Decoding RF signals vs. traditionally sound workflows?

[lilsweepy]

As a concept, I think that VHS Decode is a very exciting project that could one day reduce the need for increasingly rare equipment. And yet, I’m not the first to point out that some of the claims surrounding this process have been hyperbolic, to say the least. That being said, learning more about RF workflows by way of VHS Decode has left me with some hanging questions.

There seems to be this argument that decoding an RF Signal is fundamentally better than traditional preservation workflows. That relying on a player to decode an image is automatically a lossy process. I can definitely understand this being the case for cheap consumer VCRs, or players that can overcorrect an image, but this is admittedly where my knowledge gap begins. I’ve seen many posts, on this forum included, that suggest working from an RF signal could yield a sharper image than traditional flows. But why exactly is this the case? Is this judged on a player by player basis, or is this a uniform truth?

It’s been difficult tracking down info as so many posts I’ve seen compare VHS Decode to extremely rudimentary setups for their use cases. Or they are penned by those new to analog to digital workflows haphazardly applying concepts. I will admit that results I’ve seen from VHS Decode can be quite amazing, and in pointing out an overly enthusiastic fanbase is not my attempt to downplay the project. But having a previous career preserving content for public broadcast stations I’ve encountered the same (or better) results from a traditionally sound workflow. Are these not two means to the same end? Or is it true that a good workflow is always “lossy” and decoding an RF signal yields a fundamentally better result? And lastly, if one process has the potential to outweigh the other, are we just chasing pedantic improvements or something actually meaningful?

[lordsmurf]

Quote:

Originally Posted by lilsweepy

As a concept, I think that VHS Decode is a very exciting project that could one day reduce the need for increasingly rare equipment.

That has also always been my contention. But I'd go one further, and suggest using the FM/RF method for mere VHS is pretty stupid. Of all the analog tape formats out there, VHS literally needs it the least. There are many obscure formats with no extraction, and many non-rare formats like U-matic and Beta(max) that never really got decent hardware. S-VHS really was the height of gear for any analog format, save maybe BetacamSP. It's such a lost opportunity here.

Quote:

There seems to be this argument that decoding an RF Signal is fundamentally better than traditional preservation workflows.

It's not whatsoever. In fact, there really is no singular "traditional" workflow, aside from the very basic formula of VCR > TBC > capture card. vhs-decode is identical here, merely swapping capture and TBC to VCR > capture > TBC (and DOC/etc). Though is need be noted that this "software TBC" of sorts does not functino nearly as accurately as hardware TBCs. Lots of workflows have modular orders of operation, swapping or even merging tasks. But the basic needs cannot be avoided. It's not better, just different, and still worse.

Quote:

That relying on a player to decode an image is automatically a lossy process.

It's just false. vhs-decode uses a player too, for at least the heads.

Quote:

I can definitely understand this being the case for cheap consumer VCRs, or players that can overcorrect an image, but this is admittedly where my knowledge gap begins. I’ve seen many posts, on this forum included, that suggest working from an RF signal could yield a sharper image than traditional flows. But why exactly is this the case? Is this judged on a player by player basis, or is this a uniform truth?

Most of the "it can be sharper" is actually traced back to my discussions of it long ago, in a theoretical sense more than practical application. In hypothesis (maybe note even a theory), it can be a tiny % sharper than the best VCRs, due to processing. I do not refer to the fake Youtube BS that shows a blurry thrift store VCR, compared to vhs-decode (and noting vhs-decode isn't great there either). I refer to quality equipment, the various brands/models of suggested S-VHS VCRs. Most Youtube comparisons are actually Y/C vs. composite, or bad VCR vs. not-terrible/decent/good VCR, nothing more.

But getting that sharpness will come with tradeoffs, meaning you'll lose quality in other areas, at least as the project exists now. Consider something as "simple" as DOC (dropout compensation), which is generally handled well in VCRs, but not at all well by vhs-decode. This is simply the nature of working with analog video. It's often about trade-offs, and perfection is a fool's errand.

Quote:

It’s been difficult tracking down info as so many posts I’ve seen compare VHS Decode to extremely rudimentary setups for their use cases. Or they are penned by those new to analog to digital workflows haphazardly applying concepts.

That is entirely correct.

Quote:

I will admit that results I’ve seen from VHS Decode can be quite amazing, and in pointing out an overly enthusiastic fanbase is not my attempt to downplay the project. But having a previous career preserving content for public broadcast stations I’ve encountered the same (or better) results from a traditionally sound workflow.

