Current status of ld-decode / vhs-decode (true "backup" of RF signals)

[lordsmurf]

Me, I'm capturing my tapes very slowly using both methods, traditionally and with vhs-decode, and to be honest, I find vhs-decode better.
I won't enter a debate with anybody from digitalfaq about how wrong I am

I'm not sure who would say your method is "wrong" here. Given how tapes are now entering a self-destructive phase (2000s fearmongering is 2020s reality), and how variable vhs-decode performance can be, your method seems prudent to me. (1) Get the known quality capture, using the known quality gear. (2) After the safe capture is made, then play around with vhs-decode, in an effort to squeeze out that last little % of quality (most of which is sharpness). I definitely approve of that.

It just takes a lot of space and time to do a tape, but it's so worth it.

That's the tradeoff. Capturing has many tradeoffs, be it using budget gear, MPEG capture, reliability of the method, or whatever it is. vhs-decode is an optional under certain conditions. I think the pushback happens when overly eager devs/fans of the project get "too big for their britches" and suggest it's something that could be used in all cases, or for general use. That quickly gets hyperbolic, unrealistic, sometimes outright false.

Here's an old comparison I did between conventional capture and vhs-decode:

That "sample'" is completely invalid, bad science. What models were used for this test? What settings for those decks? I could make a very comparable comparison between a POS deck from a thrift store, and the AG-1980P with optimal settings for sharpness. And it's look similar, an "impressive" different that lacks context.

The best option will always be FM RF capture today

I have zero doubt that, someday, the FM method can be leveraged to outperform traditional methods. But it's not true today. And it may never be true for this vhs-decode project, given how "cheap" is stubbornly a primary concern (and appliances to add much-needed TBC accuracy is shunned).

you need to promote the best preservation today, not the best "picture"

That makes no sense. Preservation is about extracting the best video ("picture") and audio possible. Simply extracting random raw data may longterm be a waste if it cannot be processed later.

and a decent VCR

This is another main problem. The term "decent" is overused to describe everything from consumer crap, to not-really-great ancient "pro" decks, to the decks that excel at stability without raping the image (no over-sharpness, over-contrast, bad AGC, etc) as most decks sadly do. But even those best decks have caveats, and is why serious hobbyists and pros have multiple decks, or multiple brand and model.

I think everyone at DigitalFAQ likes to ignore facts

Not at all. The issue is that we don't get all googly-eyed at something new and shiny, nor take hyperbole at face value. For starters, vhs-decode is not FM. The FM method is theoretically sound. The problem is the cheapskate approach by vhs-decode, which makes it vastly more difficult, variable/unknown in performance, and just overall a tiny niche within the already niche of video ingest. It neither appeals to high-end desires (quality with time/funding ROI), nor low-end desires (easy).

and label this as a "product" in development when it's not a product it's a collection of tools brought together and built up by each other within one community

Yeah, that's open-source. And open-source is literally always in development, until bought or ceases. Given the total lack of organization of vhs-decode, "bought" isn't going to happen. That leaves cease. Or more likely copied, vastly improved on, then sold because it's actually viable beyond a tiny cheapskate niche. Meaning your contribution will be forgotten, not even a footnote in the longer term story. Anybody that doesn't see this is somebody that hasn't paid attention to tech for the past multiple decades. That's always how the story plays out.

AI-assisted tools could replace the current software

"AI" is an overused blind-faith concept, as it is here.

To consumers its the most cost effective entry

No.

to restoration houses and national archives

No.

FM RF is an addon method to existing workflows

Yes,

not some obscure black box fixed product

Not being an appliance, or kit, or similar, is part of the problem. No QC, no ROI.

I'm not implying that this doesn't have potential,

Same here.
FM is theoretically sound, practically accessible, but not practically reliable.
That's not vhs-decode. But it could be, if it releases that death-grip on the desire to want to be an Easycap replacement, using random VCRs from the junk store.

I've even offered financial help to the project,

This is also part of the issue. Some people have time, some have skill, some have money, some have a bit of each (but still predominantly in one camp). The vhs-decode project has too many libertarian-type personalities, which shun any assistance from "the other", and get pissy when you suggest methods to improve (outside-their-box thinking). Especially pissy when you call BS on the over-the-top rhetoric that reminds me of Taco Bell in Demolition Man, or even Tesla fanboys. No, Taco Bell will not be all restaurants, the roads will not be just Teslas, and all capturing will not be done via vhs-decode. You cannot have an intelligent discussion when cheapness, stubbornness, and outright divorce from reality dominates the conversation.

