Want to be a part of a VHS test?

[guyburns]

Before I digitise the Hans Vonk tapes (see this thread), I've decided I'd better make sure the procedure and equipment I use gives good results. So, with a technically-minded friend who is also keen, we're going to begin.

He has a JVC HR-D430EA (from 1987) and a Hauppauge HD PVR 2 capture unit.

Test Procedure
1. Choose the best-quality VHS movie he has, and decide on which 10-30 seconds to use as a Test Sequence. Opening credits are probably the best because of the lettering.

2. Record the Test Sequence six times with these file names (or whatever bit rates are available):
- Composite 5Mbps
- Composite 10Mbps
- Composite 15Mbps

- Component 5Mbps
- Component 10Mbps
- Component 15Mbps

All captures to be flat: no sharpening, colour/contrast correction, or other optional alterations to the video.

Premiere Comparison
I'll import the files into Premiere, layer and align them, and see if there is any difference in video or audio quality. Then when I get access to other VCR players and capture devices, I'll do the same thing as a comparison between players and capture devices.

Enthusiasts Wanted
This is where any enthusiast reading this comes in. If you'd like to be a part of this test, gmail me at gdburns with what equipment you have, your phone number and address. I'll talk to you on the phone, send the tape, you digitise as above, upload the files to my drop box, and post the tape, either to me or the next person on the list (if there is a list). If you require payment for your efforts, indicate how much. Payment will only come after the files and the tape are returned to me.

Is he kidding?
You may think: "Is this Burns bloke for real?" Well, 10 years ago, at the start of a 12-year project turning historically-important Tasmanian wilderness slides into AVs (see here), I did the same thing for scanning slides. After 18 months of investigation, the result was a 250-page PDF: The Art and Science of Scanning Kodachrome.

This present test will be nowhere near as substantial (to be finished by the end of February), but I do want to know how to proceed before I begin digitising VHS tapes.

Test Results will be made publicly available.

[latreche34]

How can you capture component from a VCR? Anyway, neither composite nor component are recommended for capturing VHS, For best results you would need a S-VHS VCR built in line TBC that can output S-Video, Bad tapes may need a frame/field TBC.

[guyburns]

Thanks for the response.

This isn't about theoretical best results, this is about real world comparison. I don't mind spending a heap of dollars on equipment, but I do mind if I could have achieved the same results for 10% of the cost and bother.

I don't know how you capture component from a VCR. The player has component output; the Hauppauge has component input, so I assume it can be done.

[gunzel]

I take it from the other thread that your source tapes are in PAL?

Do you have so much source material that the savings in disk space are worth testing a bunch of different compression schemes rather than just capturing uncompressed or using a lossless codec?

I have a Philips VR1600, an ATI 600 USB, and a Canopus ADVC-100, using an old Mac running Snow Leopard for capture currently but have been looking at getting Boot Camp on it with XP. Being involved in your test may help me get back into it from being distracted with other things. I will send you an email.

I look forward to reading your deep dive on Kodachrome scanning. .

[guyburns]

Thanks, gunzel, for the reply and for the email.

There are about 20 tapes, but I haven't been through them completely yet, so I'm not sure how much I want to digitise. I may digitise anything of interest at a low bit rate, for viewing on the computer. Then later, for material that finds it's way onto the Blu-ray, I'll digitise only that at what I consider to be a visually-lossless bit rate, to be determined by this test.

Most of the tapes are PAL, but some of the professionally-produced tapes are NTSC.

Re using a lossless codec, what file size per minute does that equate to for PAL video from tape?

[latreche34]

Quote:

Originally Posted by guyburns

This isn't about theoretical best results, this is about real world comparison. I don't mind spending a heap of dollars on equipment, but I do mind if I could have achieved the same results for 10% of the cost and bother.

This is not theoretical, it is real word facts, Using basic model composite VCR would never give you the same results as a S-VHS VCR built in LTBC via S-Video.

Again, even if we suppose the VCR has component out for VHS, You want to use S-Video or Y-C, component will be processed from Y-C in the analog domain and that will incur a loss. But I have doubts that your VCR model has component out, Having SCART does not mean every pin is active.

[traal]

Quote:

Originally Posted by guyburns

Re using a lossless codec, what file size per minute does that equate to for PAL video from tape?

HuffYUV which is lossless is about 0.5GB per minute of footage. So a $130 8TB hard drive can hold hundreds of tapes losslessly.