That's the difference. Newbies are fooled by bad comparisons. People who know better realize both of them suck. One sucks less, but it's still a level of suck. I'm not even referring to the image quality here, but the level of needless work (and drive space) to re-create the wheel for what already exists. It's a solution looking for a problem. It's not at all like ld-decode was, and how other -decodes (other formats) could be.

Quote:

Are these not two means to the same end?

Yes -- but not "the same", it's still very beta. I've been around video for decades, and am not fueled by hopium. This project could gain in future years, or fail entirely. It's not proven itself as a viable reliable method, merely one of tinkering with mixed results.

Quote:

Or is it true that a good workflow is always “lossy”

No.

Quote:

and decoding an RF signal yields a fundamentally better result?

No.

Quote:

And lastly, if one process has the potential to outweigh the other, are we just chasing pedantic improvements or something actually meaningful?

In terms of the VHS/S-VHS format (and likely Video8/Hi8 format, eventually), realizing that an actual 1:1 comparision may yield tiny % sharpness gains, you're looking at pedantic, asinine, OCD, placebo. It is there, but you enter the insane realm of the self-described "videophile", who will spend countless time and money for a % of a 1% gain.

I'm a pragmatist, I'm The Price Is Right. I want the best quality possible, but without going over(board). Some of the stuff even discussed at this site makes me shake my head at times, such as breaking out a histogram for every single crappy homemade camcorder tape. You've lost sight of the actual project. You enjoy playing with toys, not doing work. That's largely the realm of vhs-decode, and probably always will be. It's hard enough convincing a person tat he/she needs quality gear to not make their conversions look dreadful (worse than the VHS tapes were). It's another to tell them they have to tear apart a VCR, learn scripting, and buy dozens more hard drives to store unviewable data.

I'm sure some small % people are getting some % of results, but that's really it. I've seen very few videos that were truly interesting, and much of it was likely due to other variables.

To be a bit crass and blunt here...
Certain folks online get all pissy when I don't bless their method as bestest ever, but they also need to grow the hell up. It's a tool/method to convert a video, it's not their real-life child. It doesn't need their mama bear protection, attacking any who do not agree it's the most beautiful perfect thing ever. vhs-decode's farts smell like ass too. All of the methods fart, each can be quite stinky. Some more than others.

Does that answer it for you?

[lilsweepy]

Quote:

In fact, there really is no singular "traditional" workflow, aside from the very basic formula of VCR > TBC > capture card. vhs-decode is identical here, merely swapping capture and TBC to VCR > capture > TBC (and DOC/etc).

This I think pinpoints the biggest misconception those new to digitization generally have about the process. When I first learned of this approach I saw so many posts claiming RF/VHS Decode was the purest image you could ever achieve. I said I thought this was cool! But keep in mind your offloading processes onto software we can achieve today with equipment. I got a lot of pushback to the point of doubting my understanding, which prompted this post.

I had first come across RF decode within amateur archivist circles on Twitter (which is also a source of great disinformation). In my previous career as an archivist I did some outreach and education on AV preservation workflows so am always excited when people get interested in the process. But when I start to hear blatant misinformation in regards to preserving a rare or one of a kind tape, I can’t help but be frustrated.

Practicality is another thing, as you said. I heard the RF decoding process adds hours of time. If that’s true I’d hesitate to recommend this approach to any institution. They’re already racing against time to begin with.

Thank you so much for your input! This is my first post after being a longtime visitor so I highly appreciate the detailed response. Again I wanted this post to clear up some of my own misconceptions and I guess to highlight some frustrations. I’m a big user of open source solutions so any new addition is cool to me. It’s the info surrounding those tools that can lead to issues.

[latreche34]

I think too many members misled by non fair comparisons, where you take a composite capture using easycap and compare it to an RF pickup that bypasses composite, easycap and lossy mp4.

VHS is noisy, that's just the nature of the beast, but if we put the noise aside and compare a proper capture of say a clean S-VHS feed like this one, tell me what does VHS-decode brings beyond this? My point is VHSdecode is as close as possible to the RF recorded on tape, no doubt about that, but it can be achieved as well using proper equipment and bypassing a time consuming step of decoding that hasn't been finalized yet and takes a lot of time and memory storage for the RF files. Keep in mind the video in the link is not the raw capture, let that sink in.

[latreche34]

Sorry for dusting off this old thread, Here is another misleading comparison using composite vs vhsdecode.

[Aya_Rei]

Yeah looked at the VCR model, sure seems to be a standard composite only consumer VCR. I bet if they did a comparison with an S-VHS VCR, it'd be a beaten up one to make even more misleading comparisons, that they still use on the wiki by the way. (Actually read Harry's comment, scratch that.. guess proper S-VHS comparisons will most likely never happen, so decode will forever have unfair comparisons until somebody else steps in)

I personally feel like this boils down to sharpness and detail gain, but you want to know what else can help with that besides using an S-VHS VCR? Using EDIT mode! Not saying that to sound smart, but moreso because I feel like decode just ain't worth the hassle. As I've said, is it better than a crappy composite consumer VCR? Yeah it is, but so is a refurbished S-VHS VCR.

