What do you think of this workflow for capturing commercial VHS?

[robertzombie]

Hello,

I have a small but growing collection of commercially released VHS tapes of films, concerts, and television programmes that haven't been reissued on optical or streaming formats. I would like to capture these tapes, mostly for preservation but also to tinker with restoration tools to see what's possible. The tapes are all PAL from the late 1980s and early 1990s. They're in very good condition but the AV quality of the content does vary.

I have been reading the forum over the last few days and was wondering if the experienced members here might take a look at my proposed workflow and settings and let me know if I'm heading in the right direction. I should add that I have done VHS to digital transfers professionally before using broadcast equipment not available to me at home, however I haven't performed any video restoration. I have done plenty of professional audio restoration though.

First, my capture chain:

• VCR: LG LV880. 2009 model, capable of PAL and NTSC playback, Hi-Fi audio. In excellent working condition, clean inside, clean heads. Only real downside is it doesn't have S-video output but it does have seperate audio output. Manual.
• "TBC": Panasonic DMR-ES15. Visually in excellent condition, seems to be working well. I have a few questions about its settings though (below). Manual.
• Capture device: Diamond VC500 USB.

Video path: LV880 > scart output > ES15 > s-video output > VC500.
Audio path: LV880 > rca output > VC500.

Will bypassing the ES15 for the audio path introduce any significant AV sync issues?

I'll be capturing the video to HuffYUV AVI and audio to 16/48 in VirtualDub. I'm going to use the guides here and here to configure the software.

Here are the settings that I'm using with ES15:
Functions:

• Comb filter: On (in my opinion this offered an improvement when viewing a VHS on my TV, but I'm not sure if it's better to de-comb in software)
• On-screen messages: Off
• Grey background: Off
• TV Aspect: 4:3
• Progressive: Off (not sure whether to set this to On? I couldn't see any difference when viewing a VHS and assume it's unecessary for VHS. Setting this to "on" does open up some additional display settings however)
• TV system: PAL

Display > Picture Menu:

• Playback NR: Off
• Progressive: Off (see question above, if progressive is set to On via the Functions settings then additional settings are found here called "Transfer" auto/film/video. See here in the manual. Is it safer to just leave all that off?)
• AV-in NR: Off (when viewing a VHS on my TV I toggled this on and off and thought the NR was actually quite acceptable, hopefully I can achieve something equal or better in software)

I have a colour bars and tone VHS that I recorded from the DVD available here. I'm planning on using this to set a reference audio capture volume level in software. I was wondering if a capture of this VHS could be used to make any meaningful assessment of my transfer chain's effect on colour levels? Is it possible to create a colour correction profile by comparing the VHS capture with the master file and using that as a starting point for correcting other captures? I'm guessing not, but thought I'd ask.

So that's where I'm at so far. I'd be grateful if you could tell me what you think of this!

[manono]

Only real downside is it doesn't have S-video output but it does have seperate audio output.

As you already know, having an S-Video output will be better, but if you can't, you can't.

Will bypassing the ES15 for the audio path introduce any significant AV sync issues?

If I remember correctly, the delay will be about +200ms. It's easy enough to fix later on.

Progressive: Off (not sure whether to set this to On?

Always off.

AV-in NR: Off (when viewing a VHS on my TV I toggled this on and off and thought the NR was actually quite acceptable, hopefully I can achieve something equal or better in software)

VCR noise reduction wasn't very good. You can do better in software afterwards. I assume you'll be using AviSynth for your restorations?

[DB83]

I am probably wrong with this and the red wine is inevitably talking.


But I would have thought that using the ES15 in pass-through mode you have composite in (since your VCR does not output s-Video) and composite out. Or does the ES15 actually split the video signal internally ?

[robertzombie]

Only real downside is it doesn't have S-video output but it does have seperate audio output.

As you already know, having an S-Video output will be better, but if you can't, you can't.

Will bypassing the ES15 for the audio path introduce any significant AV sync issues?

If I remember correctly, the delay will be about +200ms. It's easy enough to fix later on.

Progressive: Off (not sure whether to set this to On?

Always off.

