Developing a analogue capture card with in built tbc

[mountaincabbage]

Hello all,

I've run a videotape digitization business for around a decade now and we're currently exploring the idea of developing a bespoke capture card with the goal to improve our existing set up and workflow. For us, our ideal card would be a 4 channel s-video/composite/L/R audio.

At the moment I'm attempting something simple, I'm currently attempting to have a PCB made with 1x composite 1x L audio 1x R audio. If that works well, we'll add s-video, and then if that works, things get interesting. I think after that point we'll try and build in a full frame TBC into the card, which given these are no longer produced would be highly beneficial for anyone attempting to digitize analogue video.

I appreciate this is an ambitious project, but I thought it would be worth while posting this here and asking fellow enthusiasts what they would like to see in their ideal analogue capture card setup?

[johnmeyer]

The problem I have had with most analog capture cards for digitizing traditional NTSC and PAL sources is that users (including me) cannot get them to easily produce an acceptable result. You don't need to look further than the posts on this site to see the issues. These include:

Dropped frames
Blown out highlights
Color issues
Sync issues, including time base flagging, jitter, and inconsistent starting points for each scan line

Adding some sort of TBC to address the last set of issues is an incredibly ambitious goal (are you planning to do custom ASICS?), but even if you don't get around to that you really need to address the first three issues above so that, out of the box and without a ton of fiddling, a user can capture without dropping frames and have the gamma and chroma properly represented.

I still often recommend using an old DV or Digital8 camera for capture because, despite their horrible compression and color space issues (the latter is not so bad for you PAL people), you almost never get dropped frames, and the levels and colors, within the confines of DCT compression, are properly represented.

So my request, if you really think you can do this, is to provide the simplicity and "idiot proof" attributes of DV capture with a card which uses modern, robust compression combined with a full color gamut. The TBC can still be handled in the deck, although those are hard to find and hard to keep working. If you really can offer a TBC, you'd sell one to every person in this forum, but I don't think you can just program an Arduino to do the job.

[dellsam34]

The biggest problem today in capturing analog materials is the video cassette players themselves, With the slow deterioration of tapes it is hard to find VCRs decent enough to cope with such tape conditions without paying a premium for high end VCRs and their costly maintenance and repairs.

On the capture hardware side, The closest thing to DV easiness is SD SDI or SD HDMI, So if you are going to design a capture device from scratches go for those two ports then find a SDI/HDMI recorder so you can ditch the computer for capturing and avoid all the headache, Multiply the cards and recorders for multiple tasks, Then transfer the files to computer later on if you want to do post processing and encoding.

[mountaincabbage]

Thanks for your responses. The capture cards we currently use do handle video well, it will be interesting to see if we can improve tracking/jitter/dropped frames with what we can do with the analogue signal processing prior to processing it to digital.

@johnmeyer -
Dropped frames - We should be able to assess the dropped frames performance as we'll have full control over the card (which is PCIe). TBC would help with any corrupt frames provided there is only a few of them.
Blown out highlights - This could be handled with adjustable gain on the input levels, something we can explore down the line. This could also be adjusted software side.
Color issues - we have a comb filter built in the card which should help with that, again we'll have to see the results.
Sync issues - attempting to address with the TBC stage.

@dellsam34 - Yes, the hardware available for playing back tapes is dwindling. Pushing the boundaries of the capture card side may help widen the players that can be used for digitization. I did work on a project with Singmai (who you might be aware of) to try and make a new VCR, however we got stuck a while ago. If we did have success in making equipment it would be our ultimate goal to produce new VCRs, although at this stage it is far-fetched idea.

Setup wise for us, PCI cards works fine, we like to have full control over the card and integrate it with our systems to start/stop and handle file management automatically. A USB device could be created easily later down the line.

I'll keep posting updates here on how things go.

[Barrythecrab]

Very sad about SingMai and Dan passing away. What has happened to his work, I wonder.

[themaster1]

So you're trying to recreate a DataVideo TBC-100 (PCI) in essence

https://www.modesdemploi.fr/datavideo/tbc-100/mode-d-emploi

[mountaincabbage]

@Barrythecrab - Yes, it's very sad, he was a lovely person

@themaster1 - I didn't know Datavideo ever made a PCI card, thanks for providing that, yes would contain a lot of the functionality we are going for.

