Audio sync drift on ~50% of VHS captures (AG-1980 + Sony GVD-800 + QuickTim

[CaliGator]

I’ve been reading a lot of older threads and guides, but I’m seeing conflicting advice. I know this topic has been discussed before, so apologies for the repeat — I’m just hoping to get a straight answer from people with real-world experience, especially for a Mac-based workflow.

Workflow
• VCR: Panasonic AG-1980 (line TBC ON)
• S-Video + RCA audio
• Sony GVD-800 as FireWire passthrough
• Mac mini 2013 (Mojave)
• QuickTime Player 7 DV capture (one file per tape)
• Must stay on Mac — WinDV not an option
• Drift exists in raw DV capture (before encoding)

Issue
• ~50% of tapes have gradual audio sync drift
• Starts in sync, off by end of tape
• Home-recorded VHS
• Drift amount varies
• No dropped frames reported
• Some tapes are perfect

What I’m trying to figure out

Is this typical VHS timing instability not fully corrected by the AG-1980 + Sony DV conversion?

Would adding one of these likely fix it?
• Panasonic ES10 / ES15 passthrough (VHS → ES10 → GVD-800 → Mac)
• Canopus ADVC instead of Sony
• Both together

Trying to build a reliable, set-and-forget capture chain.

[Alwyn]

Is this typical VHS timing instability

Play your tape directly to the TV and check the sync. That will help start to isolate the problem.

[pcspeak]

I Had Different Hardware!

I had a similar problem with audio drift, a long time ago.
I think I stopped using the S-Video connector and used the composite connections (yellow cable?)
Not sure if I got a better/cleaner video transfer, but the audio drift disappeared and the result was good enough for the videos I was converting.

[VWestlife]

Standard DV does not have locked audio, and can drift over time. They did this purposely, to force pro users to upgrade to DVCAM or DVCPRO to get locked audio.

[aramkolt]

The Canopus DV conversion products ADVC100/110 do have a switch on the bottom for locked audio, so you might want to grab one of those if you're definitely using a DV capture path. The other issue could be hardware limitations of the Mac mini, so would be good to rule that out with a more modern Mac, though that can get expensive to find a firewire 800 to thunderbolt 2 adapter required for those that lack thunderbolt ports.

Lastly, there are standalone DV recorders that would avoid the use of a computer altogether. The ones I'm most familiar is the Datavideo DN series which depending on the model can accept composite, component, and S-Video, or you can just feed it DV from another device to record. The most full featured in terms of IO is the DN-400. Most of the DN series will allow you to directly store DV files to the removable hard drive module which you can then transfer the files off of and to your Mac. Since they have analog video inputs, so you don't even technically need to use a separate DV converter. The DN-400 supposedly has improved audio sync over some of the other models.

Other question is if the AG1980 has been refurbished or not since that could be playing a role to some degree. I do fully refurb those if you need that service.

[rgr]

Standard DV does not have locked audio, and can drift over time. They did this purposely, to force pro users to upgrade to DVCAM or DVCPRO to get locked audio.

Are you sure? I digitized one tape using a Digital 8 camcorder (analog in, DV out) and the sync was perfect.
Not to mention that digitizing using the DV format is obsolete these days.

[lordsmurf]

Standard DV does not have locked audio, and can drift over time. They did this purposely, to force pro users to upgrade to DVCAM or DVCPRO to get locked audio.

Are you sure?

He is 100% correct.
"Audio lock" is a very specific term in DV video, referring to audio integrity within the stream. DV25 (MiniDV, Canopus boxes, etc) does not support audio lock.

I digitized one tape using a Digital 8 camcorder (analog in, DV out) and the sync was perfect.
Not to mention that digitizing using the DV format is obsolete these days.

"Audio lock" has nothing to do with audio sync.

Canopus (now a defunct company sold long ago) was a marketing machine whore back in the 2000s. Their marketing goons were fast-and-loose with terms, and somebody there completely misunderstood what "audio lock" is/was. Not too different from how vendors like B&H mispresented DV boxes as TBCs. Those myths just will not go away.