AV-in NR: Off (when viewing a VHS on my TV I toggled this on and off and thought the NR was actually quite acceptable, hopefully I can achieve something equal or better in software)

VCR noise reduction wasn't very good. You can do better in software afterwards. I assume you'll be using AviSynth for your restorations?

Thanks for your reply. It would certainly be interesting to see an "S-Video vs. Composite" comparison using the same commercial VHS. I always thought standard VHS was composite in nature, and S-video output was only beneficial with S-VHS tapes. Kind of like how "upsampling DACs" make no difference to the frequency response of redbook audio.

The latest guidelnes from IASA on the Preservation of Video Recordings say the following about S-VHS:

A major evolution of the format is the Super VHS (S-VHS) system (Beeching: 2001,p. 30). Introduced in 1987, the higher coercivity tape formulations allowed a much higher luma carrier frequency to be used, which improves luma signal to noise ratio and frequency response and considerably better luminance resolution. S-VHS playback machines also introduced the S-Video connector, which provided higher video quality due to separate luma and chroma signal outputs, as opposed to the older composite connector.

S-VHS tapes played back in a standard VHS machine will exhibit white speckles and picture tearing. The colour-under frequency used for the chroma was unchanged in S-VHS, so there was no improvement in the quality of the colour part of the signal, leading some manufacturers to develop Chroma Noise Reduction (CNR) replay circuits. An S-VHS cassette has a dedicated S-VHS detection hole on the underside, which can assist as a visual check to identify S-VHS carriers (Beeching: 2001, p. 56).

Older S-VHS machines will switch into normal VHS recording mode when detecting an oxide rather than metal tape. However, JVC later produced the SVHS-ET system (Beeching: 2001, p. 56), which allowed certain machines to make S-VHS recordings on standard tapes, further complicating playback for archivists.

If it's crucial to capture standard VHS with an S-VHS machine then I am happy to consider it!

I've been having a look at the Film9 software and considering using that for restoration, my understanding is that it is an AviSynth GUI. Standalone AviSynth might be a bit beyond my command line capabilities and a GUI would suit me better. That said, I'm not ruling out using standalone AviSynth.

[robertzombie]

I would have thought that using the ES15 in pass-through mode you have composite in (since your VCR does not output s-Video) and composite out. Or does the ES15 actually split the video signal internally ?

The ES15 has composite, S-video, and component outputs. It has a built in comb filter, so I'm assuming that I'd need to use its S-video output to avoid a reconversion back to composite after the comb filter.

[DB83]

Well the only real way to see if s-video out is supported from composite in is to try it. What I do know is when dealing with most vcr/dvd-recorder combos - I do appreciate that the ES15 is not one - is that s-video is only available on the dvd side. VHS output is for composite only. Yet I would equally understand that if you have s-video in then you can expect s-video out.

[Barrythecrab]

The ES-15 has S-video out, and I think you can use it even with composite input. I believe I read it will even clean up the dot crawl. Maybe I’ll test mine later, I really just use it for tapes that tear at the top.

[DB83]

Two items that could fail with your proposed work-flow.


1. Any tape that has macrovision copy-protection. Rareish on UK tapes but I have handled some in the past. In that scenario you will require a full-frame TBC to defeat the mv.


2. NTSC tapes. As you pointed out your vcr, like most UK models, has NTSC playback. But the output is neither pure NTSC or pure PAL. Some dvd-recorders can take this PAL60 input and spit out a NTSC signal. Not sure if your Panny can do this. Yeah, I know, RTFM.

[Barrythecrab]

RTFM...I had to look that up!
Please note: DB83, you look like you had a less than optimum visit for your Corona nose swabbing, I think they forgot something.

[dvd3500]

Pre-2003 there were macrovision stripping boxes you add in your chain to remove it.
They are technically illegal now but possibly available....

[lingyi]

VHS/S-VHS is Y/C and so s-video (Y/C) is native output. It's Laserdisc that is composite.

Here's a thread that discusses s-video vs composite and a comparison. There's another recent thread (that I can't find right now) in which the poster has a lot of comparisons of different capture methods.

http://www.digitalfaq.com/forum/video-capture/4915-video-cable-composite.html

http://www.digitalfaq.com/forum/video-capture/9561-composite-vs-video.html

Most of the Macrovision boxes did a poor job at best and many didn't do anything at all. As DB83 stated, the only way to properly and fully strip the Macrovision signal* is by using a TBC.

