The old "war of opinion" - Capture 8mm tapes in DV or lossless?

[Marvolo]

I have bought a Digital8 Camcorder capable of playing back Hi8 and 8mm tapes. The original 8mm camcorder (a Blaupunkt CR-8000) has long died. All my 8mm tapes were captured in 2008 by a professional television engineer who captured them right into DV via FireWire / iLink, presumably using a Digital8 camcorder himself.

All those captures look very nice to me. Even when I connect the hard drive to my 4K Samsung Smart TV, they still look very sharp and nothing like VHS or any other low resolution video. Probably the TV does a very good job at upscaling it to its 4K screen.

Anyway... I'm sort of an perfectionist. That's why I thought: what if there was a way to get even better results than those from 2008? Maybe even with the help of today's state-of-the-art technology such as Topaz Enhance AI? So I delved into video restoration and came across the old war of opinions: analog video should never be captured in DV (or a lossy format) vs. 8mm or VHS is so much lower in resolution in quality than DV that it actually won't matter at all as it will never reach DV quality anyway.

So, I bought a Digital8 camcorder intending to re-capture all those tapes that have been sitting in the basement since their last capture in 2008. As of now, I have no other means than to capture them in DV again using the DV-output of the cam and my built-in FireWire card. I captured all those tapes again in DV and then compared them frame by frame with the 2008-capture. To my surprise, most of it was literally identical, despite the fact that the tapes are now 15 years older. The only slight difference I noticed is that my cam seems to have a slightly higher brightness / gamma compared to 2008. But really only slightly. But resulting in a loss of fine structures such as details of clouds that now are just plain white areas and in 2008 you could still see a little more detail inside the cloud.

Would it help to re-capture all tapes in a lossless format? For that I would have to get a capture card. I can only capture through FireWire as of now. But would it be worth the effort and expenses?

If it helps, I could try and upload both a sequence of the 2008 capture and my own from 2023. Both DV-AVI.

THanks a lot

[Bwaak]

Would it help to re-capture all tapes in a lossless format? For that I would have to get a capture card. I can only capture through FireWire as of now. But would it be worth the effort and expenses?

What do you expect in response?

Capturing as DV is about getting reliable results, while capturing through SVideo is about trying to squeeze the maximum quality. As such, it is apples and apple crates - the optimization criteria are different.

Since both opinions are well-argued, the only way is for you to assess the validity of arguments and to do your own evaluation.

[johnmeyer]

Totally agree with Bwaak. His post, especially his second paragraph is really well-stated.

Most people don't get into trouble capturing with DV, and can often really screw up when using higher-quality equipment ... but ... if they are willing to learn how to use the better equipment, they can get better results.

DV provides decent results, but it has two downsides: compression noise and limited colorspace (especially NTSC). For most people who aren't perfectionists, given that they are starting with VHS/SVHS/8mm/Hi8mm which are all pretty lousy to begin with, the damage done by DV's limitations are often the least of the problems in the capture.

I personally don't think that a lossless codec is absolutely required for consumer analog. A modern lossy codec, if used at a reasonably high bitrate, can do a pretty good job. Having some compression sure makes the workflow and final storage a whole lot simpler.

[Alwyn]

A modern lossy codec, if used at a reasonably high bitrate, can do a pretty good job.

John, could you expand on this. From my readings, the only lossy codec in this scenario (Video8/Hi8 capture) is DV, or are you suggesting capture into something like H264?

What do you expect in response?

A reasonably definitive answer from someone who has done both and can compare. That is, after all, the purpose of asking advice, so you don't make the same mistake or learn the same lessons over and over again.

[Bwaak]

What do you expect in response?

A reasonably definitive answer from someone who has done both and can compare.

I presume he already have read a bunch or reasonably definitive answers, all of which fall into two categories, which he correctly identified. What is next? Suppose someone does a test and comes up with DV looking as good or - gasp! - better than SVideo/uncompressed. The result will be dismissed on the grounds that they don't know how to capture analog video and that their A/D converter is crap - which very well may be true. The only result that anti-DV group can accept is that DV sucks. Hence, any comparison for the purpose of convincing other people is pointless, and only personal experience matters. Q.E.D.

[Alwyn]

Interesting YT video comparing VCR, analogue passthrough with DV camcorder, and VCR+TBC.

https://www.youtube.com/watch?v=UfUz69fztT0

Not exact comparisons (use of Denoise filter in the last, for example) but informative, especially the brighter brights of the DV conversion.

[Sharc]

Interesting YT video comparing VCR, analogue passthrough with DV camcorder, and VCR+TBC.

https://www.youtube.com/watch?v=UfUz69fztT0

Not exact comparisons (use of Denoise filter in the last, for example) but informative, especially the brighter brights of the DV conversion.

Maybe elevated darks for the DV, rather than 'brighter brights'. The brights are not off the kilter. No real loss of details in the DV brights. I would even prefer the DV variant in this case. The head of the dog is best for config#3 (details) though.

[Marvolo]

A reasonably definitive answer from someone who has done both and can compare. That is, after all, the purpose of asking advice, so you don't make the same mistake or learn the same lessons over and over again.