The only dev that talks sensibly is oln/hodgey.

I just have to be honest with newbies who think this is a cheap way of getting tapes digitized, They must know that this is not a plug and play method, requires knowledge in both software and hardware and requires a lot of tweaking, When it's ready for prime time I will be the first one promoting it.
Not to mention that you will have to modify every tape machine for every format, Not everyone has that knowledge, Not sure who you are adressing here, restorations firms, sure, I'm addressing the average person who comes here looking for solutions after facing problems with cheap chinese devices, those folks my friend cannot use VHS-decode.

^ Quoted because accurate.

Easy craps are the market this project is slowly helping kill

About as slowly as a lethargic snail on downers.
Easycap doesn't even know/care that you exist, and the lower end market has zero interest in the complicated method.

but to do so requires education and dedication

Again, this is a main problem. Those who are educated, know enough to see that it's not any sort of workflow replacement. It's something very niche, for certain projects that are willing to pay for the level of work needed. You really fail at the ROI aspect of all this. It's just not there. Realize that the "I" if not just money, but time. Time is so important that "time is money". Video is already irritatingly long to process, yet this "advancement" makes it slower.

That doesn't change the fact that it is not the average person's tool, you can argue all you want but as it is right now, it is not a plug and play device like the easycap, and as it stands it doesn't make a dent in easycrap sales, When it becomes a marketable box with video in digital out with a driver on Windows 10/11 we can come back to this discussion.

It's not marketable. There is literally no main audience for it. The top and bottom ends are excluded, so what's left? Some cheapskates with lots of free time? No business model, no targeted demographic, just randomness. You've heard the phrase "a solution seeking a problem". Will here it is. FM (and potentially vhs-decode) are great theoretical solutions to certain niche problems. But no such mass problem exists, which is what this solution tries to fix. That lends itself to wrongly seeking cheap for the masses, rather than targeting the actual niche problems facing archivists, restorationists, etc.

However it will never be wide spread like easycap

We need to stop discussing Easycap. There are so many more cards available, some arguably give better results that vhs-decode with image values (excluding the ultimate sharpness from FM methods).

Moded VCR's will be certainly available online with an added price tag and depends on the demand vs supply prices could excede high end S-VHS VCR's with line TBC.

Ah, yes, good call. That's where money will be made, as suggested above. And likely improved on, which is when it will get more reliable. But as stated, for much higher costs that currently available workflows, with suggested VCR/TBC/card.

I know people looking at it for 2" and 1" capture where modern demodulation, timebase correction and decoding techniques could outperform the integrated stuff, and for older consumer formats like EIAJ and Skip Field that need a lot of TLC to make them watchable.

Excellent use case. And yet, vhs-decode still doesn't fit here. But FM methods do.

[VideoMem]

I see great success here.

It's niche, yes, as it is GnuRadio and SDR in general.
Is its public technically inclined, talented and skilled?, yes.
Will it be a commercial product by itself?,
Maybe, in the realm of lost radioshack's kits memories.

Tape decoding in general will not be a commercial product anyway.
We're talking about signal study and improvement exploration here, that's timeless.

Vhs decode/ld decode is made by contributions of people of different countries, speaking different languages with no other interest than improve our understanding of the field.
Another part on where it is successful is in its relation complexity/contributions.
This project has so valuable skilled people working improving them only because we think the same, and we tend to resolve the problems from input to output like any other data ingest.

At cbvs there's lot of filtering already applied.
It's lossy at that point, that's a fact.
Could it be enough for the sake of practicity? Sure
But if you want to recover, understand, adapt and and improve (even with low budget) these are very good tools to start.

And in the process you will learn a lot of things.

[dellsam34]

Well said VideoMem.

[lordsmurf]

It's niche, yes, as it is GnuRadio and SDR in general..

I don't think that's an accurate comparison, GUI software for the masses.

lost radioshack's kits memories.

I can definitely see it as this, 1970s style hardware hacking. Most of that stuff was a dead end, tinkering for the thrill of the tinkering. Not as much interested in the actual quality of the end product of the tinker toys.
Anybody else remember these? https://archive.org/details/byte-magazine

We're talking about signal study and improvement exploration here

That part is nifty.
Insisting it can replace existing methods, using any cheap random VCR, is the BS.