[themaster1]

Well maybe make more research before trying to reinvent the wheel. Makes sense imo.

[dellsam34]

Yes, the hardware available for playing back tapes is dwindling. Pushing the boundaries of the capture card side may help widen the players that can be used for digitization. I did work on a project with Singmai (who you might be aware of) to try and make a new VCR, however we got stuck a while ago. If we did have success in making equipment it would be our ultimate goal to produce new VCRs, although at this stage it is far-fetched idea.

Yes I'm aware of the Singmai VCR project and have exchanged e-mails with Daniel about some ideas, may his sole rest in peace.
The idea of a VCR is not far-fetched if you use the current supply of VCRs mechanisms circulating in the market and make your own playback and system control electronics and a simplified external case, Playback only should not be that complicated, you have 3 motors (Drum, capstan and cassette loading) cassette sensors and a mode switch, a RF module, and A/C head preamp, All this should be made in a small PCB.

I didn't know Datavideo ever made a PCI card, thanks for providing that, yes would contain a lot of the functionality we are going for.

It is not an active PCI card, it just conveniently mounts to a PCI slot, all connections are made externally, the PCI connector is blank, it has no coper pins. If you open a TBC-1000 you'll find something similar with an added distribution amp.

[lordsmurf]

the goal to improve our existing set up and workflow.
I appreciate this is an ambitious project, but I thought it would be worth while posting this here and asking fellow enthusiasts what they would like to see in their ideal analogue capture card setup?

You're missing something. We shall discuss.

Blown out highlights
Color issues

Also IRE/pedestal.
Image values (color fidelity, luma, etc) are easily the most damning issues. Most cards fail at good reproduction.

Blown out highlights - This could be handled with adjustable gain on the input levels, something we can explore down the line. This could also be adjusted software side.
Color issues - we have a comb filter built in the card which should help with that, again we'll have to see the results.

No. This should not require user intervention. What we refer to is cards screwing up the quality that existed on the tape. Cheap chips, terrible default settings, or both, are to blame.

Adding some sort of TBC to address the last set of issues is an incredibly ambitious goal (are you planning to do custom ASICS?),
If you really can offer a TBC, you'd sell one to every person in this forum, but I don't think you can just program an Arduino to do the job.

Correct, far more ambitious that people realize. It's been a dead-end conversation for many years now. The chips needed for SD analog TBCs have not been fabbed since the 00s. And it would take very custom 1's/0's ground-up type coding to effectively make a new chip. It's not merely an exercise it rounding up some off-the-shelf chips/boards.

I still often recommend using an old DV or Digital8 camera for capture because, despite their horrible compression and color space issues (the latter is not so bad for you PAL people), you almost never get dropped frames,

It's merely a lack of reporting that is the feature of DV boxes, and not a lack of drops/inserts. Those happen, just silently. Same with many DVD recorders. There's nothing special about a DV codec, or the hardware inside DV boxes, that would make it drop/insert any more/less than any other ingest device. The difference always comes down to reporting. "If I close my eyes, you can't see me?"

ditch the computer for capturing and avoid all the headache,

Preview would still be the issue.

So you're trying to recreate a DataVideo TBC-100 (PCI) in essence

Not for capture.

If you open a TBC-1000 you'll find something similar

Emphasis here, because it's not the same.

[dellsam34]

Preview would still be the issue.

Not at all, connect a monitor to the HDMI monitor port. That's how I'm setting mine up.

[mountaincabbage]

Thanks for your comments.

The idea of a VCR is not far-fetched if you use the current supply of VCRs mechanisms circulating in the market and make your own playback and system control electronics and a simplified external case, Playback only should not be that complicated, you have 3 motors (Drum, capstan and cassette loading) cassette sensors and a mode switch, a RF module, and A/C head preamp, All this should be made in a small PCB.

I don't want to go too off topic as this is about capture cards - I thought it was simple until we started trying to do it with Singmai, and that was utilising existing playback mechanism and the video/audio heads. To jimmy something together might be possible with a lot of effort, but then re-creating the loading, playback and video head is a large project.

Also IRE/pedestal.
Image values (color fidelity, luma, etc) are easily the most damning issues. Most cards fail at good reproduction.

We should have control over video parameters with this card, I'll have to play around with it when we get the first version back and see how it compares. It might be we can make some pre-sets for certain formats, or dynamically vary the parameters based on the input.