[aramkolt]

Quote:

Originally Posted by gunzel

I take it from the other thread that your source tapes are in PAL?

Do you have so much source material that the savings in disk space are worth testing a bunch of different compression schemes rather than just capturing uncompressed or using a lossless codec?

I have a Philips VR1600, an ATI 600 USB, and a Canopus ADVC-100, using an old Mac running Snow Leopard for capture currently but have been looking at getting Boot Camp on it with XP. Being involved in your test may help me get back into it from being distracted with other things. I will send you an email.

I look forward to reading your deep dive on Kodachrome scanning. .

What drives are you using for the Mac with the ATI600 USB? And what capture program while using it? Thanks!

[gunzel]

Quote:

Originally Posted by aramkolt

What drives are you using for the Mac with the ATI600 USB? And what capture program while using it? Thanks!

I neglected to add “clone”, not sure if that makes a difference. I think I’m just using VideoGlide. It’s been a little while and I am travelling, I will check next week and come back.

[guyburns]

Thanks for the responses. Two fellas have signed up for testing, using several different bits of gear on Mac and PC.

If anyone else is interested, forget the idea of captures at different bit rates. Just one lossless capture is required if you're able to do that. I can downgrade it from there.

[Gary34]

I’m relatively inexperienced also but to me this whole test seems unnecessary. This is 2023. We have some advantages over people digitizing there videos earlier like cheap storage hard drives, faster editing computers, and the fact that the pros and cons of this gear have already been examined by people that know what they are doing. Obviously we have a bunch of disadvantages too but ow well we don’t have time machines. Mostly disadvantages in capture. Anyways why not just look into what latreche34 is saying and look back into the forum for your information unless you just like testing.

[lordsmurf]

Quote:

Originally Posted by guyburns

This isn't about theoretical best results, this is about real world comparison.

What does that even mean?

In the real world, you access the Y/C data recorded onto the tape. That's been possible for home users since the 1980s, using non-crap VCRs, using s-video (separated video, referring to separation of Y and C, the luma and both chroma).

Quote:

I don't mind spending a heap of dollars on equipment, but I do mind if I could have achieved the same results for 10% of the cost and bother.

Don't falsely compare the prices between fully refurb'd quality gear, and a POS bought from a thrift store (that a kid stored his candy and LEGO in pre-internet).

With a 10% comparison, I can guarantee you won't get "the same" results, or even close to it. There can be instances, where certain sources, allow for lesser quality gear, for half the price, maybe a bit less. But not 10 times less.

Good video gear should have a budget comparable to any of quality tool, such as a laptop or desktop computer, or a lawnmower. Not a cheeseburger, not a month of Netflix.

Quote:

I don't know how you capture component from a VCR. The player has component output; the Hauppauge has component input, so I assume it can be done.

Component output is up-processed internally, not passed native (or near-native) Y/C, nor simple composited from Y/C. It is always sucks on cheap (sub-$500) consumer devices. Internally, some low-end off-the-shelf part. Very often -- in fact almost always -- the "component" output is processed internally to composite, even if Y/C input and component output. That's best the devices are all low-end that have this "feature".

[aramkolt]

If you are going through a rack mount full frame TBC as part of the test, wouldn't it be better to capture the component or SDI out if the TBC has one as opposed to the S-Video out since it would be a higher bandwidth signal than S-Video from the temporary digital version in the digitally stored frame?

I know rack mount TBCs aren't on the recommended list, but if people have them to test, seems reasonable to do so.

[lordsmurf]

Quote:

Originally Posted by aramkolt

If you are going through a rack mount full frame TBC as part of the test, wouldn't it be better to capture the component or SDI out if the TBC has one as opposed to the S-Video out since it would be a higher bandwidth signal than S-Video from the temporary digital version in the digitally stored frame?

No.

Quote:

I know rack mount TBCs aren't on the recommended list, but if people have them to test, seems reasonable to do so.

Just realize many of us conducted those tests 10,15,20,25 years ago already. You're just recreating what we did a lifetime ago.

[Gary34]

Svideo is really what you want for VHS. It prevents crosstalk by separating your luma and chroma. Also your line TBC needs the luma separated for timing and the chroma for DNR. Composite VCRs with TBCs never worked out well.

[aramkolt]

Quote:

Originally Posted by lordsmurf

No.
Just realize many of us conducted those tests 10,15,20,25 years ago already. You're just recreating what we did a lifetime ago.

