VHS tapes won't track, audio speeds up?

[matt314159]

I'm 16 tapes into capturing my family's VHS home movies, and so far I've just run across two tapes that I can't get a good playback lock on with this deck.

Recap.
Hardware Chain: JVC S5400U > S-Video > ES15 > GV-USB2
Software chain: Virtualdub2, capturing 480i HuffYUV avi masters, then Hybrid for QTGMC deinterlacing (faster, bob, making it 59.94fps), cropping, resizing to 720x540 and encoding in H.264 CRF 17.

The two tapes seem like a bad scenario. It seems first, OTA TV was recorded onto these tapes in SP mode, and then my folks stuck them in the camcorder and recorded over it, maybe in SLP mode. I notice my S5400U doesn't show SP/LP/SLP when playing back the 'bad' portions of these tapes, and it just can't get a solid tracking lock, and there's lots of visual artifacts. I've disabled auto tracking and messed with manually adjusting the tracking, and I can make it worse, but not better than what the auto was giving me. Based on the speed of the FF drive as it goes past the cut from bad footage back to old recorded video, you can hear it speed up when it gets to the SP section, and the VCR displays SP when we get back to that point.

I cleaned the heads on this VCR on Friday with 99.9% alcohol and chamois swabs just lightly moistened, and see no change.

I did a one-minute capture of the span where it goes from camcorder footage back to the soap opera recording so you can see what it acts like.

Here is the raw capture in HuffYUV AVI (~580MB)
Here is the clip in H.264 MKV (~85MB)

I think I'm probably just coming up against the limitations of my low-end capture setup, but I wonder if there's any tips or tricks y'all can recommend. A chat with Gemini suggested that maybe a thrift store VCR with less sophisticated tracking might be more easily able to brute force its way through the tape, but as with anything concerning AI, I'm extremely skeptical.

Thoughts? suggestions?

[Aya_Rei]

Only dealt with this once, was an EP Mode VHS copy of a VHS-C or 8mm camcorder tape.

My JVC MV45U could not lock onto it at all (blue screen + chipmunk pitched audio) and my VS30U could lock onto it after 10 seconds of play (normal audio + picture), but the video kept cutting out. So was overall unusable. Tried manually adjusting the tracking too and that did not help.

I suspect the tapes were recorded on a misaligned VCR, so really the only option besides brute forcing and trying to find some VCR that can track it correctly, is to manually aligned your VCR so it can play back these problem tapes correctly.

But that's a very risky process as you are changing the factory standard alignment, from what I heard it's not easy changing it back.

So you'd probably rather misalign a throwaway VCR, not your JVC.

[lordsmurf]

Why is Calibration on?
Why is Overlay (Superimpose) on?

The wrong settings in the JVC VCR will make alignment/tracking issues worse.

[matt314159]

Quote:

Originally Posted by lordsmurf

Why is Calibration on?
Why is Overlay (Superimpose) on?

The wrong settings in the JVC VCR will make alignment/tracking issues worse.

I've tried disabling auto-calibration and manually adjusting tracking, but as noted, I can make the issue worse, not better with manual adjustments to the tracking.

You're right I should have turned superimpose off, but that's neither here nor there for the issue with these two tapes.

-- merged --

When I got home from work, I disabled Superimpose (a setting I missed when I was getting this all set up), disabled auto-calibration in the menu settings (as opposed to temporarily by pressing the Ch+/Ch- buttons on the VCR simultaneously like I was before), as well as turned on "tape dub mode" which sounded like it might give a more transparent signal playback. And I turned on "Video Stabilizer".

This was the resultant capture of roughly the same section of tape. I also tried turning Video Stabilizer back off, and it looked the same or slightly worse.

So that's where things stand right now. I'm going to grab a cheap 2010 funai deck from the storage room at the college where I work, and we'll see if that can manage to brute-force its way past the odd tracking with this tape.