Vhs decode/ld decode is made by contributions of people of different countries, speaking different languages with no other interest than improve our understanding of the field.
Another part on where it is successful is in its relation complexity/contributions.
This project has so valuable skilled people working improving them only because we think the same, and we tend to resolve the problems from input to output like any other data ingest.

ie, open source.

And in the process you will learn a lot of things

In the 2000s, I spent a ton of time pouring over analog signal theory text books. It is interesting, worthy of learning. And FM extraction has potential.

But the BS, over-eagerness, and desire to be cheap, is the issue that I always point out with the vhs-decode project. Most of the "comparisons" as seen on Youtube are disingenuous at best, ignorance at worst. It's often cherry picking the worst possible capture hardware. Instead, what about uploading some quality captures that some of us do on a regular basis, using tried-to-true workflows?

There were great comments on the Hackaady article, about the "new and improved" sample:
- But … isn’t VHS-decode the worst of the four? I don’t understand…
- Feels like what a good S-VHS recorder could do on a new tape back in the day.
- You are still on the same place where you was 5 years ago, quality is same bad quality as before 5 years and you, the group of deep amateurs cant do it better. So 1,2 or 5 years before, the quality and reliability is the same – BAD.
- What you’re “planning” is a very low-res version of what Sony did in the early 80s.

^ The fact that one of the devs thought that I'd left those comments is amusing. Until just now, I'd never ever seen that article. He thinks I'm the only one critical, but as I've stated before, others that truly know video capture aren't impressed whatsoever, especially the high end (archivists, restorationists, etc). That leaves the low end, and they have Easycaps, HDMI adapters, essentially fodder crap. So you're essentially just BS'ing newbies that don't yet know if they want to go for quality, or succumb to the Chinese USB crap. That is literally the only positive comments I ever see on this project. People that don't capture video, and wouldn't know a capture card from a network card.

A few folks have decided to attempt it, and gotten results they want. That's fine. But it doesn't override any of the above.

[VideoMem]

Well, this gonna be the slowest response against FUD.
Some of these article responses are FUDing us, FUD over the project and gaslighting the collaborators.

I say us because I spent long hours programming some parts.

The last piece I contributed was hifi-decode.

I really don't not know why, and at this stage I'm afraid to ask why it is worth the effort.

If you state vhs-decode is not better than the best solution, please tell us where you see that difference.
We can change it, when spare time, it is still in development, it is not discontinued.

Professional sales are like noise reduction of technical pitfalls.
They inpaint defects, they mask, they discontinue when not sold enough.
We don't do that here, I think this is for other kind of purpose.
DNR and noise reduction maybe adding to the comparision error.
Unless non linear noise reduction activated, vhs-decode does not have other kind of DNR or inpainting besides dropout correction.

Tell me one manufacturer/supplier/provider of tape playing devices that give you support for understanding what's going on inside their black box signal pipeline IC in a way even if they discontinue the line or go bankrupt their clients won't fall into the abyss of technical debt.

We're doing a technical support and reverse engineering for a technology considered obsolete.
It goes in a paralell universe where people still have analog interests.

-With documentation lost and sites going offline.
-With no active factories making newer tape players.
-With a lot of hype because nostalgy.

And they go in flames (in response to the fud of these comments) because word 'replace' without 'maybe', or better, 'complement or coexist' in a sentence.
If we caused trouble to someone's business, we're sorry, there is no businesses intentions here (at least not at my side), and no hurry.

[VideoMem]

Now I see what are you pointing out, and I agree on some of them.

But some are wrong, vhs-decode is not cheap if you sum the time it takes to master it.
It is not optimized, so it could be power hungry if done at big scale.
There is no gui. The gui will not be cheap.
There is ffmpeg also.
Only thinking about the parameter mappings into knobs and/or valueboxes/sliders, cross platform considerations, look and feel and user experience.

The capture setup is tricky, like any other player, a professional grade one is often made to last more hours of continue use without service.
Consumer grade ones in other hand are prolific and share parts between.

Vhs-decode does not limit you to a specific type of hardware, but as always, the final product or experience is never one isolated part of the signal capture, in this case vhs-decode can't produce good results if the input signal is low or it has poor frequency response, has aggregate noise.

Personally I see an early overoptimization search of the cheaper way and agree there's lot of edge case decoding samples out there.
Most of these are indeed difficult portions on where the decode process had (and sometimes has) its challenges, TBC failures and so on.

