked to *incoming video*, the vertical sync would get reestablished everytime the incoming vertical sync was significantly off. (Let's assume that the TBC is NOT waiting til the frame buffer is full before starting to output it.) If I'm right about vert. sync driven by incoming video, then maybe the video delay might be fluctuating between 0 to 1/4 or 1/2 of a frame. Based on that and on the BBC guy's comment that audio delays are a more forgivable psychoacoustical error than video delays, I'm thinking I should delay the audio an average value of 16 ms (averaging 0 and 1 frame), barring some way to actually measure it. (And of course if it's constantly fluctuating, there's no way to exactly follow it all the time.)


I supposed I'll have to do some (ughhh) TESTING. I'll be processing volleyball, so the sound of hitting the ball might be a good spot for me to tune my "timing ears". Just wish there was an objective (and cheap) way for even a deaf-mute to do this. Also, getting the Amiga out, set up after all these years, and crammed into a tiny space in my rented room is something I wish to avoid unless there's good reason to think it will help. Heck I don't even know if the thing will boot up, or if it does, that the TBC cards will still work. Maybe there will be help files in the software to explain all the switch and buttons I remember being there. I can't even remember the brand or whether I might still have the printed manual.

[davideck]

But one thing seems a bit in error, at least based on my own understanding. It was Davideck's comment (2nd thread):

The video delay through a TBC/Frame Synchronizer ranges from 0 to 1 Frame, and it is constantly changing as the video output reference timing drifts with respect to the video input. The delay can also step abruptly between 0 and 1 Frame whenever an input frame is repeated or dropped in order to maintain frame synchronization.

I think he's referring to a TBC with sync lock to an external sync source.

Some TBC/Frame Synchronizers do not have an external sync lock (or gen lock) input, but they do have an internal timing reference generator that is independent of the incoming video timing. The TBC/FS output sync is NOT locked to incoming video, but rather is derived from the internal reference. This is what allows the TBC/FS output to provide continuous uninterrupted sync regardless of the incoming video status.

The TBC/Frame Synchronizers that do have an external gen lock input use it to synchronize their internal reference to an external source, but the TBC/FS output timing still remains independent from the incoming video timing.

[TCmullet]

Thanks, David. What you said makes sense. I had thought the sync was derived either from external or video source ONLY. Hadn't thought there might be an internal sync generator independent of video source. Yes, some sync source other than video input would be necessary to ensure a solid vertical sync lock throughout the process. I'll have to check my docs (if I can find them), but I still think some full-frame TBCs (including my own) have "sync to incoming video" as an option. Perhaps (just thinking theoretically) those do have an internal sync generator which takes it's *initial* setting from the incoming video but then continues on it's own.

(Been a LONG LONG time since I dealt w/any of this. Tape. Ugh.)

[davideck]

Well, I do know that some older professional TBCs had a "sync to incoming video" option. They would gracefully servo their output timing to keep the incoming video centered within their timebase correction window. Kind of a low pass filter in the timebase error domain. It prevents the need for a whole field or frame of memory.

It also avoids the occasional frame stepping and maintains a more consistent processing delay. However, the recovery performance after a record gap or dropout can be an issue.

[zestefraicheur]

My conclusion 6 years later :
There is no "real TBC" or "false TBC", and the expressions "line TBC" and "full frame TBC" don't mean anything. There are only people who bought an expensive TBC 1000 frustrated to know that a Panasonic DMR-ES10 or a Philips DVDR5520 used as a passthrough can do the job (or even better) for 30 bucks.
And if you want to avoid audio/video sync issue, buy a Canopus ADVC 55/110 with a built-in TBC and a strong audio sync lock system.
If you can't find or afford it, use a VCR/DVD combi as a TBC passthrough and then capture the signal with vMix. BTW Any capture card (Hauppauge or AverMedia pci-e cards) will do the job.

[LMotlow]

My conclusion 6 years later :
There is no "real TBC" or "false TBC", and the expressions "line TBC" and "full frame TBC" don't mean anything. There are only people who bought an expensive TBC 1000 frustrated to know that a Panasonic DMR-ES10 or a Philips DVDR5520 used as a passthrough can do the job (or even better) for 30 bucks.
And if you want to avoid audio/video sync issue, buy a Canopus ADVC 55/110 with a built-in TBC and a strong audio sync lock system.
If you can't find or afford it, use a VCR/DVD combi as a TBC passthrough and then capture the signal with vMix. BTW Any capture card (Hauppauge or AverMedia pci-e cards) will do the job.

Nonsense. This guy either can't read or can't see well, or both (probably both). If you can't tell the difference between a line-level tbc and a frame-level tbc (yes, they do different things), and you can't see the difference between multiple stages of lossy DV processing, you shouldn't be handing out advice to people who know better. And I see you also don't know that much about capture cards.

This is the same old "if it moves and makes noise it's GREAT!" brand of VHS conversion that has been discounted for decades. OK if mediocre results make you think you've accomplished something, but since anyone can get the same results playing a tape into a DVD recorder why would anyone need this "advice" about capture gear resurrected from a long-deceased thread? And you'd think someone would learn something after 6 years.

[lordsmurf]

My conclusion 6 years later :
There is no "real TBC" or "false TBC", and the expressions "line TBC" and "full frame TBC" don't mean anything. There are only people who bought an expensive TBC 1000 frustrated to know that a Panasonic DMR-ES10 or a Philips DVDR5520 used as a passthrough can do the job (or even better) for 30 bucks

What utter nonsense.

And if you want to avoid audio/video sync issue, buy a Canopus ADVC 55/110 with a built-in TBC and a strong audio sync lock system.

#1 - Canopus DV boxes don't have TBCs, that was marketing the loose term "TBC".
#2 - Canopus DV boxes are not magic, and can lose sync just as much as anything else. The "audio lock" is more marketing, and one obviously written by somebody that does not understand what "audio lock" actually is. (Hint: a pro feature not found on this device, and "audio lock" doesn't have anything to do with capture sync anyway.)

If you can't find or afford it, use a VCR/DVD combi as a TBC passthrough and then capture the signal with vMix. BTW Any capture card (Hauppauge or AverMedia pci-e cards) will do the job.

vMix? I'm 99% sure this is just a spammer. I've seen several spam posts for it this fall/winter. That's some sort of streaming editor, not at all a capture program. Just another spammer ******* bumping ancient posts.

[davideck]

There is no "real TBC" or "false TBC", and the expressions "line TBC" and "full frame TBC" don't mean anything...And if you want to avoid audio/video sync issue, buy a Canopus ADVC 55/110 with a built-in TBC and a strong audio sync lock system.

It's always been so simple...

Whenever an analog video signal is digitized, it is captured with an inherent timebase profile. Good or bad, "Timebase Corrected" or not. A TBC is supposed to reduce timing variations in the analog signal during capture. YMMV.

A Frame Synchronizer (or full frame TBC as it is often referred to here) will provide a stable uninterrupted analog video output signal which makes it easier to capture without audio/video skew, etc. YMMV.

Digitizing analog video directly to 1394 makes the above tasks easier in part because there is no need to construct a stable timebase corrected analog video output signal during capture. You referred to this as a strong audio sync lock system.


Merry Christmas LS!