[timtape]

As I understand it, the longer play recordings use a different set of video heads. Possibly your JVC's longer play heads are dirty, worn/damaged or circuitry associated with longer play is faulty. It might be worth giving the video heads another clean just in case. Be careful. Video heads and drums are easily damaged. Definitely no "brute force".

Another possibility is a faulty video recording. That would be checked by playing on a good condition VCR which can play the longer play recordings.

[matt314159]

Yeah it's weird. Since AI theorized that a cheaper, dumber VCR might be able to get a signal lock, I grabbed one from the storage room at work and brought it home, and though I'm having to capture via composite, this is able to get a lock and plays it fine.

So either something is wrong with the JVC or perhaps by its very nature, it's expecting a more refined, stable signal. I don't know. I have been happy with the captures so far, even when I had the on-screen display left on, so I think I'm just going to use this 2010 Emerson unit to capture the tapes the JVC gets cranky with like this one.

But I call this a success!

[timtape]

Of course it's success and in your case success was as near as the next VCR you happened to try. But I suspect success was not due to the Emerson being a modest machine, but because it was in good, or at least better working condition (on the longer play tapes), than your JVC's internal condition.

These decks are now old. These days we cant just go out and buy a brand new one. Increasingly the watchword is, or should be, "condition, condition, condition".

[lordsmurf]

Quote:

Originally Posted by matt314159

disabled auto-calibration in the menu settings (as opposed to temporarily by pressing the Ch+/Ch- buttons on the VCR simultaneously like I was before), .

Are you relying on AI again?
Calibration has nothing to do with channel up/down in the unit or remote.
- calibration is calibration
- channel up/down is manual tracking

Quote:

as well as turned on "tape dub mode" which sounded like it might give a more transparent signal playback.

That is not correct. "dub mode" is for unused recording functions.

Quote:

Originally Posted by timtape

- As I understand it, the longer play recordings use a different set of video heads.
- Possibly your JVC's longer play heads are dirty, worn/damaged or circuitry associated with longer play is faulty.
- It might be worth giving the video heads another clean just in case. Be careful. Video heads and drums are easily damaged. Definitely no "brute force".

- correct
- likely
- correct

Quote:

Originally Posted by matt314159

Since AI theorized that a cheaper, dumber VCR might be able to get a signal lock,

AI is really ignorant when it comes to these video topics. Most of the bots have jumbled up low-knowledge myths/ramblings (especially from Reddit) with the factual writings from sites like this, VideoHelp, Doom9, and some others. It's smushed it all together, and falsely given it equal rankings in quality.

That "cheaper, dumber" theory is technobabble nonsense. It's not technical, not accurate.

Quote:

I grabbed one from the storage room at work and brought it home, and though I'm having to capture via composite, this is able to get a lock and plays it fine.
So either something is wrong with the JVC or perhaps by its very nature,

That 5400 deck was my main concern several threads ago.
Refer back to https://www.digitalFAQ.com/forum/vid...check-vhs.html
It's not suggested, and for a reason.

The mostly likely issue is 30+ years of age and lack of proper maintenance. In proper refurbished condition, it's still not a great deck, but it should outperform a low-end consumer VHS "combo deck" Emerson (Funai) for tracking. It's not fair, not 1:1, to compare a ~2005/2010 deck to an early 1990s deck, when neither likely had much maintenance. If nothing else, that JVC has an extra decade of gravity weighing on the alignment guides.

Quote:

it's expecting a more refined, stable signal.

No.

Quote:

I don't know. I have been happy with the captures so far, even when I had the on-screen display left on, so I think I'm just going to use this 2010 Emerson unit to capture the tapes the JVC gets cranky with like this one.
But I call this a success!

And even with the above all stated -- and it needed stating -- you're doing what is needed here. Good plan, you're getting results.

Could it be better? Easily. But at a higher gear cost.

When money is a project priority, what you've built seems to be working fine for you. Not that I need to give it, but what you're doing gets my approval. Though I'm assuming results are at least above average, given ES15 and not-worst GV-USB2 in use.