At this point I can see how these tapes are suppose to look (unless it's clearly a copy, then the quality gets all muddy) so this stuff ain't a some dealbreaker for me. My night and day moment was using a recommended workflow.

[Gary34]

This video is another bad comparison. https://youtu.be/xOq7BRDHPBs?si=HQOcAweLSeHQRuD1 At 6:06 in this video https://youtu.be/foLPX4YTwHk?si=QX1LNIADnxN9pO5A he says he got that for 400 dollars on eBay. He thinks it’s fine because it’s playing but that 1980 hasn’t been recapped. He says around 13:00 that he uses OBS for his captures.

[Aya_Rei]

Think that Linus video isn't about VHS-Decode specifically but it does show off bad means of doing the job, do remember a thread on the video when it came out.

Clickbait-y YouTube at its finest, same with that TechLinked video (Fun fact, both channels are own by the same company)

Seriously, who is decode even for anyway? Archivists and Government institutions/film companies? No, archivists already have their own methods of doing what they do. The general population? No, most people would just want a digital copy of their media, no matter the quality. So this barrier to entry would be too high for them. DIY hobbyist people like us? No we already use recommended equipment that does what we need it to do and, when all is working, a dang fine job. People who are cheap? No, the time and money spent trying to set that up won't be worth it.

It's like, this pleases nobody. Everybody who this could be for already has his or her ways of digitizing tapes. For people like us who use recommended equipment, the benefits, if there are any, are just not worth the drawbacks and time spent trying to get it all set up, same with transfer services who use equal quality equipment as us, they already have their own methods of doing what they do, and if it works dang well for them, and has been working for them for years, then why change now?

Of course if you want to take a crack at using decode in order to experiment and make some proper comparison tests for example, then go for it! But I feel like it'd be impossible to convince most people that this will be the only way to go moving forward.

Capture now and worry about decode later? Sounds like a big joke, it's kind of the same as doing a traditional capture and doing restoration work on it, and then deciding the redo the restoration process again the future (because you might've forgot to use some filters, change some settings, make the audio restoration sound better than your previous attempt, etc)

Really feels like this came out all too late in my opinion... we already have the traditional method that works great, and have been for 20+ years.

[latreche34]

It's a small Discord/GitHub closed community who like to tinker with codes, I don't see anyone outside of that circle using it.

[lordsmurf]

Quote:

Originally Posted by Aya_Rei

Seriously, who is decode even for anyway?

• Archivists and Government institutions/film companies? No, archivists already have their own methods of doing what they do.
• The general population? No, most people would just want a digital copy of their media, no matter the quality. So this barrier to entry would be too high for them.
• DIY hobbyist people like us? No we already use recommended equipment that does what we need it to do and, when all is working, a dang fine job.
• People who are cheap? No, the time and money spent trying to set that up won't be worth it.

It's like, this pleases nobody.

Fixed the formatting in the quote.

I see two main user bases.

1. "Because I can" -- ie, the smart-ass users, often young, and generally low-knowledge at video.

2. Model rocket builders. To them, it's just fun to take something apart, screw around with it, see how it works, use it in ways never intended. They're not worried about comparative quality, or ROI, or anything else. It's just good wholesome geek fun. (I understand this crowd. I did that myself many times -- but mostly when I was younger, had time to blow/waste. Not VCRs, but other tech, other tools.)

It's not for serious use, too many issues.

Quote:

Really feels like this came out all too late in my opinion...

That's it entirely. It's a solution to a problem that hasn't existed in 30 years (1990s and earlier). The days before TBCs existed, before quality capture cards existed. I can look back and think "gee, I wish that existed back then". But it didn't. And what we eventually got was better anyway, and still is.

[SHGMC_2]

I think people lose the real goal of VHS-Decode in the long run.

I think VHS-Decode is a great project. I see this as a way to adapt to the times where a lot of the hardware required to do great captures become rarer and rarer but other solutions such as RetroTINK (probably the best in its class and probably the only one I'd recommand) brings the possibility to bring a decent capture with modern capture hardware.