Correct, far more ambitious that people realize. It's been a dead-end conversation for many years now. The chips needed for SD analog TBCs have not been fabbed since the 00s. And it would take very custom 1's/0's ground-up type coding to effectively make a new chip. It's not merely an exercise it rounding up some off-the-shelf chips/boards.

Our approach would either to try and make progress with either a FPGA or attempting some software side TBC functionality. Will have to see if phase 1 works first

[dellsam34]

What's the goal here? marketing or just a setup for your own capturing business? if it's the latter then why don't you continue on the SM03 project capture device? Having few of those should not break the bank.

[mountaincabbage]

What's the goal here? marketing or just a setup for your own capturing business? if it's the latter then why don't you continue on the SM03 project capture device? Having few of those should not break the bank.

Initially internal, at the moment we use multiple individual capture cards, which works reasonably well, but in our case we would like multiple channels and better control. To my knowledge, there are no multi channel analogue capture cards on the market, and this is beneficial for digitization companies or enthusiasts which have multiple devices. Also, developing our own board allows us to learn how it works, and customise the signal processing. The idea that we can push forwards technology in this area appeals to me. If there is a way that we can increase the number of VHS machines that can be used moving forwards will increase the longevity of our business, we have more broken VCRs than working ones now, so we also need to dig into how to repair these machines this year.

It's probably going to fail, but we can learn along the way

[dellsam34]

Have you considered the RF capture route for stand alone VCRs that work on their own with tapped RF test points? And then process the raw digital signal in computer or in FPGA environment then into computer using PCI or USB3/4.

[pandy]

To my knowledge, there are no multi channel analogue capture cards on the market,

There is plenty cards with multiple analog inputs - if they are dedicated for video then CCTV market offer lot of models - some of them are HD capable, if your goal is general ADC card then search for National Instruments ready to use solutions and what is most important to you they offer support.
Also multichannel oscilloscopes without display offering possibility to push lot of samples trough USB3 or similar interface.

Also, developing our own board allows us to learn how it works, and customise the signal processing. The idea that we can push forwards technology in this area appeals to me.

Push forward technology? Technology is gone - this niche market where most already available solutions is beyond what you need.

It's probably going to fail, but we can learn along the way

Prepare for expenditures...

[dellsam34]

As far as I know CCTV cards do not cope very well with noisy consumer video tape formats, They work well with stable live feed from analog composite cameras sensors.

[mountaincabbage]

Have you considered the RF capture route for stand alone VCRs that work on their own with tapped RF test points? And then process the raw digital signal in computer or in FPGA environment then into computer using PCI or USB3/4.

We tried the VHS-decode project out, but I believe the processing was extremely slow so was not viable for our use case. I see they are still making progress which is great. We use top end VCRs so I think the quality we get is comparable/better than with the direct RF capture. I wonder how the quality works with U-Matics, as we often have trouble with the quality of that format.

There is plenty cards with multiple analog inputs - if they are dedicated for video then CCTV market offer lot of models - some of them are HD capable, if your goal is general ADC card then search for National Instruments ready to use solutions and what is most important to you they offer support.
Also multichannel oscilloscopes without display offering possibility to push lot of samples trough USB3 or similar interface.

CCTV - no audio inputs, quality issues. Checked out National Instruments and couldn't see a card? Just to clarify this device is for digitising videotapes.

Push forward technology? Technology is gone - this niche market where most already available solutions is beyond what you need.

Just because the technology is old and niche doesn't mean that improvement can be made. There is a new set of problems to address, primarily loss of available playback equipment.

[mountaincabbage]

Have you considered the RF capture route for stand alone VCRs that work on their own with tapped RF test points? And then process the raw digital signal in computer or in FPGA environment then into computer using PCI or USB3/4.

processing was extremely slow so was not viable for our use case.

Just a further thought on this, perhaps some card developed for handing the RF signal to improve the processing side for the vhs-decode could be worth exploring

[aramkolt]

It's an interesting concept to try to tackle making a whole new video capture card, but I don't know what you'd expect to accomplish that companies that devoted 10's to 100's of millions of dollars doing the same when relevant analog DACs were more mainstream. These days, there are more advanced FPGAs (which are basically hardware emulation chips), but you'd have to have some experts on how to program those to have a chance a project like this working.