As a comparison, I recently built a workflow for $1250, which had a JVC deck, ES15, Cypress composite TBC, and good-version Pinnacle USB. The difference is my gear was all rebuilt (and realigned, cleaned, re-lubed), and both the frame TBC and ES15 were modified. It will make a lovely setup for the gent who bought it! But it was somewhat of a PITA, and I probably won't build more.

Quote:

Originally Posted by timtape

Of course it's success and in your case success was as near as the next VCR you happened to try. But I suspect success was not due to the Emerson being a modest machine, but because it was in good, or at least better working condition (on the longer play tapes), than your JVC's internal condition.
These decks are now old. These days we cant just go out and buy a brand new one. Increasingly the watchword is, or should be, "condition, condition, condition".

^ This. We cannot say it enough.

[matt314159]

Thanks, folks. Yeah there very well could be something wrong with the JVC. I'm just relieved that it's working well for 90% of my tapes. The S5400U purchase was probably my biggest regret of this project, but I'm making it work.

As I've been capturing and processing these home movies at a pace of 1-2 a night, I've been uploading them to a Google Drive folder shared with my family, and my parents especially have really been enjoying watching them every night on their TV. Mom even said the quality looks better than she remembers it, so suffice it to say, it's been a hit with my family.

[lordsmurf]

Quote:

Originally Posted by matt314159

I've been uploading them to a Google Drive folder shared with my family, and my parents especially have really been enjoying watching them every night on their TV. Mom even said the quality looks better than she remembers it, so suffice it to say, it's been a hit with my family.

Awesome.

[matt314159]

I'll admit I do wonder how much quality I'm permanently leaving on the table by using this non-recommended setup. Am I currently getting 90% of the picture information from the tape, when a proper high-end prosumer workflow would get 98%? That kind of thing.

I'm definitely going to hang onto these tapes and store them carefully and if I find myself in a better financial position in a few years, I suspect it might be worth re-doing them right. I know the best time to capture these was 20 years ago, and the next best time is today. In five years, the tapes will be older and that there's the risk of damage or destruction, and the VCRs, TBCs, and cards to play, stabilize, and capture them will be older and refurbished units will be fewer and further between, and certainly more expensive.

But with my parents already almost 70 years old, I want them to begin enjoying these tapes again NOW, and I wanted to do a better job than dubbing on a VCR/DVD-R combo unit or chain mail-in transfer service (they probably do the same thing), so here we are. And surely there is some quality that is permanently lost by not doing them right, now, even if I end up re-doing it properly in five years.

[lordsmurf]

Quote:

Originally Posted by matt314159

I'll admit I do wonder how much quality I'm permanently leaving on the table by using this non-recommended setup. Am I currently getting 90% of the picture information from the tape, when a proper high-end prosumer workflow would get 98%? That kind of thing.

With that non-suggest non-TBC JVC, and the Emerson VHS combo deck, you're getting maybe 60% of image quality. But the ES15 adds back at least another -10% in various noises. So net about 50% quality. It can get much worse, but it can also get much better.

Better deck, better TBCs, better card = 90-99% of possible quality.

(BTW, since we're talking about quality of methods, FM/RF/vhs-decode is maybe 85% of quality at best. But it then strips out -20% quality, due to "halo/ringing vs. blur" and lack of decent DOC. So it's still only 65% net quality. It's mostly an upgrade to low-end consumer VHS decks, not at all S-VHS decks with line TBCs. And sometimes the results are worse than 50%. That method is all over the place, no consistency in results. That project has a lot of FUD/misinformation surrounding it, due to the current kid leadership. I don't like discussing that young peckerwood or his pet project, but sometimes it needs to be mentioned.)

Quote:

I'm definitely going to hang onto these tapes and store them carefully and if I find myself in a better financial position in a few years, I suspect it might be worth re-doing them right. I know the best time to capture these was 20 years ago, and the next best time is today.

Yep.