If you have access to hi end equipment, there's no way I recommend using just a capacitor, a resistor, a meter of coax, one bnc plug, a recently made pci-e card with a refurbished chip and a desktop pc running linux with a hacked/improved driver (unless you know what you're doing).
I would suggest an oscilloscope probe in 10x mode, and an amplifier at least, and asking engineers to conduct noise and tape calibrations tests of the unit with or without the probe connected to ensure there is no ground loop nor issue causing the setup to introduce noise into the player, verify that the input signal has enough amplitude, it is continous power waveform and the machine is tracking properly.

Even the heads can become dirty and degrade the result.

The software enginners by itselves cannot outperform the result they could if they work along the specialists in the field to solve issues.
That kind of fertile ground for science is not cheap, not easy, hard to coordinate and its pitfalls are not only the tool used by itself for the task.
Quality involves organization, discipline, commitment along with the tools. Knowledge and expertise are their results.

[VideoMem]

A fair comparision benchmark could be same device captured at conventional but without dropout compensation nor DNR and rf vhs decode without DOC also. A sort of heterodyned followed by TBC 1:1 comparision.

In SP mode or the fastest tape speed the machine can handle.

[peppi001]

My opinion:
I am very happy to have found the vhs-decode project.
Of course there are mistakes, but the quality is still very good.
I have already digitized several cassettes with this.
Big thanks to all developers.

(P.S. I'm currently stuck with a secam tape.)

[lordsmurf]

I say us because I spent long hours programming some parts.
....
Now I see what are you pointing out, and I agree on some of them.

You seem to be the 2nd reasonable dev I've met.

What I have problems with is the utter BS some are spouting, the hyperbolic nonsense that is divorced from reality. Some from various devs, some from fanboys (that are mostly video ingest illiterate). It's just noise from a peanut gallery, and serious discussion cannot be had. At some point, as with everything else open-source, there may have to be a fork, to separate wheat from chaff, before the project actually matures.

[dellsam34]

Would a VCR with the RF head preamp built in the drum itself bypassing the rotary transformer yields better signal for the VHSdecode card? The Toshiba M-751 uses 8 heads, 4 dual azimut video heads for LP/EP and two video heads for SP and 2 heads for HiFi audio, It uses a direct rotary connector like those used in mode switches and potentiometers instead of a rotary transformer, I though I would ask if this VCR is a better condidate for this task since it doesn't have the drawbacks of rotary transformers. Not sure how marginal the difference is but according to Toshiba there is a difference in signal quality.

[lordsmurf]

but according to Toshiba

"grain of salt" needed. Toshiba was an aggressive marketer in the 90s-00s, to be the point of being BS at times. Similar to Canopus. The engineers don't write product documentation, but rather marketing dept copy writers do. They can get hyperbolic, and it was definitely the case on VCRs and DVD players with Toshiba in those decades. Also, much of it was translated from Japanese, so lots of lost in translation at times. I still remember some confusing manual text and on-screen menus with my favorite DVD player in the 00s, and how IRE was described as "color boost".

[oln]

Depends on what extent the flying preamp helps, which I have no idea about really. JVC also had this in a few of the big ass "broadcast" models but not in any consumer one afaik. I have a PAL Toshiba with it (V-804W) but haven't tested rf capture from it.

[dellsam34]

That would be great if you can do some test captures with the Toshiba, In theory having direct feed from the heads by hard wiring them should be better than using an isolated rotary transformer but I'm not an expert in RF transmission so I wouldn't know what the drawback are of carrying RF via an isolation transformer.

[Orikson]

Hello everybody,

I'm new to this forum, but I already read some pages about vhs decode. I'm currently about to safe my old family tapes, mainly Video8 but also some VHS and betamax back from 1987. I already tried the "Gubbel" method (VCR > S-Video > Panasonic HDD Recorder > HDMI > Blackmagic Intensity Pro 4k) as well as the more professional version (VCR > S-Video > Canopus Edius NX PCIe Card) and like the results. But now I found out about vhs decode (and HiFi decode) and thought about giving it a try, to squeze out even more details. To be sure that I understand that right, I wanted to ask a few questions.

The digitizing of the video is basically split into two parts. First step is to record the raw RF data, that is stored on the tape. Second step is to convert this raw data to the actually video. This conversion is not yet working perfectly for Video8 tapes, is that correct?