100% agree that I am waaaaay late to the party in wanting to test these things now, but I think that's kind of what the original poster is wanting to do - Convert via a variety of methods and give a better *visualization* of the advantages and disadvantages of different capture chains in short video clips with only one variable changed at a time.

In a vacuum, individual captures could look great, but without a side by side comparison to something else, it can be much less clear as to "could it be better?" and to determine *visually* what the superior chain is excelling at specifically and whether that justifies any increased costs. Say there's just a hue shift that can be fixed with a proc amp (either hardware or in software), that's probably not a great reason to exclude that chain as a contender unless it's also not superior in anything else.

The true test I think would be to post unlabeled samples of different popular and some less popular chains all using the same starting VCR/tape and have people determine for themselves what is visually the best to them so that any bias on equipment choice is removed. Such a test is difficult though because it everyone has different goals for capture such as archival lossless vs a one time capture with fewest steps/complications (such as audio sync problems). You also may not visually be able to see the benefit of using a full frame TBC in a short clip test as evidence of audio sync and image stability may not be present over such a short duration and not all chains have good capture statistics.

Quote:

Originally Posted by Gary34

Svideo is really what you want for VHS. It prevents crosstalk by separating your luma and chroma. Also your line TBC needs the luma separated for timing and the chroma for DNR. Composite VCRs with TBCs never worked out well.

Agree at the end of the chain we don't want crosstalk and S-Video does help with that. I would argue that there's even less potential for crosstalk if the signal is digital (SDI) or component out of a full frame TBC if you're using one. You'd probably want to still want S-Video into the full frame TBC assuming the VCR's internal Y/C filter is better than the TBC's internal one.

There was a lot of talk back in the day about how laserdisc players had wildly variable image quality of their S-Video output compared to composite, and in many cases an external YC comb filter to get to S-Video worked MUCH better. Though I get it's not an exact apples to apples comparison.

The internet doesn't seem to agree if regular VHS is encoded as composite or if there's some YC separation on the tope. If it truly is just composite, then S-Video performance should be determined by the quality of the YC filter - and in some cases a YC comb filter external to the VCR could do better, unless noise is added to the composite signal within the VCR on the way out?

[Gary34]

Quote:

Test Procedure
1. Choose the best-quality VHS movie he has, and decide on which 10-30 seconds to use as a Test Sequence. Opening credits are probably the best because of the lettering.

2. Record the Test Sequence six times with these file names (or whatever bit rates are available):
- Composite 5Mbps
- Composite 10Mbps
- Composite 15Mbps

- Component 5Mbps
- Component 10Mbps
- Component 15Mbps

All captures to be flat: no sharpening, colour/contrast correction, or other optional alterations to the video.

Premiere Comparison
I'll import the files into Premiere, layer and align them, and see if there is any difference in video or audio quality. Then when I get access to other VCR players and capture devices, I'll do the same thing as a comparison between players and capture devices.

I don’t think that is a very good testing method.

1. You are choosing the best quality prerecorded VHS movie you have that might be different as far as timebase errors from what you are intending to digitize. That definitely doesn’t seem like a good way to judge different TBCs.

2. Composite and component are not good for VHS. If you would use the search tool on the forum there are technical reasons why composite and component are bad for VHS. If you are dismissing all typed information as theoretical then I don’t know how much good this forum and the typed information from people with 20 and 30 years experience can help you.

3. There’s too many other variables. Everyone has such different gear, different source tapes, different computer hardware and software. It’s hard to judge any hardware like that. Also if you are judging 20 year old gear and you get it from eBay then it can’t really be used for testing because it’s condition becomes another factor. Someone needs knowledge, tools, and a reference to see if those are functioning properly.

4. You are saying to judge your vcr and capture card against recommended gear without any color corrections or optional sharpening but recommended gear can have hardware sharpening and a pro amp and if you are judging it without that then you aren’t really judging the gear. Since video capture is an art the users skill is another factor. If they get aggressive with the proc amp that can mess with things.

5. Something I ran onto when I tried capturing without a frame TBC was my capture would crash. I don’t know for sure if that was my bad software of lack of frame TBC or both. I still have my old AVIs but that’s something you can’t tell by the video samples. How frustrating it can be working without gear you need. Now since I found this site I have became super interested in everything but at that time it just stopped me and I just kind of out everything down for a really long time. That seemed like a real hassle to me.