Thank you!

Anyway, seems like I will have to find out myself by getting all the equipment needed for analog capture... We'll see.

That said, speaking of video restauration, has anyone had some experiences using AI such as Topaz Video Enhance AI to restore or enhance analog video? Looking at several youtube videos, the results seem to be amazing.. Many of these vids show analog footage upscaled to 1080p or even 2k.

I'll post a snippet of the 2008 DV capture and my own DV capture for you to compare and evaluate how the overall capture result has been. All in all, I had to learn that the results may differ GREATLY just by using different D8 camcorders. I've had playback with a D8 cam that was terribly dark in picture - you could barely see anything. The same tape played just fine with my current D8 cam and it looks more or less identical to the 2008-capture.

[DB83]

Well it's your eventual call. Yet you will inevitably come across many discussed issues about analogue transfer and then having invested in that equipment you might suddenly think was it worth that effort.

Being someone who can provide you a direct comparision of a dv transfer to a lossless one I am quite prepared to do that if it helps you make a decision BEFORE you jump in.


BTW Your assumption of how the tapes were originally transfered may not be accurate. True that the 'pro' might have done it as you assume. But he equally could have done it with Hi8 camera into a ADVC box. The differences you now see in your new capture to the older one could be down to the quality of the camera, as you have noticed, or, if he used a higher-spec ADVC (the 300 is the one I have) that has some control over visual quality beit it brightness, contrast etc.

[lollo]

There are some recent comparisons between lossless and DV approach in these threads:

https://forum.videohelp.com/threads/409928-Videohelp-forum-messages-mentioned-in-YouTube#post2693235

https://forum.videohelp.com/threads/410062-Which-capture-of-my-old-VHS-looks-the-best#post2694196

That said, speaking of video restauration, has anyone had some experiences using AI such as Topaz Video Enhance AI to restore or enhance analog video?

Yes, we did several experiments in this thread:

https://forum.videohelp.com/threads/399360-so-where-s-all-the-Topaz-Video-Enhance-AI-discussion

and many other examples on the forum.

In general, there is nothing Topaz VEAI can do that cannot be matched/beaten by AviSynth/VapourSynth.

[Marvolo]

Being someone who can provide you a direct comparision of a dv transfer to a lossless one I am quite prepared to do that if it helps you make a decision BEFORE you jump in.

Please do.

BTW Your assumption of how the tapes were originally transfered may not be accurate. True that the 'pro' might have done it as you assume. But he equally could have done it with Hi8 camera into a ADVC box. The differences you now see in your new capture to the older one could be down to the quality of the camera, as you have noticed, or, if he used a higher-spec ADVC (the 300 is the one I have) that has some control over visual quality beit it brightness, contrast etc.

True, but I have strong evidence pointing to the fact that at least a D8 camcorder must have been used. I see the typical blue screens in areas where there is no signal (especially at the beginning of each tape) and other signs that are identical to my D8 camcorder. But yes, there still might be a small chance that a different player could have been used.

[DB83]

^^ Give me some hours. Have to go out now.

[Sharc]

...From my readings, the only lossy codec in this scenario (Video8/Hi8 capture) is DV, or are you suggesting capture into something like H264?

One can for example capture to DV, ProRes, slightly lossy H.264/H.265 using a low CRF in x264/x265....
Just be careful not to overload the CPU which could cause dropped frames.

[lollo]

Also (obsolete) MPEG2. Hauppauge USB-Live 2 + Hauppauge Capture software can capture with realtime compression 10mbps with good quality (although inferior to lossless) and no dropped frames with appropriate hardware. I never use it, except for testing.

[Marvolo]

So, I decided, I'd give it a try and capture analoge. The youtube video above kind of convinced me that it could be possible to get even better quality.

Since I'm in Germany, and I usually browse German video forums, I found this very detailed step-by-step guide to capture analoge. (It's in German but you can have it auto-translate by Google Translate or your browser).

I just wanted to ask if they do the right thing. They say:

1.) Proper Playback Machine
2.) Use a DVD/HDD-Recorder that works as external TBC, either Panasonic:
DMR-EH65
DMR-EHxxx (all „EH“ with 3-digit odel number)
DMR-EXxx (all „EX“ with Digital tuner)
DMR-EZ49 (VHS-DVD-Kombi without HDD)
3.) Connect DVD Recorder via HDMI cable to HDMI Splitter to remove HDCP from HDMI signal
4.) Connect HDMI Splitter to either Blackmagic Intensity Pro Card or Canopus Edius NX PCI Card as they are the ONLY TWO cards that capture according to analog specifcations.

I was going to get the Canopus Edius NX Capture Card and one of those DVD recorders listed above. They can be found relatively cheap on eBay.

I hope it'll work that way?

[lollo]

Connect HDMI Splitter to either Blackmagic Intensity Pro Card or Canopus Edius NX PCI Card

The BM card is used to capture the digital HDMI stream converted and stabilized by the Panasonic.

Canopus Edius NX captures from Analog.

Different workflows.