Can I anyway already record the raw RF data and convert it as soon as this second part is working perfect? Or does the expected data format for the raw RF data change over time and will not be compatible with newer versions of the conversion? The tapes degenerate over time, so I would at least like to get the current state in a non-degenerating state to work with them later.

For the hardware I'm not quite sure what to get. First of all I thought about getting the a PCIe card with CX23881, so that the 40 MHz quartz mod will work fine + AD8367 amp. Even though this might be not necessary for Video8, I would like to play around with oversampling. Or is the quality with the Domesday Duplicator even better, if only a little bit?

I also can provide various video snippets from various VCRs if this will help to improve the tool.

And last but not least: Thanks to everyone contributing to this great project

[oln]

Video8 (and betamax for that matter) isn't as refined as for VHS yet yeah. So provided you get a actually good capture of the rf signal the decoding steps can be re-done with improved software later yes, the rf data won't be incompatible. For Video8 also we haven't quite managed to sort out quite how to best capture from the newer camcorders as they seem a bit finicky about having stuff connected to the test point. So while for vcrs the difference between cx card + amp and Domesday Duplicator is probably very minor I'm less sure when it comes to video8, especially if using a newer camcorder - details are a bit unclear there still.

On the flip side there is might not be quite as much to gain from rf capture of video8 it compared to vhs/betamax since the conventional options there are already quite decent with Sony camcorders sans the issue with Chroma on the right side of the image on PAL on them which can be a bit annoying.

[Titan_91]

Here are a few side-by-side comparisons.

https://forum.videohelp.com/threads/394168-Current-status-of-ld-decode-vhs-decode-(tru...25#post2629620

Notable differences:

Proper time base correction
Improved detail recovery
Improved black level and white clip restoration (dynamic range)
More accurate color recovery
No post-processing filters to smoothen or sharpen the image as most VCRs do
Drop out correction
Proper framing of the image including greater overscan area captured
No VCR on-screen overlays like "PLAY" or "TRACKING" obscuring the image

All of these improvements result from digitally capturing the RF signal from the head drum itself and bypassing 20+ year old demodulation, de-emphasis, and other processing circuits normally used for conventional video capture in the past. Everything is done losslessly in software except the bare minimum process of playing the tape, tracking the RF envelope, switching between the 4 heads, and amplifying the output from the head drum to a test point on the VCR's circuit board. Other steps, such as connecting a traditional hardware time base corrector, involve more physical connections and cables, and these steps introduce signal loss and distortion. These losses are a result of physics and how electrical signals behave when heavily processed by traditional analog circuits.

https://imgsli.com/MjAxNTE4

https://imgsli.com/MjAxNTE5

https://imgsli.com/MjAxNTIw

This 35 year old VHS-HQ tape from 1988 was captured conventionally about 10 years ago, and RF captured again last week with a $10 PCI card. Even though the tape has aged 10 years between the two captures, and its RF signal has gotten weaker, the one done last week still looks miles better.

Conventional capture:



RF capture using VHS-decode:



Conventional capture:



RF capture using VHS-decode:



Conventional capture:



RF capture using VHS-decode:



VHS-decode is the new standard for videotape archival, as LD-decode is the new standard for LaserDisc archival. This improves not only the audio and video playback experience, but you fully preserve everything stored on the tapes and discs as they were physically mastered. No process in history has achieved this before, at the consumer level.

[Orikson]

Impresive comparison! Can you tell which setup you have been using for the old conventional capture and for the RF capture?

I read that some are using a hardware low pass filter, e.g. Mini-Circuits BLP-10.7+ with 14 MHz cutoff, but why? Couldn't this filtering be done in software?

[Alwyn]

Wow!

[dellsam34]

Something really wrong with the output of that VCR or the capture device in the comparison, The details are completely washed out, How did you capture from the VCR, what output you used and what processing steps have you done? It doesn't make sense that the head switch is shown on the vhsdecode screen shot and no head switch on the conventional capture, If the same VCR is used the head switch should be identical.

[rgr]

Here are a few side-by-side comparisons.

There's something wrong with conventional capture -- even white looks blown out.

[Sharc]

Yes. The conventional VHS look heavily denoised, and the blown out brights are clipped at luma = 235, loosing any details.
The results of the VHS decode look amazingly good though.

[Titan_91]

Impresive comparison! Can you tell which setup you have been using for the old conventional capture and for the RF capture?

I read that some are using a hardware low pass filter, e.g. Mini-Circuits BLP-10.7+ with 14 MHz cutoff, but why? Couldn't this filtering be done in software?