Quote:

In five years, the tapes will be older and that there's the risk of damage or destruction, and the VCRs, TBCs, and cards to play, stabilize, and capture them will be older and refurbished units will be fewer and further between, and certainly more expensive.

Yes, all risks to waiting longer.

Quote:

But with my parents already almost 70 years old, I want them to begin enjoying these tapes again NOW, and I wanted to do a better job than dubbing on a VCR/DVD-R combo unit or chain mail-in transfer service (they probably do the same thing), so here we are.

Perfectly reasonable.

Quote:

And surely there is some quality that is permanently lost by not doing them right, now, even if I end up re-doing it properly in five years.

Maybe, maybe not. Lots of factors.

Whenever you're ready, get back with me. (Although, in 5 years, I may be retired. Not from everything, just video hardware. But I should still be online, still helping others, as I've done for 25+ years now.)

[Aya_Rei]

Quote:

Originally Posted by lordsmurf

(BTW, since we're talking about quality of methods, FM/RF/vhs-decode is maybe 85% of quality at best. But it then strips out -20% quality, due to "halo/ringing vs. blur" and lack of decent DOC. So it's still only 65% net quality. It's mostly an upgrade to low-end consumer VHS decks, not at all S-VHS decks with line TBCs. And sometimes the results are worse than 50%. That method is all over the place, no consistency in results.

From a quality stand point it starts at basically the same as a prosumer setup, but you get nasty ringing artifacts (and/or low signal to noise ratio if you aren't doing the soldering required setup)

So you start from 90 to 99% but the percentage automatically gets reduced due to extra ringing.

You might be like me and go 'ew ringing bad' and remove it, but that is at a cost of sharpness. So you are losing whatever preserved sharpness gains you were getting compared to a prosumer setup to remove what decode introduces to every tape.

Basically decode introduces ringing that was not present on the original tape (or more ringing if a tape has some ringing baked into it) and to remove/reduce it, you sacrifice sharpness, you sacrifice any potential quality gains to combat this problem the method introduces.

Decode captures could also be more noisy overall, so you may need to run a stronger degrain filter than usual. Though the noise on a tape depends on the tape. I've done noisy looking SP mode tapes and more clean looking SP mode tapes.

At the end of the day the quality of the source tape dictates the quality of your capture. There is only so much that can be done to improve the quality of some EP mode copy of a copy.

With Decode you trade all of these false claims of quality improvement for the time spent trying to set the dang thing up. It beats out whatever consumer composite VCR you use, but that is it.

If tapes were encoded the same as LaserDiscs (where Decode is 100% the best way to extract them), we wouldn't be having this talk now would we.

[lordsmurf]

Quote:

Originally Posted by Aya_Rei

So you start from 90 to 99% but the percentage automatically gets reduced due to extra ringing.

No. You're still limited by the consumer-grade heads and transport of the low-end consumer VHS VCRs. It definitely does not start at 90-99% image quality, unless using a recommended S-VHS VCR mech/model. It starts reduced (~85%), and gets further reduced (~-20%) by adding issues that were not actually inherent to the tape video. It nets better than a composite consumer VCR, but not a suggested S-VHS VCR with line+frame TBCs. --- And that only refers to image quality, audio is still a mess with that method.

But let's not sidetrack OP's thread too much here.

[matt314159]

I'm actually glad y'all brought up the vhs-decode project, because I had been eyeballing it. It appealed to me as a tinkerer, but I do get the feeling that at best it's not ready for primetime. Maybe software will improve down the road to better interpret the captured signal?

LordSmurf, the DigitalFAQ main site mentions offering capture services. Are you the one who does/supervises these? I do have one VHS from 1988 that was such a hit with my family, and myself included that I'd be interested in sending it in to have a HuffYUV AVI made of it using a proper high-end workflow. It would also help me get an idea of what kind of improvement I might potentially see if I re-capture all the tapes later with a good workflow down the road. It's a first-generation SP camcorder recording in good shape that runs about 2:02. I poked around but didn't see an order form, so I'm sure I'm missing something.

