TBC still required for RF capture? (VHS-decode)
 

I was reading around on the forum and interwebs, of a few VHS enthusiasts using Domesday Duplicators that capture the RF signal from VHS heads directly.

They claim you don't need a TBC if you do this. They claim it's "software based TBC". They also claim you're getting sharper images.

I'm not sure about the second part, but I'm interested about the first part.

The Domesday Duplicators states that it has TBC corrections, so I'm pretty sure RF signal decoded still requires a form of TBC for capturing.

edit: (repeat question. Thread already exists)

http://www.digitalfaq.com/forum/vide...c-frame-2.html

No, not really.

The signal will require timing correction, period. Both frame and line, perhaps field. But trying to recreate TBC in software has not worked to date, and may never work. So hardware is likely still going to be required. But it'll have to be a new kind of TBC, not the standard A>D>A.

This projects has a lot of wild claims. And most of the claims come from people not developing it, nor understanding the fundamentals of digital video ingest. (Upscaling, Topaz, and "AI" are the same way. Lots of claims, most are half truths or nonsense.)

vhs-decode is not viable for actual VHS capturing yet, especially not NTSC. Anybody doing it currently will have rough results, worse than a standard workflow (using quality suggested gear). I dismiss a lot of these claims outright, because I've seen it over and over again. "VCD quality is great", etc etc. Functional, but not quality.

Personally I find that RF capture is being used, more or less as an excuse for not buying a good and proper SVHS deck. It's being used as a way to bypass the internals of a cheapo VCR. Then because you don't have line correction, people try and compensate that with software but it will never be as good as an SVHS deck.

Oh and by the way. I mentioned to the curious onlooker who was interested in KeyWest BV10s. I warned him about buying used models on the internet and directed him to DigitalFAQ, after he told me he "already spent a lot of money on the project with terrible results". I warned him against buying on FleaBay, he thanked me for the advice and I never heard from him again. Oh well...

I can honestly say spending the few grand last year on a proper setup was the best decision I ever made. I would still be tinkering with broken equipment if I hadn't. I spend a grand on the latest iPhones so I really had no excuses.

People underestimate the value of a good TBC and SVHS deck. Nothing is like the real thing.

Just because you are used to having a separate TBC and a capture device it doesn't mean that's the rule. Having one device for capturing and signal timing is not new, it was used since the early 2000's, S&W made devices that have ADC, line TBC and frame TBC all under one hood, a very sophisticated computer with FGA, proprietary OS and ADC and TBC chips, An example is the TBS-800.

What's different with VHS-decode is instead of starting from the Y-C signal, they go back all the way to the RF signal recorded on the magnetic tape and do all the processing steps, Amplification of the RF signal, decoding luma and chroma, horizontal and vertical timing, chroma timing ... etc. In theory a TBC is just a capture device.

Excuse me? :huh1: ld-decode has performed software TBC on the demodulated RF signal since the first prototype code in 2013. If it didn't, the decoded video would warble around nonstop.

ld-decode in stages

vhs-decode uses the same TBC code base, and the developers have improved upon it to better handle the crappier and more varying sync signals found on VHS (consumer recordings) as compared to LD (always professionally-mastered pressings).

This is not the same concept as attempting to perform TBC on a digital component (4:2:2) file like the "Software TBC that doesn't need the frame edges" or jmac's attempts. The *-decode projects are working from the complete signal including horizontal & vertical sync.

Perhaps that's the problem.

Sure. But those are closed-loop systems, somewhat proprietary (and non-standard to classic capture norms), and have weaknesses to certain types of source. That's the main reason I've never used such a setup. But if you can use that sort of setup, it's viable. Those were mostly made for pro settings, pro sources, not consumer analog sources.

Sort of like "six one way, half a dozen the other". vhs-decode is still capture, just in a different way. However, certain constants/non-variables must still exist. ie, TBC.

Laserdisc isn't VHS. No mention of ld-decode is made here. Many would argue that even analog capture of LD needs no TBCs, and that would be correct in theory (just not practice).

I know. But it's not anywhere near being viable. I check on it from time to time. Progress is extremely slow (snails move faster), and I have my doubts it will ever work. But still, I've always wished them well. I'll do what I can to support it (suggestions, samples, space, etc), but I'm no cheerleader.