I am lucky to have a ATI TV Wonder 600 and datavideo TBC-3000 TBCs to work with. We also have the ATI AGP cards (which I hate working with, but I'll spare the details) but getting systems with those will get rarer and rarer. I know that, if any of these systems fail, my boss will not search to replace them but will rather invest in current-tech solutions and, unfortunately, comes with drawbacks such as making keeping the de-interlacing quite difficult for those that want it. However, VHS-Decode is a way to keep the original interlacing in a world where good (S-)VHS hardware become increasingly difficult to come across. Not to mention other formats such as EIAJ-1 (aka the first format with video on tape), U-Matic, Betamax/cam, etc.

My 2¢.

[Gary34]

Nicholas Serra’s comment here https://www.digitalfaq.com/forum/vid...p-4-print.html and here https://www.digitalfaq.com/forum/vid...hs-decode.html made me wonder how tough it is to get working.

[lordsmurf]

Quote:

Originally Posted by SHGMC_2

I think people lose the real goal of VHS-Decode in the long run.

I'd suggest using it for "VHS" is losing sight of it's potential benefits. Because, as you state, and as I've stated before many times for years...

Quote:

Not to mention other formats such as EIAJ-1 (aka the first format with video on tape), U-Matic, Betamax/cam, etc.

This has always been the "elephant in the room". The main vhs-decode crowd cries about costs (yet ironically spend like "drunken sailors" on the excessive SSD/HDD storage needed), and non-VHS decks have costs. Betamax has especially become pricey in recent years, and those decks are largely junk even when "good". So this project would greatly benefit from it. And yet, it's mostly overlooked. Missed opportunity. It's not mature, not even alpha-grade decoding.

Always keep this in mind: this project is about costs, not quality. In order for quality to take the lead, like-minded archivists will need to get together, to truly make RF/FM viable for these non-VHS sources. And for their own sake, they need to fork it, in order to cut away the immaturity/antics of the low-end VHS users.

Had the tech/project existed 15-20 years ago, I probably could have orchestrated it. I knew so many non-VHS users back then, from hobby to pro. But many are now retired, have passed, or simply lost interest over the years/decades. I'm sure some are still out there, but they won't be found acting immature/unprofessional in the Reddit/Discord sites. Connecting with them will be the issue.

Quote:

I see this as a way to adapt to the times where a lot of the hardware required to do great captures become rarer and rarer

This isn't a concern yet, and may never be. The VCRs are degrading faster than TBCs are becoming extinct (mostly due to abuse and stupidity), and decent/good capture cards are somewhat plentiful. (But noting newbies, even seasoned users, need help to acquire those good cards. It's not sold new on Amazon, and eBay is not reliable. Too many variables.)

The FM/RF/-decode methods all still require the VCRs. Cheap junk VCRs still result in bad decoding (hence why so many bad examples exist), so quality decks are still needed. It cannot be avoided.

Quote:

I am lucky to have a ATI TV Wonder 600 and datavideo TBC-3000 TBCs to work with. We also have the ATI AGP cards (which I hate working with, but I'll spare the details) but getting systems with those will get rarer and rarer. I know that, if any of these systems fail, my boss will not search to replace them but will rather invest in current-tech solutions and, unfortunately, comes with drawbacks such as making keeping the de-interlacing quite difficult for those that want it.

Current tech doesn't care about arcane formats. That's always been the issue.

If you ever need tips on that ATI AGP, I'll do my best to help you.

I need to re-work one of my own AGP ATI systems, so perhaps I'll make a new guide for it. Your input would be considered and addressed in it. I want to update everything ATI.

Quote:

Originally Posted by Gary34

Nicholas Serra’s comment here ... made me wonder how tough it is to get working.

The main problem is that the goal posts always move.

As I wrote on VH a couple months ago:

Quote:

Originally Posted by lordsmurf

I actually had a AG-1970 that I was going to use for it, but it bricked while stored.
... but it didn't work anyway.
... oh, but now it does, if you do this.
... oh wait, no, that's bad now, do this instead.
... nevermind, this is needed now.
... oh hey, try this instead.
... etc

By contrast, for VHS/S-VHS/Hi8/Video8, you can buy:
- recommended JVC/Panasonic S-VHS VCR with line TBC (or Hi8 camera)
- recommended frame TBC
- recommended ATI/Pinnacle type capture card

... and done.

- No more "fixing it later" or "next update". Nothing more to screw around with.
- Nothing more to buy.
- Buy the gear, use the gear -- and later resell it (if needed/wanted), as it holds value.
- The end.
- Done.

You can stray from the recommended/suggested "VHS conversion recipe", but it has consequences. And yet, even then, you're largely done, and it can still often out-perform the RF/FM/-decode method.

Setting up these RF/FM workflows is an endless "rabbit hole" within the video capture "rabbit hole". Sort of like the movie Inception, where you start dreaming within a dream. It's a murky mess, easy to get lost -- and never actually get anything accomplished.