If a DV box has "kept sync", it will always mean that audio was resampled to adjust time. There is nothing special about DV capture boxes. Those can, and do, drop frames, like any other capture card. One of the "tricks" to hide drop/insert is to mess with audio sampling. Resampling was a go-to especially in the late 90s to early 00s, the era of Canopus boxes.

Adam Wilt wrote about this in depth, several places, back in the 2000s.
https://www.adamwilt.com/DV-FAQ-tech.html#LockedAudio
I had conversations with him decades ago, but I didn't realize who he was at the time. I have more papers on this topic, some also from Wilt. But I'll have to locate those. I'm currently sorting my decades worth of documents.

The Canopus DV conversion products ADVC100/110 do have a switch on the bottom for locked audio,

That doesn't work how you think it does. Nothing can "lock" audio in DV25.

Other question is if the AG1980 has been refurbished or not since that could be playing a role to some degree.

Worth repeating. OP, tale note. ^
And contrary to misconception, AG-1980P is not "the best" S-VHS VCR, just outstanding at certain tasks. It has weaknesses, especially to several JVC models

Standard DV does not have locked audio, and can drift over time. They did this purposely, to force pro users to upgrade to DVCAM or DVCPRO to get locked audio.

I'm really impressed that you know this. I don't often get impressed.

[aramkolt]

I'd always kind of heard that DV captures didn't have quite as many audio sync issues as many other capture chain options, so this is interesting to me. I'd imagine there'd have been a lot of DV conversion box returns if audio sync issues were common with consumer VCRs of the era. Should be less likely with the AG1980 having a multiline TBC I would think.

Any chance OP is willing to send me one of their tapes that does it the worst so I can test if it is reproducible with the Canopus products and one of my AG1980's? I can try both the "locked" and "not locked" audio. Also, how long does it take for a meaningful/obvious audio delay to show up when capturing? Ideally it'd be a tape that you don't need back, but I can mail it back if you really need it still.

Other thing you could try is capturing with iMovie instead. It'll save as .dv files within the "project" contents. Not sure why that'd make a difference, but you never know and it should still be free.

Also, while not explicitly stated, I'm guessing that the audio gets more and more "behind" the video as the capture goes? That'd suggest that frames are being dropped, so when the frames are played back, there's "less" video than audio to last the duration of playback compared to audio. I'm also not sure how you'd see "dropped" frames with your current setup as QuickTime won't report dropped frames. Something like DV rescue might give some clues, but I'm haven't really used it myself to say for sure.

[aramkolt]

I did play around with DV rescue a bit - it can tell you about both audio and video errors in a DV file and might have some ways to correct them. You'd be surprised how many errors a short DV clip can have. I had a random 34 second clip that I transferred 20 years ago from a DV tape and there were DVrescue says there are no video errors, but there are 89 lines (blocks?) of audio errors, all seemingly sequentially, see attached screenshot. Also of note, these 89 lines must not represent much, because the "percentage" of lines with errors for the entire clip is reported to be 0.00%


[Attachment 92070 - Click to enlarge]

Also a little strange that frame 0 has a different set of timecode from the rest despite this being a single clip.

I don't fully understand what I'm looking at with DVrescue, but it might reveal something to those who use it more often that could be helpful.


Lastly, some other Googling shows if you happen to have audio sample rate incorrect between the playback and capturing device (32 vs 48khz), that'll usually cause audio drift. Problem is that not all modern capturing apps will let you specify the audio sampling rate. DVrescue I image will though.

[The_Doman]

What I’m trying to figure out

Is this typical VHS timing instability not fully corrected by the AG-1980 + Sony DV conversion?

Possible you could do some repeated test captures from a known stable video source like cablebox/DVD/HDD player/recorder or whatever you have with composite/S-Video?
Also you then can switch channels/stop/start playback during captures to see how your capture setup reacts to that.
I often do this myself when testing capture devices/tools.