Pfft. No. I know what's needed for this to happen. We're not seeing it yet. Lemmings that are not involved in the project, and have scant knowledge of video ingest whatsoever, parrot and regurgitate BS all the time. This mentality goes back decades. I still remember "KVCD is as good as DVD" and more recently "all you need is an HDMI adapter" and "Topaz is AI". From a technical perspective, all hogwash, easy to dismiss without further investigation. When something new is mentioned, that merits a look. Not the same old nonsense.

While that may be part of it, I think equally it's an issue of nerds playing. That's great and all, but it's a distraction from actual work, and quality work. Serious hobbyists, pros, they're really not interested in something like this right now. Not viable, not worth messing with. We want to spent our time doing video, not farting around hoping/trying to maybe do video. That's the difference. This excludes the project devs and contributors, of course.

Also never underestimate the negative economics, the stupidity, of ultra cheapness. Some people would rather throw away hours (days, weeks, months, even years), as opposed to just buying the tool needed. Often times, amusingly, the "cheap" method longterm actually cost more (money, time, or both). The simpleton idea of "any VCR works" is the clickbait for vhs-decode.

Sure, closed loop it is. But lets say you have a way to get the perfectly timed digital signal out of your TBC-1000 and feed it to the computer to be stored as it is without converting it back to S-Video, wouldn't you want that? That would make it closed loop wouldn't it? Unless you think converting the signal back to analog improves it further, If so I would like to see a technical explanation of that.

There cannot be. Staying in digital from the beginning is an added value for sure.

I think LS refers to the fact that having separate components is more flexible because you can change one if sometime the capture chain does not behaves properly (loosing quality compared to the all in one approach if it worked)

Even better would be the option to use all digital chain, or A>D>A when needed.

Improve? No.
Not make worse? Sometimes. Too often, in fact.

That's it exactly. VHS is rarely simple matter of one-size-fits-all workflow. Either that, or you have to set aside tapes that will not cooperate, and have those done elsewhere.

I can't get behind this, not in the technical argument anyway.

S-VHS players have the benefit of playing S-VHS tapes properly. There's nothing inherently magic about an S-VHS machine for playing back standard VHS tapes. S-VHS machines use the conventional S-Video connector which is very useful for capturing, but the machines don't really do much extra for 'vanilla' VHS.

Broadly, S-VHS machines are later in VHS' lifecycle and were, by virtue of being built for a more demanding market, better quality machines than many standard players, but they're no panacea.

Many S-VHS machines can restore (to a degree line-timing), they often restore sync-tips, colour bursts and neaten the non-visual portion of the line which is a worthy task, but they're often not the best at it. These machines are all ageing though, they command high values and many of them are (personal opinion) not worth the prices some try and command for them, I'm fortunate that I can purchase 2nd/3rd rate machines and repair them, but this isn't feasible for many.

vhs-decode is proven in principle, I don't think that's up for debate anymore - there are arguments for and against when it comes to the technicalities and overall result, but it is improving, whereas some of the knackered old TBCs that are commanding such huge values are ageing like fine-milk bluntly and most are getting to the stage where components need replacing; where shall we be in five years? The same with many cards and capture devices, how long do we pretend these items aren't ageing?

Where this is interesting (and as a tape head, it's a well-known phenomenon in the tape community) is the following:

1) "Old tape stock is too expensive/rare!";
2) Formulations are designed, or re-releases of old tape stock, perform brilliantly;
3) Market sits back and waits for the prices to drop on the old-tapes, as this is what people 'actually' want, pristine, wrapped 1988 MA-XGs, not modern formulations, no matter how technically superior they are;
4) Gate-keepers do a round of 'well poisoning', 'you shouldn't look to new formulations', 'they make the shells in China y'know', If you truly cared about your recordings, you'd be happy to spend $100 on a That's MR-X Pro' and a hundred others.
4) Nothing happens for six months - nobody buys or sells anything whilst everybody watches on expectantly;
5) Lack of interest, new stock is withdrawn/deeply discounted - sells out in minutes at 75% reduction;
6) Gatekeepers holding the old stock sell again, with a +20% on the previous prices.

Rinse and repeat.

I've seen a few cycles of this, it's almost now a trope in the tape community. I think anybody who hangs around a few analogue forums will testify.

Now there's no distinct parallel between this and vhs-decode, after all, vhs-decode is not a tangible product - but I'm just watching a few reactions and I can't help but wonder if there are a few, very faint shoots of this springing up over VHS-decode? It could be my eyesight, however.

vhs-decode will work, there's a long way to go to make it a feasible exercise for general hobbyists but arguments about things like the concept of a software TBC are settled now, are they not?