Doesn't sound like you really are wanting to do lossless capture if you're going for a "multi-channel" card since I don't think most computer can handle that amount of data via a single PCIe interface. Kind of sounds like you want to make a card that does compression in hardware and then outputs compressed streams akin to MPEG2 or MP4 that aren't particularly high bitrate.

If I was to contemplate on how to take on a project like that, I'd probably look into writing different/modern drivers for an older capture card that already does the task well - but I'm not aware of ones that do the task well that are multi-channel.

Alternatively, you could just base the design on most of the hardware from a commonly available device to then transplant key chips from them onto the new card. This could be something like analog devices chips that are contained within the Blackmagic Analog to SDI cards - they are capable of advanced features that aren't implemented within the Blackmagic products due to the lack of supporting hardware, namely additional RAM for TBC and TBC-like features such as ADLLT (adaptive line length tracking) which is essentially an advanced line TBC. However, since you are already using high end VCRs, those sorts of features may not matter to you.

The other chip you could look into would be the Theater 200 chip which can do hardware MPEG2 compression and is what the recommended the all-in wonder cards have. This would require you to figure out a way to write modern drivers for them and I don't know how you'd go about that. The chips themselves you'd have to desolder from AIW cards since they were never sold separately and aren't made anymore.

You could theoretically put as many Theater 200 chips as you need channels on one PCIe board and get your multiple interlaced high quality MPEG2 streams off of if, but again, no idea how you'd write the drivers for it.

[pandy]

There is plenty cards with multiple analog inputs - if they are dedicated for video then CCTV market offer lot of models - some of them are HD capable, if your goal is general ADC card then search for National Instruments ready to use solutions and what is most important to you they offer support.
Also multichannel oscilloscopes without display offering possibility to push lot of samples trough USB3 or similar interface.

CCTV - no audio inputs, quality issues. Checked out National Instruments and couldn't see a card? Just to clarify this device is for digitising videotapes.

Some CCTV offer auxiliary analog audio channel - audio is easier to process so you can rely on general market solutions and later only sync audio with your video (see no problem in this as even if capture done trough separate devices then it should be easy to align) - trying to capture everything at one is possible but still you need develop lot of things.

I also checked NI site and seem they indeed radically reduced available portfolio - just search for DAQ boards from different vendors - if the offer some solution capable to perform 30Megasamples or more then it should be sufficient for your needs in case of video capture.
Alternatively search for IC vendors such as Analog Devices, Texas Instruments etc, they offer fast analog digital converters IC's and provide also evaluation kits so you can check if solution suit your expectations.

Push forward technology? Technology is gone - this niche market where most already available solutions is beyond what you need.

Just because the technology is old and niche doesn't mean that improvement can be made. There is a new set of problems to address, primarily loss of available playback equipment.

What kind of improvement - i see only one possibility to improve - you sample your video with sufficient oversampling and perform all processing in software domain so you can use more refined approach than offered in HW solutions - this is not so far from RF capture proposed by dellsam34 - this is truly innovative approach as you remove your signal from uncertainty of analog processing - especially now when HW equipment has 20 or more years.
If you searching for real innovation then go for magnetic camera (so device to visualize magnetic field) - you will be independent from ancient and outdated equipment and able to recover magnetic flux also from very difficult situations.

Don't get me wrong - i'm not sure what is your goal, how deep you are decided to dive - if you happy with already existing digital demodulation solutions (such as https://www.analog.com/en/products/adv7802.html ) then i see no innovation in this.

[lordsmurf]

Push forward technology? Technology is gone - this niche market where most already available solutions is beyond what you need.

Agreed. Converting videotapes has an expiration date. The days for "new hardware" is largely behind us, especially if you're starting "from scratch" in 2025 (as these things can take literal years to perfect, and be ready for a "come to market"). Additionally, anything new will have relatively high costs (even worse now, with a tariff/trade war worldwide), so you really have to know your target market before bankrupting yourself.

ADLLT (adaptive line length tracking) which is essentially an advanced line TBC. .

This is false. Not factual. I don't recall off-hand, but I don't believe ADI even made that claim. It was merely some forum members speculating a decade ago. This myth needs to go away..

The other chip you could look into would be the Theater 200 chip which can do hardware MPEG2 compression and is what the recommended the all-in wonder cards have. This would require you to figure out a way to write modern drivers for them and I don't know how you'd go about that. The chips themselves you'd have to desolder from AIW cards since they were never sold separately and aren't made anymore.