6. This really seems pointless because this is the OPs capture card. Hauppauge - HD PVR 2 Gaming Edition High Definition Game Capture Device with Digital Audio https://a.co/d/h1Axw2u That is clearly designed for gaming. It might not even be compatible with virtualdub. You should do the same thing I did with my Intensity Shuttle and sell it on eBay. HDMI connections are not good to see on an analog capture card. Your current gear is in the awful category. I kind of had to struggle with bad gear though before I was ready to really learn. I would just try not to let it stop you. It can get frustrating working with bad gear. I can see that leading to some people not finishing there projects. Also be careful when you start buying gear. Good luck with it. The Hans Vonk project sounds interesting.

[latreche34]

Quote:

Originally Posted by aramkolt

The internet doesn't seem to agree if regular VHS is encoded as composite or if there's some YC separation on the tope. If it truly is just composite, then S-Video performance should be determined by the quality of the YC filter - and in some cases a YC comb filter external to the VCR could do better, unless noise is added to the composite signal within the VCR on the way out?

Not just the internet, all white papers, books, magazines agree that video tape is a Y-C recording format not composite, with the consumer ones being "chroma under", meaning that chroma is downconverted from x MHz to y KHz (x,y depend on the format) to fit on the tape, during playback the chroma gets upconverted back to x MHz for playback. down and up conversion incur a small loss.

[Gary34]

I don’t think anyone is going to want to hear this one but me nor anyone in this group can test a TBC. The test you are proposing surely doesn’t.

If you are going to test your shocks out on your car you are not going to just drive on a flat road. If you are going to test your cardio then you are not going to walk at a pace that is easy for you. You have to stress test. That takes certain tapes knowledge and tools.

[lordsmurf]

Quote:

Originally Posted by Gary34

I don’t think anyone is going to want to hear this one but me nor anyone in this group can test a TBC. The test you are proposing surely doesn’t.
If you are going to test your shocks out on your car you are not going to just drive on a flat road. If you are going to test your cardio then you are not going to walk at a pace that is easy for you. You have to stress test. That takes certain tapes knowledge and tools.

Correct.

I like the scocks analogy . That's the sort of BS test that a shadetree mechanic or week-one automotive student might do.

Any idiot can "test" anything. Just look at all the eBay sellers than have zero idea what a TBC even does, but they sell them as "tested". The quality and depth of testing matters. Most people don't even know how to stress test TBCs, even if they had the tools. (Buyers can be dumb too, not just sellers. Think of all the truly stupid Amazon reviews that are left, but the real problem is user error.)

A large part of stress testing is having an extensive library of tapes, and knowing details on those tapes (age, on what camera/VCR made, upstream source, etc). You can't just have/buy some random tapes, and make a quality test.

Lots of people will claim "worked for me!", but it may fail for the other 99% of the population. Worse yet, in many cases, the person uttering that phrase was blind to the obvious errors. We see that on Youtube a lot, where those slipshod "how to convert VHS" videos show awful quality, and the clueless Youtuber dismisses it with technobabble nonsense. Far too many items aer claimed to be "TBCs", but many actually are not, and many have zero TBC functionality of any kind (placebo effect).

Quote:

Originally Posted by latreche34

Not just the internet, all white papers, books, magazines agree that video tape is a Y-C recording format not composite, with the consumer ones being "chroma under", meaning that chroma is downconverted from x MHz to y KHz (x,y depend on the format) to fit on the tape, during playback the chroma gets upconverted back to x MHz for playback. down and up conversion incur a small loss.

^ This.

VHS/S-VHS is 100% physically Y/C encoded on the magnetic tape, no exceptions, zero debate to be had here. And there general rule is best extraction will also be Y/C (via s-video DIN cables, aka separate video, because it separates luma from chroma) -- with a few exceptions, oddball, rare, fluky.

Quote:

Originally Posted by Gary34

1. You are choosing the best quality prerecorded VHS movie you have that might be different as far as timebase errors from what you are intending to digitize. That definitely doesn’t seem like a good way to judge different TBCs.

The most damning element here is that many (most?) retail tapes were contact duplicated, not recorded, thus shielding it from typical timing errors. Make no mistake, TBC is still need, timing errors do exist, as that's the nature of VHS. But many kinds of standard errors will be absent. So it's not a test, beyond some simple initial testing (ie, does anything exist at all, maybe, perhaps?).