One more advantage of using S-VHS machine for VHS tapes is avoiding the extra composite stage, It keeps the luma slightly sharper, chroma cleaner and avoids the chroma/luma cross-talk. It would not make a VHS tape look like a S-VHS tape but the difference is noticeable to the average viewer in a side by side comparison.

You're right, and I didn't consider that in my reply - which I should have done.

It's still a heterodyning stage though, so whilst it is a (great) improvement, it's not flawless. Only U-Matic from memory can do true Colour Under ->Colour Under throughput but we're wandering into the weeds at that point. Please don't rush out to play with UMatic! Colour was very fragile on U-Matic anyway, hence it was probably a requirement rather than a convenience.

You're right to clarify that though.

So what?

"trying to recreate TBC in software has not worked to date, and may never work"

This quote is objectively false, because it has already been done. Your statement said nothing about VHS, so a software TBC that corrects LD already proves it false. But again, this same TBC code has been extended to work better for VHS sources. So it's doubly false.

I've already found an example tape where Panasonic's line TBC fails on a particular shot while vhs-decode's software TBC succeeds.
Attachment 15098
Attachment 15099

Every LaserDisc player ever produced includes a TBC. It's literally a requirement to play back the format.

https://youtu.be/zj8RE1EV_Q4?t=1878

I need to get some images up of some of my experiments, this view is also quite useful for clearly visualising line distortion between conventional TBCs as well which I have used it for.

I've found that the vhs-decode software TBC is more consistent than some hardware, it's also able to rectify some issues that other equipment could not manage, or at least greatly improve some very wonky tapes.

Arguing that it 'doesn't work' is being a bit delusional now. Sure there are arguments regarding practicality, user-friendliness, things it doesn't do so well and no doubt a host of other obfuscations, but the argument, in essence, that it 'doesn't work' is just denial at this point.

This can be improved upon at least, unlike [insert brand X] hardware unit.

... And one of the main reasons why it is not ready for prime time yet is that there is so many variables in extracting an analog video signal from a tape that constant tweaking never ends, not to mention things can change from a tape to another, One cannot deny that even with conventional VCR's, playback compatibility is not consistent across multiple tapes, that's why multiple VCR models are preferred for maximum tape compatibility. However as of right now you get more tape compatibility using one VCR than using VHS-decode with one set of settings.

<sigh>

It's only "magic" when people don't understand what's going on. As mentioned, a primary reason is non compositing the output. Another is that the decks are simply made better, which lends itself to things like less timing errors. It's very much a situation of "you get what you pay for", with the cheaper/worse option having more problems, less quality. Often many more problems, substantially less quality -- not a nuanced amount.

Don't confuse newbies here. The decks are not suddenly higher in cost. In reality, most are less than MSRP. Some at. Refurb'd units have the added costs of the refurb parts/supplies/time. And none of those numbers includes inflation of 15-25 years, namely that many AV items have doubled in price in that time, often with a lower quality. So in reality, gear now is a bargain compared to then.

That's been my stance from the beginning.
- Does it work? Yes.
- Does it work well? No. That may change, or not. This isn't an easy task, and has some ugly challenges. And the tern "challenges" doesn't mean overcoming it is inevitable. Sometimes challenges are insurmountable (something I can attest to, in terms of health). In those cases, where you cannot overcome, alternatives must be used. And we already have the capture alternative, and have for many years.

Again:
- I do support it.
- But I'm not a fanboy or cheerleader about it. I'm a realist, not a blind optimist.

In terms of video capture hardware, we'll be fine. (I'd be more worried about geopolitics at this point in time.) The need to capture will of course be dimiinshed.

Capture cards isn't as big of an issue. In fact, not really a worry at all. The bigger issue is OS, and stubborn users refusing to use legacy OS (not merely latest-and-"greatest" OS), Newer is rarely better with video capture, newer is often just newer (and more often worse).

I can't buy that. Why? 20% isn't worth the buy of stock. From fundamentals alone, I'd suggest there's merit to the argument of avoiding new tape stock. And any tape bought on a dip is finally fairly valued. Or did you mean sold at 120%, with it bought at -75% (aka, 95% gap). But even then .... eh.