You could theoretically put as many Theater 200 chips as you need

That's not even a theory, just a mere hypothesis.

ATI Theatre 200 is excellent, but I actually prefer LSI Logic Domino 86xx. The problem is that nobody ever used LSI well, largely because documentation for implementation sucked (as do most chips, quite frankly). And actually, I'd assume the ATI documentation is terrible or even non-existent, seeing as how it was ATI proprietary, unlike LSI which was kitted. You can see the quality of LSI in various devices, but no single device ever had all good settings at once.

But off-the-shelf chips, even extant/outdated ones, is a dead end.

We tried the VHS-decode project out, but I believe the processing was extremely slow so was not viable for our use case. I see they are still making progress which is great. We use top end VCRs so I think the quality we get is comparable/better than with the direct RF capture.

This is a common finding among serious users. I have lots of private conversations, where others have come to the same conclusion. It was a complete waste of their time, waste of their money. Most of them are "afraid" (more like unwilling to deal with the BS) at saying so publicly, because the trollish members (and the "documentarian") of that project pull out their pitchforks, libel any who disagree, etc.

I wonder how the quality works with U-Matics, as we often have trouble with the quality of that format.

They're too myopically focused on VHS (where it's not needed), and ignore formats where it could be truly beneficial (Beta, U-matic, etc).

Just a further thought on this, perhaps some card developed for handing the RF signal to improve the processing side for the vhs-decode could be worth exploring

As I've long said, the only way that project will ever succeed long-term is if it gets put into hardware "black box" of sorts. Easy kits ("dummy friendly"), easy instructions (limited scope of VCRs), real-time processing, etc. I've long said somebody will take it, vastly improve it beyond the "tinker toy"/"model rocket"/homebrew on Github, and commercialize it for a tidy profit.

Just be aware that the project kiddies will beg you to "share the code on Github", then try to threaten you if/when you refuse, then maybe even try to "force you to share" with an internet mob (DDoS, whatever). Ignore it. The clever kiddies, who want to steal the code from the "black box" will find the chips locked/encrypted.

But it will take dedicated work, and years, and maybe even $10k of initial/R&D budget. Given how tape conversion is late cycle, who is willing to take on that risk?

Even then, after all that, is it really better than a standard capture card and TBC? Probably not. The main strength would still be non-VHS formats, which is even more niche, tiny fraction of current capture needs

[dellsam34]

Just a further thought on this, perhaps some card developed for handing the RF signal to improve the processing side for the vhs-decode could be worth exploring

That's what I've been saying but they insist on keeping the processing inside the computer, Not only it makes the process faster but it relieves the CPU.

[pandy]

ADLLT (adaptive line length tracking) which is essentially an advanced line TBC. .

This is false. Not factual. I don't recall off-hand, but I don't believe ADI even made that claim. It was merely some forum members speculating a decade ago. This myth needs to go away..

I think this is statement from Frank Kearney so indeed Analog Devices claimed such thing.

The latest techniques for the correct regeneration of the missing or inaccurate synchronization information, such as advanced digital line-length tracking (ADLLT) technology, used in the ADV7188, ADV7184, and ADV7180 decoders, ensures optimized processing of a wide range of nonstandard signal types. Even for the weakest signal strengths or the poorest quality VCR sources, it is now possible to display an image on a user's TV that is both stable and viewable.

[mountaincabbage]

Agreed. Converting videotapes has an expiration date. The days for "new hardware" is largely behind us, especially if you're starting "from scratch" in 2025 (as these things can take literal years to perfect, and be ready for a "come to market"). Additionally, anything new will have relatively high costs (even worse now, with a tariff/trade war worldwide), so you really have to know your target market before bankrupting yourself.

For sure they have an expiration date, but this isn't as close as what was previously thought. I've seen claims in 2010 that there was only 10 years left, and we can still get decent transfers off 95%+ tapes today. I'd expect we still have another 10-20 years, maybe even longer, where tapes can be transferred. The demand will go down, and the condition will get worse. In that time we can look to find solutions to maintain / improve capture quality.

It's an interesting concept to try to tackle making a whole new video capture card, but I don't know what you'd expect to accomplish that companies that devoted 10's to 100's of millions of dollars doing the same when relevant analog DACs were more mainstream. These days, there are more advanced FPGAs (which are basically hardware emulation chips), but you'd have to have some experts on how to program those to have a chance a project like this working.