Yes, the old capture used an off the shelf composite DVD/VCR combo, Hauppauge HD PVR, and h.264 compression at 5Mbps, more than enough for standard definition content.

I am using a Mini-Circuits 13MHz low pass filter, yes. I do this for VHS and Laserdisc captures. While you can filter out high frequency noise on the software side, I do this because my Conexant CX2388 capture card has been modified to remove the low pass filter on the input pin I'm using, VMUX2. This minimizes the possiibility of a second or third harmonic of the captured signal aliasing and corrupting the fundamental signal.

Something really wrong with the output of that VCR or the capture device in the comparison, The details are completely washed out, How did you capture from the VCR, what output you used and what processing steps have you done? It doesn't make sense that the head switch is shown on the vhsdecode screen shot and no head switch on the conventional capture, If the same VCR is used the head switch should be identical.

Yeah I'm not sure what caused that but all my captures I previously did looked like that. Normal tapes looked fine in the same machine. And the Hauppauge HD PVR looked fine with other content. These are VHS-HQ tapes. So the tapes may have been slightly out of spec on the luma pre-emphasis. Very possible the $1600 Quasar camera that recorded them wasn't 100% on spec. The head switch being higher is likely due to it being a different machine for playback.

[lollo]

Titan _91, we all love vhs decode and have great expectations, but for a fair comparison, use a capture with the recommended workflow (high end S-VHS VCR with Y/C output and lineTBC, eventually a frameTBC, and one of the suggested consumer capture cards (Hauppauge USB-Live 2, IOData GV-USB2) or a high-end cards (BrightEye 75, Canopus NX, ...)

[Titan_91]

I realize that, but respectfully you're missing the point. There are plenty of other comparisons on YouTube and in the Discord server using more expensive VCRs and hardware time base correctors. And while the software approach is a work in progress, assuming you have a good RF capture from the head amplifier, it's still more capable than a traditional hardware chain and is improving every day. Heck, someone just discovered the ideal pre-emphasis curves for Hi8 today in the Discord server. And I've been working on Video8 improvements as well.

The less electron wrangling you have to do, the better the signal to noise ratio and lower the distortion and losses of the information originally recorded on the tape. So basically while what has been used traditionally is very good equipment, it's still 4+ connections and several more processing circuits than a simple RF capture.

https://www.youtube.com/results?sp=mAEA&search_query=Vhs-decode+comparison

[lollo]

but respectfully you're missing the point

Respectfully, you posted a useless comparison.

I know all the rest, because I follow vhs decode improvements step by step, even if I do not participate to the development

[dellsam34]

Not just useless, it's misleading.

an off the shelf composite DVD/VCR combo, Hauppauge HD PVR, and h.264 compression at 5Mbps

that's not putting the thumb on the scale. more of sitting on the scale.

[Titan_91]

Ok, let's establish a level playing field then. Post a full rebuttal, in NTSC, NTSC-J, and PAL, using your recommended equipment workflow, with the following:

Comparison with and without time base correction
Spectrogram of your video capture
Spectrogram of your Hi-Fi capture
White and black SNR graphs
Use of different comb filters
Use of different luma filters
Use of different chroma filters
Use of various de-emphasis curves, including non-linear curves
Use of different burst ABS references
Use of automatic gain control
Use of chroma trap
Use of luma comb filter
Use of comb filter vs. no comb filter
Use of sync level clipping
Video waveforms for luma and chroma channels
Full frames including blanking areas
Results with and without dropout correction, including different thresholds
Results of your decoded vertical blanking interval data
Results of your decoded teletext data
Results of your decoded TV Teddy audio data
Use of manual vs. automatic video level detection
Use of blanking DC offset clamping/compensation
Use of chroma automatic frequency control

And anything else you would like to include in your comparison.

[lollo]

Post a useful comparison then.

Once I'll have a workflow for vhs-decode (not my priority, but it will happen), it will be my pleasure.

In the meantime please compare apples to apples, there are already too many people criticizing vhs decode without need to feed them further

edit: dellsam34 is not one of them, he has a nice and not prejudicial approach to vhs decode

[dellsam34]

Post a useful comparison then.

Not interested in such hardware, but I will be happy to send you my comparison side file if you would send me the exact same tape.

[lollo]

Not interested in such hardware, but I will be happy to send you my comparison side file if you would send me the exact same tape.

Yes, good idea. Same for me with my hardware! It would be a nice demos for the community!