*As lordsmurf reminded me, there are numerous variations of Macrovision and is one of the reasons the Macrovision removal boxes did poorly or didn't work at all. Memory came back that back in the day, the boxes were constantly updated, promising to work on the newest protection scheme (similar to what DVD ripping programs have to do).

[usually_quiet]

For Macrovision removal, the XDIMAX GREX-7.4 Video Stabilizer is still in production. A full frame TBC is better. If you can find a good used one, the AVT-8710 was a popular consumer model.

[dvd3500]

The one I ..saw... definitely worked. It was a smalls series done in Germany.

[lingyi]

Per lordsmurf:

"Do not buy a Grex. That thing is overpriced junk that messes with the signal. It's a cheap wanna-be TBC, and claims to remove copy protection. It does not. Not entirely. It will "remove" some aspect, but leaves behind issues with color, luma, flickering, etc."

http://www.digitalfaq.com/forum/video-workflows/9240-necessary-hardware-capturing.html

And only the green AVT-8710 is good:

"The AVT-8710 hasn't been good since 2010. Don't buy the black boxes. Only the green ones are any good."

http://www.digitalfaq.com/forum/news/7964-avt-8710-vs.html

[orsetto]

And only the green AVT-8710 is good:

"The AVT-8710 hasn't been good since 2010. Don't buy the black boxes. Only the green ones are any good."

http://www.digitalfaq.com/forum/news/7964-avt-8710-vs.html

And THIS is precisely where the entire house of cards collapses, and has been for years already. Good luck finding one of the two recommended external TBCs in both good operating condition and at a price that isn't more than you would pay a service to do the job for you. The DataVideo TBC1000 has the seemingly sturdier build, but I've been thru several that worked normally for a few months then inexplicably went defective: at the going rate of $400- $500 if you even find one, there is some risk. The AVT8710/CBT100 sideshow has been a confusing mess for as long as I can remember: I think you could buy one with no trepidation for about the first twenty minutes they were mfr'd, but they've been a total dice roll since no matter what the color (green may up your odds of success, but the basic, very variable "barely qualifies as a TBC" overheating design remains the same).

So if you have MV problems with some of these tapes, sometimes the only easily available recourse today is in fact one of those benighted dedicated MV filter boxes. You don't need a Grex: it is overkill for the task because 80% of its circuitry is oriented to other forms of protection irrelevant to VHS. All you need is a recent generic filter that looks like a black plastic cigarette pack with two jacks (composite in/out) on one end, powered off a 9v battery. Inelegant, and video professionals sneer at them as garbage, but if the choice is a mediocre result or no result at all you go with mediocre. And in truth, protected tapes throw a wrench into the final result even if you employ a $500 TBC, because even TBCs don't filter it without adding their own artifacts. I've used multiple examples of the DataVideo and AVT/CBT with MV tapes, and all of them left me non-plussed: they're more consistent than a Grex, but remarkably sometimes worse than one of the little 9v filter boxes + a dvd recorder used as framesync. Depends on the particular TBC or filter box, and particular capture chain.

Ideally, get a nice TBC if you can afford one. If you can't find or afford a good example, or the TBC results are unexpectedly disappointing, try the little black box between VCR and ES15, with ES15 feeding your capture device. Analog MV for VHS stopped evolving a long long time ago: the days when filter boxes cost $250 and needed firmware updates semi-annually ended around 1995. Any $30 - $50 black box made since can cope with most any tape (one of my 9v boxes dates from 1993, is held together with duct tape, and still delivers cleaner colors than any DataVideo I've used as MV filter). What works for non-commercial tapes doesn't always work as well for MV tapes: multiple approaches may need to be tried.

[DB83]

Ooh. I may own one of those 'ciggy boxes'. It has the words 'Video Decoder' emblazoned on its side and DX-11 and, confusingly, 'DVD' on the top. So maybe not . Never tried to use it since my Hauppauge card operating no later than WinXP actually defeated mv all on its own.

[lingyi]

Don't know about for digital capture, but I had a black box circa 1995 that didn't do anything for VCR to VCR copying. Fortunately, the Sony Betamax(s) I used ignored most types of Macrovision. I later got a TBC-1000 primarily for use as a distribution box.