Quote:

4. You are saying to judge your vcr and capture card against recommended gear without any color corrections or optional sharpening but recommended gear can have hardware sharpening and a pro amp and if you are judging it without that then you aren’t really judging the gear. Since video capture is an art the users skill is another factor. If they get aggressive with the proc amp that can mess with things.

That's actually backwards, too. Lots of low-end gear foists "corrections" (often making it worse), without yuor ability to change it. At least most better gear allows to to adjust the levels of filtering for certain aspect (sharpness, NR, etc). All decks have some % of forced corrections (because VHS and even S-VHS has a baseline of suckiness).

The reason to own the better gear is due to the presence and the quality of these various value adjustments (or non-adjustments). It's a main reason that I think buying a quality JVC deck, then turning off all filters (ie, EDIT mode) is truly stupid. I can understand EDIT in some situations, I use it myself. But not always-on.

Quote:

5. Something I ran onto when I tried capturing without a frame TBC was my capture would crash. I don’t know for sure if that was my bad software of lack of frame TBC or both. I still have my old AVIs but that’s something you can’t tell by the video samples. How frustrating it can be working without gear you need. Now since I found this site I have became super interested in everything but at that time it just stopped me and I just kind of out everything down for a really long time. That seemed like a real hassle to me.

^ This!

Lots of videeo newbies want to "see" what a TBC does. But some have a hard time comprehending that you're not supposed to "see" a frame TBC functioning. It's supposed to ne transparent (ie, the transparency I often refer to), and make capturing hassle-free. You can only "see" a lack of TBC, all the dropped frames, audio skews, etc. And experience the PITA that makes you hate capture. Or for the goobers in society, what they'll wrongly refer to as "VHS quality" instead of the more accurate "user error quality" (or cheapskate quality, lazy quality, etc).

Line TBC is for visuals.
Frame TBC is for the signal integrity, allowing capture, the on-visuals.

I always ask this: You breathe air/oxygen, right? Can you see it? No? Then you don't need it, right?
(Right? Oh crap, he stopped breathing. Medic! )

Quote:

6. This really seems pointless because this is the OPs capture card. Hauppauge - HD PVR 2 Gaming Edition High Definition Game Capture Device with Digital Audio https://a.co/d/h1Axw2u That is clearly designed for gaming. It might not even be compatible with virtualdub. You should do the same thing I did with my Intensity Shuttle and sell it on eBay. HDMI connections are not good to see on an analog capture card. Your current gear is in the awful category. I kind of had to struggle with bad gear though before I was ready to really learn. I would just try not to let it stop you. It can get frustrating working with bad gear. I can see that leading to some people not finishing there projects. Also be careful when you start buying gear. Good luck with it. The Hans Vonk project sounds interesting.

Correct, wrong tool. Don't slap at nails (long metal hardware that you drive into wood/etc) with a fingernail file. Both are for "nails" and yet not the same. Video conversion (VHS) and video games (consoles, whatever) may both have "video" in the name, but not at all the same.

Quote:

Originally Posted by aramkolt

There was a lot of talk back in the day about how laserdisc players had wildly variable image quality of their S-Video output compared to composite, and in many cases an external YC comb filter to get to S-Video worked MUCH better. Though I get it's not an exact apples to apples comparison.

Laserdisc is a composite format, and the s-video output had to have Y/C separation. The quality of that separation depends on the filtering quality, and many sucks. Very similar to how VHS Y/C to component/HDMI is complete crap 99%+ of the time. Composite often bad too.

Quote:

The internet doesn't seem to agree if regular VHS is encoded as composite or if there's some YC separation on the tope. If it truly is just composite, then S-Video performance should be determined by the quality of the YC filter - and in some cases a YC comb filter external to the VCR could do better, unless noise is added to the composite signal within the VCR on the way out?

"The internet" is the general population, and many of them are so stupid that I wonder how they're able to breathe. You must learn how to vet sources.

VHS is Y/C, period, the end. Not composite whatsoever.

Anybody that claims otherwise is a moron that should be wholly ignored. Because when I person says one stupid thing, without correction, odds are they'll say a lot more stupid things. Lead you down the wrong path entirely. Most people double down on stupid, they never admit fault/wrongness. Too many "experts" in this world. More people need to simply admit when they have no clue. (I don't know a lot of things, and thus never write about them, no Youtube videos about those topics. Much easier that way!)