I'm guessing most of the dev was on the ICs, we're not trying to build those.

Some clarification on this project, initially the PCB is being made with the TW6865, as it supports 4 channels which is perfect for what we need:
https://www.renesas.com/en/products/analog-products/audio-video/video-decodersencoders...ornyp#overview

The drivers are out of date, so we'll see how much work it takes to interface with it. I should have the first prototype here with a month.

If this chip doesn't work, there is an alternative approach we can try, although I am undecided if we'll take the project that far.

Don't get me wrong - i'm not sure what is your goal, how deep you are decided to dive - if you happy with already existing digital demodulation solutions (such as https://www.analog.com/en/products/adv7802.html ) then i see no innovation in this.

Interesting, the other approach is to use a analogue devices chip, I note that has "Advanced time-base correction (TBC) with frame synchronization". The rest of the specs of that one look good also.

[dellsam34]

You could also get in touch with Ensemble Designs and see if they can make a special rig based on their chips, I believe they should have chips that went into the manufacturing of the BrightEye 75 which I believe they are still producing on order or at least have recently stopped doing so. BE75 is similar to SM03 from Singmai.

They have some multi channel cards:
https://plurainc.com/products/5350/

Edit: I've just learned that Ensemble Designs is Plura now.

[johnmeyer]

Any "expiration date" on VHS transfer is likely to be determined more by the ability to keep the players running, rather than on degradation of the tape itself. I realize that "tape shedding" did affect certain brands of tape, but AFAIK that was a problem with the binder in a specific tape formulation and is not a generic problem across all brands. I have transferred audio tape from the late 1940s, just after magnetic tape replaced wire recording, and the audio was pristine and the tape did not shed. I still occasionally play VHS tapes from the early 1980s, and they are OK.

I also have some movies from the late 1920s, just after acetate replaced nitrate, and they are in fine shape, even after almost 100 years.

I mention movie film because, while the older something gets, the more it degrades, the degradation of any specific piece of media can vary dramatically depending on countless factors. With movies, the acetate sometimes gets "vinegar syndrome" which utterly destroys the backing layer. But, when I've seen/smelled VS myself, I usually find that one can of film delivered by a client absolutely reeks, and the film is shriveled and ruined, while the can next to it -- stored under the same conditions -- is perfectly fine. (I always insert a packet of Kodak's Molecular Sieve into the good cans before I return the film in order to stop them from getting infected.)

I suspect the same thing is true of VHS tapes, and that many of them will survive for another fifty years, like my 1948 audio tape, and still be playable.

[lordsmurf]

ADLLT (adaptive line length tracking) which is essentially an advanced line TBC. .

This is false. Not factual. I don't recall off-hand, but I don't believe ADI even made that claim. It was merely some forum members speculating a decade ago. This myth needs to go away..

I think this is statement from Frank Kearney so indeed Analog Devices claimed such thing.

The latest techniques for the correct regeneration of the missing or inaccurate synchronization information, such as advanced digital line-length tracking (ADLLT) technology, used in the ADV7188, ADV7184, and ADV7180 decoders, ensures optimized processing of a wide range of nonstandard signal types. Even for the weakest signal strengths or the poorest quality VCR sources, it is now possible to display an image on a user's TV that is both stable and viewable.

That's still not really a "line TBC" (nor "frame TBC"), especially not as defined by the analog tape>digital community.

Interesting, the other approach is to use a analogue devices chip, I note that has "Advanced time-base correction (TBC) with frame synchronization". The rest of the specs of that one look good also.

All I'll tell you is this: an idiot chooses to "make his own mistakes", a wise person learns from the mistakes/experiences of others.

Multiple persons have already looked into Analog Devices chips, for a decade now, and the TBC claim is bunk.

Realize these chips exist in other devices, and have for years. Those companies turned off the so-called "TBC"/ADLLT function. Why? It made matters worse, not better. I've seen this with my own eyes, I have prototypes that were attempted around it. I knew this before the protos were built, and the results just confirmed it. No amount of tweaking anything did a damned thing.

If I wanted you to fail here, I'd encourage you to go down that dead-end path.

Any "expiration date" on VHS transfer is likely to be determined more by the ability to keep the players running, rather than on degradation of the tape itself. I realize that "tape shedding" did affect certain brands of tape, b
I suspect the same thing is true of VHS tapes, and that many of them will survive for another fifty years, like my 1948 audio tape, and still be playable.