[Marvolo]

Yes, of course. That way I wouldn't need the Panasonic, UNLESS (so they say in the article) you would want an additional TBC in the line before the Edius NX.

[Alwyn]

Connect HDMI Splitter to either Blackmagic Intensity Pro Card or Canopus Edius NX PCI Card as they are the ONLY TWO cards that capture according to analog specifcations.

The Startech USB3HDCAP will capture HDMI at 720x576i 25fps analogue from my Panasonic EZ-48 (VHD/DVD Combo) and my Panny EH-57 (DVD/HDD recorder).


[Attachment 71823 - Click to enlarge]

Both the Pannys require one of these splitters to defeat incorrect HDCP signalling to the HDMI output:

https://www.amazon.com.au/dp/B091GG183D?ref_=pe_2361882_297267902_302_E_DDE_dt_1

For what it's worth, given the alleged "straight through" nature of the digital signal, I can't see any significant difference between the HDMI capture and a normal analogue capture with any of the GV-USB2, USBLive2, USB-710 or USB3HDCAP in analogue mode. It feels cool doing an HDMI capture but it also cost $200 (USB3HDCAP) +$25 for a splitter verses $70 for a GV-USB2.

Your mileage may vary with the flasher Blackmagic or Canopus cards.

[Marvolo]

OK, seems like I cannot use the Canopus Edius NX card as my mainboard (from 2009) doesn't have any PCI(X) slots.. Only PCI-E.

Which means, I'll probably have to go the Blackmagic Intensity & Panasonic Player route...

[Bwaak]

YT suggested this to me. Not a lot can be gleaned from it aside of the confirmation that BM Shuttle does not like unstable v-sync: https://youtu.be/loipkgzAbCE

[Marvolo]

YT suggested this to me. Not a lot can be gleaned from it aside of the confirmation that BM Shuttle does not like unstable v-sync: https://youtu.be/loipkgzAbCE

Oh, I wasn't gonna get the BM Shuttle, but the PCI-E card itself:

https://www.kleinanzeigen.de/s-anzeige/blackmagic-design-intensity-pro-4k/2470325098-225-13532

I only need the HDMI-in port anyway... No need for the big shuttle.

[lollo]

Not a lot can be gleaned from it aside of the confirmation that BM Shuttle does not like unstable v-syn

Sorry to repeat if I have misunderstood your reply, but in the proposed workflow the BM is only used to transfer the digitized signal from the Panasonic to the PC, using the HDMI output. No A/D involved. The capture card role is played by the Panasonic itself

[DB83]

Apologies for the delay.

Here is the sample lossless (huffyuv) clip. For the record - capture device = hauppauge usb-live2 - no TBC and compositeAnd then the DV version of the same sequence.

[Marvolo]

Thanks, DB83.
Unfortunately, if both files are not in one video next to each other, I find it very hard to see differences if I have to watch them seperately in a new window. They kind of look similiar to me.
You should compare them frame by frame in an NLE, turning each track on and off as they are playing.

[DB83]

Well those comparisons, if achieveable, are for you to do.

I merely provide the source.

[Marvolo]

Well, I just ordered a Blackmagic Intensity Capture Card as well as a Panasonic DVD recorder and hopefully, I will soon be able to make a comparison with my own footage between DV and lossless capture.

I will post here and show the (if any) differences.

[Sharc]

Apologies for the delay.

Here is the sample lossless (huffyuv) clip. For the record - capture device = hauppauge usb-live2 - no TBC and compositeAnd then the DV version of the same sequence.

The lossless Huffyuv looks flawed to me: Blended fields. See for example fields around #850 (attachment). Something went wrong it seems.
The DV capture looks cleaner (apart from the expected DV compression artefacts). Less field blending. So my vote goes for the DV variant for this case.
Maybe I am missing something?

Edit: Just noted that the DV is also blended. Maybe the movie should have been IVTCed at some stage instead of poorly framerate converted?
So this is just an example that the problems are usually somewhere else than with the subtle differences between DV or analog captures ...

Anyway, everything about the topic has been said in post#2 already

[Alwyn]

To be of any value in this scenario, any comparison has to be using a D8 camera, not a VCR and an ADVC.

Unfortunately, if both files are not in one video next to each other, I find it very hard to see differences if I have to watch them seperately in a new window.

A technique I often use is to use 2 x Virtual Dub (or you could use 2 x VLC player, although frame-stepping with VLC is harder). I open each video in a separate instance, then drag one left to fill the left half of screen and drag the other right to fill the right side (or use Win 11's layout gadget at the top of screen). Use View>Pane layout>Input pane only. Then both videos will be exactly the same size. Then just drag the cursors around to match frames.


[Attachment 71845 - Click to enlarge]

[Sharc]

To be of any value in this scenario, any comparison has to be using a D8 camera, not a VCR and an ADVC.

Agree.

[DB83]

Well I do not agree. The OP was seeking a relative comparison between analogue/Lossless and DV. True that it merely confirmed what was already written >> higher colorsubsampling, bitrate etc. The method of creating the clips, as long as one did not 'cheat' by, for example, merely converting the DV to lossless surely is not relevant.