[dellsam34]

In addition to having a VCR with S-Video out as the folks suggested above, while at it get a model with built in line TBC, Line TBC works well on commercial tapes than home made tapes, you will see an improvement with it being ON.

[orsetto]

In addition to having a VCR with S-Video out as the folks suggested above, while at it get a model with built in line TBC, Line TBC works well on commercial tapes than home made tapes, you will see an improvement with it being ON.

So true! Commercial tapes tend to have a starkly binary bell curve: either they're very well recorded, or have terrible chroma/luma noise that is best cleaned up by a VCR with built-in DNR circuits before it enters the capture chain.

Unfortunately finding one of these VCRs thats still performing up to spec is nearly as hard as finding a good TBC today. Worth the effort if you have important tapes that would benefit from hardware noise cleaning (but only buy from a reliable seller with a good history dealing in VCRs: getting a worn one repaired is difficult and can be very very expensive depending on model).

[robertzombie]

Thanks for all the interesting replies!

I spent the day setting up the VC500 and doing some test captures. It seems to be working fine with my Windows 10 laptop, which is a relief! My first film length test capture is about to finish, so I'll be able to see how it held up soon enough.

[oln]

I am probably wrong with this and the red wine is inevitably talking.


But I would have thought that using the ES15 in pass-through mode you have composite in (since your VCR does not output s-Video) and composite out. Or does the ES15 actually split the video signal internally ?

Nah, it doesn't work like that. It seems to fully digitize the signal in the process - possibly to 4:2:2 Component ITU 601 format. It's what more off-the-shelf analog video decoders typically do, but I don't know for sure what the all-in-one panasonic chip does internally since there aren't any datasheets available. There isn't a separate path for different input types - it looks like the same line is used for Y in Y/C mode and composite when that is used. Presumably the A/D chip will turn on/off comb filters and switch whether it takes color from the Y or C input etc.

You could think of it a bit like say capturing the video to a computer with a capture device, and then playing it back again to an analog output.

Shouldn't be any reason to use composite output, use either S-Video or Component.

For the original question:

The Progressive settings only mean something on the component video output - if it's on the ES15 will output a full deinterlaced 50P signal there, otherwise it's a normal interlaced output similar to the S-Video out. The auto/film settings for progressive out I think it's just to change between de-telecine (for film content that is originally 24 frames per second) and normal deinterlacing.

The PAL ES15 and other PAL Panasonic DVRs do support a normal NTSC signal, though not PAL60 like this and most PAL VCRs output. You can find some multi-system ones that have proper NTSC 3.58 if you're lucky, though they're rare and not always hifi-stereo. I do not know if it works with NTSC-4.43 that some PAL VCRs (especially Sony ones) can do.

I've also had some issues with the audio gain being too high on the VC500 (and many other capture cards for that matter) with no way to adjust it on windows. This could result loud audio (typically hi-fi) clip and get distorted. Maybe they've altered this in later revisions.

[manono]

Well the only real way to see if s-video out is supported from composite in is to try it.

Meaning you might not get a viable capture if doing it that way? I've done it many times using composite out from a VCR to a Panasonic DMR ES-15. Then S-Video from the ES-15 to my USB capture device.

[DB83]

^^ Fairy Nuff

[jjeff]

As Orsetto said, the Grex is about as good as you can currently find for DVD copying and it does work for VHS but again as Orsetto said, it's kind of an overkill for VHS. I have a Grex and it removes any CP I've tried but does very slightly raise the black level and softens the picture a hair. I have many Sima CT-2's that handle resolution better than the Grex but only one has a decent black level, the rest raises it worse than the Grex
Another trick mentioned is to use an old Betamax inline with your VHS player, I've used a SL-5400, SL-5600 and SL-5800 and they all work in that capacity, note I don't believe the Betamaxes will remove DVD CP, just VHS.

[robertzombie]

I've also had some issues with the audio gain being too high on the VC500 (and many other capture cards for that matter) with no way to adjust it on windows. This could result loud audio (typically hi-fi) clip and get distorted. Maybe they've altered this in later revisions.