Well, no, this is misleading. Yes, software TBC has been possible for at least 10 years now. But the problem is it only functions well on a per-tape basis. It's not a general use button on a deck, press it, and it works. Software TBC has always needed lots of tweaks, and even then it can still be inferior. I've seen many cherry-picked samples, and am actually thrilled that a rare video of mine was made better enough that I could use the "beat it to death with temporal NR" that it became viably viewable again.

I'm glad you found a scenario where a Panasonic TBC was outperformed. But that's always been possible. After all, it's why most of us own multiple decks, multiple devices. JVC, ES10, etc. That's not unusual, it's just par for the course.

No. You can get away with a lots of inefficiencies when the TBC is weak, or outright a "TBC" (in name only). So a TBC (or "TBC") for LD would fully flop with VHS sources. I've always laser focused on TBC performance with consumer analog sources. Not broadcast sources, not video games, not LD, etc. Those other needs tend to be far less demanding, often requiring only the most base of corrections.

That's it. :congrats:

Time spent tweaking, sometimes (often?) with same or worse results.
Time ROI is negative.
Money ROI is arguably negative.

Exercise for the reader: compare and contrast the two bolded statements.

Exactly how many hours of footage have you fed through vhs-decode's software TBC? From what you've said, it's exactly zero. Why are you now pretending that you have experience with this?

I think he means by "it doesn't work" it is not a reliable alternative for the average user, Yes some samples look rock solid but for the most part there is a lot of work to be done based on the samples I've seen over at videohelp and youtube, This is one of the reasons I haven't jumped on it yet, a lot of inconsistencies and trial and error. As of right now I can get better results using a high end S-VHS VCR and a studio capture device/TBC with digital out.

In my opinion for this project to really have some fruits the team should think outside the box and start considering building their own RF preamp bypassing the VCR's own, as well as the motors servo control to completely have control over the signal, That way they don't have to worry about what model VCR is and what signal level of the RF is, Also having control of the capstan and head drum motors means with one VCR you can have more tape compatibility in terms of video standard (PAL, SECAM, NTSC), The recording speed (SP, LP, EP ..), shrunk or stretched tapes speed compensation, prediction and replacement of damaged control track sync pulses for better tracking...etc.

Those are not contradicting statements.

Software TBC "works" in the sense that it can function in a very limited way (many tweaks needed), often as "proof of concept" only. Software TBC "doesn't work" (isn't viable) in the sense that it's not something that can simply be enabled and disabled, and used in the traditional sense.

That's exactly it.

Agree. :congrats:

You and I, and others, have discussed this somewhat before. This sort of project will eventually require dedicated hardware to ever get anywhere. Random VCRs, trying to software code (inefficiently compared to hardware), etc -- that's really a road to nowhere. But therein lies the problem.

As a geeky freebie open-source projects, it's nifty and all. But to be viable, investment will be needed. In a recent (past months) TBC thread here, investment was part of that discussion. A box of old random parts, and some "duct tape and chicken wire" programming won't cut it (the ill proposed TBC code from the TBC thread, not vhs-decode).

Furthermore, a project like this won't be for John Q., nor cheapskate DIYers. At the current dev rate, average users will have long ago converted videos; many already have. So the target marker will be more serious, more pro, and demand something that has both investment and time ROI. Endlessly fiddling and tweaking is poor time ROI, which drives up end customer/user costs to unsustainable levels. Money/capex to the pro/serious capturer/ingester is less of a concern, as we all expect hardware to cost $$$$. So avoidance of hardware is really quite silly from a longterm macro view.

Remember, video was my accidental career, due to serious hobby, quality of my work. My prior career path was both financial and technical, so I don't BS easy with some things. I'm not an artsy-fartsy type that gets lost in fantasy, but am far more grounded in reality. While I'd like to see vhs-decode succeed, I'm not seeing it currently. I'd love to see an "outside the box" hardware complement. Both NTSC and PAL, noting that the PAL usage is still more advanced.

Again, I can support vhs-decode, but I won't be a cheerleader that overlooks current limitations and flaws, and that includes the project trajectory and viability. The devs are trying, I know, but sometimes it's just not enough (something I also know).

Out of interest, what is your professional role in video? I've seen you allude to it a few times but some of your speculation about videotapes doesn't add up to somebody with a tremendous amount of experience with professional videotape, only VHS?

I've seen you say "my studio days" a few times, I'm just intrigued as to what you actually did? I've found a few vestigial websites where you sold material, but it's not clear what your professional video work is?