I refer to the need or desire to convert tapes to digital, rather than longevity.

As I said, the overall "scene" of conversion has been a downward trajectory for 15+ years now. This was largely a 2000s task, heavily VHS>DVD. The 2010s was for redo, often to un-DVD the quality (better). Current conversion work is mostly procrastinators.

Tapes have a 35-65 years shelf life (something I chemically calculated in the 2000s, contrary to the 10-20 years BS at the time). We're only now seeing mass degradation (such as early 80s BASF breaking down en masse, at about the 45 mark in that 35-65 range). I'm sure many will hit 65, and some even 10-20 beyond that.

[johnmeyer]

As I said, the overall "scene" of conversion has been a downward trajectory for 15+ years now. This was largely a 2000s task, heavily VHS>DVD. The 2010s was for redo, often to un-DVD the quality (better). Current conversion work is mostly procrastinators.

We may actually see an uptick in the demand for VHS conversion for awhile. I say this because I am seeing something now that I saw 10-15 years ago with film: the parents die and the kids find these strange cassettes in a box in the basement and want to see what's on them, but don't have the equipment. That's exactly what played out with film a long time ago.

I think you are right that the first wave is dying down. That first wave was people who wanted to see their own videos "one last time" and therefore did the transfer. The second wave is what I describe above.

I got involved with the "one last time" just a few days ago with a 85-year-old widow who wanted to see her 1960s movies one last time. She wants me to transfer them but first wants to look at them using her old projector. The problem is the projector doesn't work. It is an Elmo Super 8 and the fabric belt was not just shot, it had disintegrated. I looked at it and realized that I might just have the enormous 18" belt that is needed because Roger Evans, the MovieStuff guy, gave me a lifetime supply when he refurbished my 1st-generation Workprinter 15 years ago. Sure enough, the belt fit perfectly, and I fixed her old projector in less than 20 minutes.

[aramkolt]

You could also get in touch with Ensemble Designs and see if they can make a special rig based on their chips, I believe they should have chips that went into the manufacturing of the BrightEye 75 which I believe they are still producing on order or at least have recently stopped doing so. BE75 is similar to SM03 from Singmai.

They have some multi channel cards:
https://plurainc.com/products/5350/

Edit: I've just learned that Ensemble Designs is Plura now.

I was not aware of that product line or that they changed names either. Seems it's composite only though. Big achilles heel of composite is how good the comb filter is in separating the luma/chroma as the quality to do that varies widely.

If it's composite only, then I guess that's the same spec-wise as the BE25 which is composite to SDI. The brighteye line uses Analog Devices chips for the actual ADCs, but I'm not sure what they use for the TBC part.

ADLLT meets my criteria for line TBC since it corrects horizontal line timing errors. Issue is it is only known to to be actually implemented on the TBS800 by Snell and Wilcox. See post #13 here for samples which show the actual correction it performs:
https://forum.videohelp.com/threads/402431-New-Device-Tested-for-TBC-Functionality


ADLLT info from the ADV7402 spec sheet (other 7xxx series have it as well):
The ADV7402 implements a patented adaptive-digital-line-length-tracking (ADLLT) algorithm to track varying video line lengths from sources such as a VCR. ADLLT enables the ADV7402 to track and decode poor quality video sources such as VCRs, noisy sources from tuner outputs, VCD players, and camcorders. The SDP also contains a chroma transient improvement (CTI) processor. This processor increases the edge rate on chroma transitions, resulting in a sharper video image.


It doesn't talk about resizing how long each line is, but there'd really be no point in tracking the line lengths if you then just immediately ignore how long each line is. I believe what they left out is that it linearly scales each line to be the same ending length and has each line start at the same place on the left which is what an ideal line TBC would do. That at least appears what is happening in the forum post above anyway, though it could be that some other feature besides ADLLT results in that effect. At the very least, I think it is fair to say that it regenerates horizontal sync for each line.

Hundreds of devices contain chips that are ADLLT capable, but virtually none have the additional (and always separate) RAM required to actually use it. So while the ubiquitous Blackmagic Analog to SDI converter has an ADLLT capable chip, that feature cannot be active because the Blackmagic hardware lacks the supporting RAM for that feature to work.