Yep, I had to install a passive pre-amp between the VCR and VC500. I recorded an EBU Colour Bars VHS with 1 kHz tone @ -12 dBfs and used that to set a reference level in VirtualDub by adjusting the volume level on the passive pre-amp.

[robertzombie]

Here are some colour bars then. Interesting to see the difference between the DVD source and VHS transfer in the two screenshots. I think the colours are fairly close, all things considered.
I'm slightly worried about the audible harmonic distortion, though it's not something I've heard when testing normal VHS tapes...

[robertzombie]

Thanks for all the interesting replies!

I spent the day setting up the VC500 and doing some test captures. It seems to be working fine with my Windows 10 laptop, which is a relief! My first film length test capture is about to finish, so I'll be able to see how it held up soon enough.

First film length transfer: video played fine without any dropped frames. However, the audio sync drifted significantly over the duration of the 90 minute film. I need to have a play with VirtualDub's 'Timing' settings. Anyone got any tips for setting that correctly for the VC500?

[lingyi]

As Orsetto said, the Grex is about as good as you can currently find for DVD copying and it does work for VHS but again as Orsetto said, it's kind of an overkill for VHS. I have a Grex and it removes any CP I've tried but does very slightly raise the black level and softens the picture a hair. I have many Sima CT-2's that handle resolution better than the Grex but only one has a decent black level, the rest raises it worse than the Grex
Another trick mentioned is to use an old Betamax inline with your VHS player, I've used a SL-5400, SL-5600 and SL-5800 and they all work in that capacity, note I don't believe the Betamaxes will remove DVD CP, just VHS.

Most, but not all Betamax machines ignored most (but not all forms) of Macrovision because of the recording method and AGC. They just ignore the Macrovision, but it's still there. Proven by my having copying a VHS tape to Beta, then back. I never tried it the my SL-5800, but the later machines I had didn't work as a passthrough and you state. As stated above, Macrovision when through a number of changes over the years.

DVD copy protection isn't embedded in the signal like Macrovision, it's part of the file structure and needs to be removed by the ripping software. Completely different than Macrovision in any form.

[jjeff]

As Orsetto said, the Grex is about as good as you can currently find for DVD copying and it does work for VHS but again as Orsetto said, it's kind of an overkill for VHS. I have a Grex and it removes any CP I've tried but does very slightly raise the black level and softens the picture a hair. I have many Sima CT-2's that handle resolution better than the Grex but only one has a decent black level, the rest raises it worse than the Grex
Another trick mentioned is to use an old Betamax inline with your VHS player, I've used a SL-5400, SL-5600 and SL-5800 and they all work in that capacity, note I don't believe the Betamaxes will remove DVD CP, just VHS.

Most, but not all Betamax machines ignored most (but not all forms) of Macrovision because of the recording method and AGC. They just ignore the Macrovision, but it's still there. Proven by my having copying a VHS tape to Beta, then back. I never tried it the my SL-5800, but the later machines I had didn't work as a passthrough and you state. As stated above, Macrovision when through a number of changes over the years.

DVD copy protection isn't embedded in the signal like Macrovision, it's part of the file structure and needs to be removed by the ripping software. Completely different than Macrovision in any form.

I guess when I was talking about DVD CP I was talking about the MV part as in copying a DVD realtime, not the file structure part you'd run into copying the DVD on a computer. I've never seen a DVD that had digital CP but not MV although I suppose it could be possible? I did frequently see DVDs with MV but not the digital CP and in this case, they wouldn't copy realtime but would copy freely with a computer, even one without special ripping software.

[DB83]

Thanks for all the interesting replies!

I spent the day setting up the VC500 and doing some test captures. It seems to be working fine with my Windows 10 laptop, which is a relief! My first film length test capture is about to finish, so I'll be able to see how it held up soon enough.

First film length transfer: video played fine without any dropped frames. However, the audio sync drifted significantly over the duration of the 90 minute film. I need to have a play with VirtualDub's 'Timing' settings. Anyone got any tips for setting that correctly for the VC500?

I'm pretty sure there is a setting in vdub to drop frames to keep audio in sync. The dropped video frames may not be noticeable but only you can tell from the resultant capture.


Even so it may still be more productive to capture to an external HDD so that the OS, which could be trying to write to the same HD as your cap at the same time, does not overtly influence the capture.