Thanks for all your help, I appreciate you guys!

[lordsmurf]

I'll PM you about the tape conversion.

And since you had interest, here's a few more details for...

The main problem with vhs-decode is that it was co-opted by an unemployed GenZ'er who thinks everything should be cheap/free. But that's just not how the world operates. And that's not even what other developers think, but he's the loudest mouth in the room. It works against the project, because it's a race to the bottom. Quality gear is shunned/mocked, while cobbled-together gear from Aliexpress and thrift stores is celebrated. The irony is that, aside from the VCR (and lack of TBCs), the method requires:
(1) "gamer" computers (hefty resources) for decent results -- and ties up said computer for many hours
(2) at least 5x+ the storage space
(3) the Domesday/clone cards that cost 2x+ as much as normal capture cards
(4) plus all the random parts/pieces (amps/etc) that are often needed, because the VCR modding doesn't work consistently
So it's not really saving any actual money, when all prices are netted out. But it does require at least 4x the time per tape.

The software may, or may not, ever really improve. I don't mean little tweak updates, but substantially, like DOS turning into Windows. I've said this for 5 years now. The promise, from day one, has been "just use the method now, and trust the software will get better later". But that's also not realistic, the "later version" is often vaporware, and you're stuck using what actually exists now. The "do it now" ("or you tapes will fade!") is also the scaremongering tactic of fly-by-night "conversion" slop-shops of the 2000s. I've always taken a dim view of that. Tapes are aging, yes, but spare me the FUD/BS.

I try to be upfront with costs, time, and effort needed for conversion. They hide it, for whatever reaosn.

[matt314159]

Quote:

Originally Posted by lordsmurf

(2) at least 5x+ the storage space

That alone is enough to terrify me especially with how AI has upended the storage and memory chip market for consumers for the foreseeable future.

[Aya_Rei]

Yeah the viewpoint is "be cheap by doing all the work yourself"

There are people out there who would rather pay somebody to setup decode for them than to do it themselves. The biggest barrier to entry is the requirement to perform VCR surgery. They view soldering as some easy task anybody can do..

And yeah the main drawback besides overall complexity and noob unfriendliness is that it is way slower overall (and that all depends on computer specs) plus higher storage space requirements.

One 2 hour tape using my mobile workstation laptop for example would take over a day to decode, and I've dealt with needing to leave the thing running overnight during the restoration process.

And the resulting compressed file would be double in size to a Huffyuv AVI file.

[lordsmurf]

Quote:

Originally Posted by Aya_Rei

Ythe resulting compressed file

Just to emphasize, multiple files are required. Not at all like a single lossless capture file, at "only" ~35gb/hour (which is still not small, but at least manageable).

Quote:

Originally Posted by matt314159

That alone is enough to terrify me especially with how AI has upended the storage and memory chip market for consumers for the foreseeable future.

Did you see that Samsung/Micron/Hynix all recently updates their outlooks, and don't see memory shortages giving up until 2028? They don't plan to ramp production, both for profit and their own longer-term safety (not overproducing, not overspending into inventory). So this will likely go on for several more years.

Taiwan saber-rattling may up-end CPU/GPU, too. We're wasting munitions on Iran, and depletion could allow China to attack Taiwan, we cannot defend. Without TSMC, everybody is screwed.

Ugly times.

[Aya_Rei]

Quote:

Originally Posted by lordsmurf

Just to emphasize, multiple files are required. Not at all like a single lossless capture file, at "only" ~35gb/hour (which is still not small, but at least manageable).

Yeah one compressed file for the video, another for the HiFi audio. Both would be 2x the size of a Huffyuv avi file so 4x total.

Anyway feels like decode is only better at tracking messed up signals because the comparisons you see online pair against cheap Chinese cards.

It is all unfair junk to stroke the project leader's ego.

Proper fair comparisons against prosumer gear shows it's true colors as it only being better than junk consumer gear.