Comparing two different 8mm (video8) capture attempts & different players

[Marvolo]

Hello from Germany,

In the following I would like to compare 2 different capture attempts, which are 15 years apart.
Of course, this also means that the Video8 tapes had not been touched since the first capture in 2008 until the new attempt a few weeks ago.

The first capture in 2008 was done by a specialist company because the service we first contacted could not play our tapes. Allegedly, the tapes had been recorded with an incorrectly adjusted camcorder, so that normal players could no longer play them. Therefore, we were told to turn to a special company that could play all the tapes.

I can't say which device was used by that specialist in 2008. It was a professional television technician who presumably had correspondingly high-quality equipment at his disposal.

This screenshot here shows what can be seen at the beginning of each of his recordings - perhaps this allows conclusions to be drawn about the equipment he must have used. Before each recording I have a grey screen with purple spots for a few seconds until the tape starts and plays. Whenever there are empty spaces on the tape the screen goes grey as well until the next recording begins.


[Attachment 70917 - Click to enlarge]

With Digital8 camcorders this screen would be rather blue, so I suspect that the playback device he used was probably not a Digital8 camcorder.

A few weeks ago, out of curiosity, I dug out those old tapes again that had sat in the basement for 15 years and captured them with a borrowed Digital8 camcorder, which (interestingly) was possible without any problems, at least with this tape here, even without special equipment. Here you can see the two comparisons:

Screenshot 1 shows the capture attempt from 2008 with unknown (presumably) special (professional) equipment, screenshot 2 shows the attempt with a Digital8 camcorder via FireWire.

Both recordings (2008 and now) were captured in DV-AVI.

It is noticeable that the 2023 recording is, firstly, blurrier and, secondly, the colours are much darker and richer in contrast than the 2008 recording. You can see much more detail in 2008 because the overall image is brighter and the colours are less intense.

BEWARE: the 2023 images have been cropped to hide the jitter noise at the bottom, the 2008 capture is unedited.

The 2008 recordings seem to have small horizontal / parallel lines visible which the 2023 recordings don't show, presumably because the 2023 recordings are less sharp and more blury by nature.

My question now is: how does this happen? Is it because the tapes were younger and not as heavily deteriorated in 2008 compared to now? Which shot do you think is better?

Capture 2008:

[Attachment 70915 - Click to enlarge]

Capture 2023:

[Attachment 70916 - Click to enlarge]

Capture 2008:

[Attachment 70918 - Click to enlarge]

Capture 2023:

[Attachment 70919 - Click to enlarge]

Would it be worth capturing all the tapes again with this Digital8 camcorder as additional backup even though the image seems to be kind of worse than the 2008 attempt? Could this be a player thing? Different players produce different pictures?

Thanks

[rgr]

Yes, each player displays the image differently, sharpens differently, the colors will look different (brighter or darker).
I have several VHS players - the picture looks different on each one, eg Panasonics display a less sharp but very stable image. Toshiba sharp (too sharp), but less stable. Philips somewhere in the middle. Of course, this can be adjusted, but this is the default setting.

And I doubt that the first recording was done by a real specialist

[Marvolo]

And I doubt that the first recording was done by a real specialist

What makes you think that?

Yes, each player displays the image differently, sharpens differently, the colors will look different (brighter or darker).

So, in other words, I was kind of lucky that all the tapes were (apparently) captured with a good player which produced a far better picture 15 years ago than the Digital8-camcorder today? That means, the chances are very slim that I would be able to yield even better results than the ones from 2008, right? Unless I would try a ton of different players to see if it will get any better.

I thought that the DV-AVI file format was not ideal, but back in 2008 it was just the right thing for 8mm. If there was a way to yield even better results than the ones from 2008, I would re-capture everything, but this time in a lossless format for longtime archival purposes... But the question is whether the results would get any better than the ones I already have from 2008 in DV-AVI and if recapturing was worth the effort... That said, the tapes have aged 15 years by now. Maybe they won't be as good as in 2008.

[rgr]

And I doubt that the first recording was done by a real specialist

What makes you think that?

Poor deinterlacing (if any). (But later I read that it is the original DV-AVI file).

Yes, each player displays the image differently, sharpens differently, the colors will look different (brighter or darker).

So, in other words, I was kind of lucky that all the tapes were (apparently) captured with a good player which produced a far better picture 15 years ago than the Digital8-camcorder today? That means, the chances are very slim that I would be able to yield even better results than the ones from 2008, right? Unless I would try a ton of different players to see if it will get any better.

The first image is just sharper. You can do it yourself.

I thought that the DV-AVI file format was not ideal, but back in 2008 it was just the right thing for 8mm. If there was a way to yield even better results than the ones from 2008, I would re-capture everything, but this time in a lossless format for longtime archival purposes... But the question is whether the results would get any better than the ones I already have from 2008 in DV-AVI and if recapturing was worth the effort... That said, the tapes have aged 15 years by now. Maybe they won't be as good as in 2008.

Capture from 2008 is good enough, you'll get a better improvement in post-processing than re-capturing.

[dellsam34]

Given the nature of tapes being Video8 which means analog, Yes you can do better than both captures, You would need a camcorder with S-Video out and built in line TBC, Capture lossless AVI 4:2:2 720x576 for archival and de-interlace then encode to h.264 for playabck.

Screenshots can never tel the whole story, If you want an opinion about those captures you would need to upload two unaltered samples of the footages at the same point of time.

[Marvolo]

and de-interlace then encode to h.264 for playabck..

Is de-interlacing even recommended? Given that video8 (such as all analogue formats) is an interlaced format and has been recorded in interlace mode, wouldn't it be better to leave it like that and let the respective player and/or TV do the interlacing? All modern Smart-TVs or file players would know how to do a decent de-interlacing and up-scaling to 1080p or even 4K.

Isn't it wiser to do as little changes as possible to the natural features of analogue video? The only thing I'd probably do is capturing in loss-less to avoid digital compression. I'd like to keep the analogue stream as natural and untouched as possible for the master files which might at the same time also serve for playback.

[dellsam34]

You can leave the playback version interlaced but no player beats QTGMC de-interlacing, Most modern cheap TVs do poor job at displaying SD interlaced sources, I highly doubt a 100GB/hr is manageable for sharing and playback, Master files are master files and playback files ar as such.

[rgr]

+ denoising + sharpening

[dellsam34]

A proper capture does not need much of the destructive post software processing, Nothing you do in software is harmless, Sometimes it is needed to address an issue but for a clean analog source, de-interlacing is as far as I can go in my opinion because in itself is a big loss, I've seen over processed files and they look like plasticy animation.

[lollo]

de-interlacing is as far as I can go

Without appropriate parameters setting, a QTGMC() call alone already denoises and sharpen in a significant way

[rgr]

de-interlacing is as far as I can go in my opinion because in itself is a big loss, I've seen over processed files and they look like plasticy animation.

Every analog player does some image processing anyway, and without any control (or with little). Of course, you can improve the video, you can also spoil it.

[rgr]

QTGMC() call alone already denoises and sharpen in a significant way

I recently tried to set QTGMC to sharpen the video, with little success. Only unsharp (ffmpeg) on default settings gave the desired effect.

[lollo]

QTGMC() call alone already denoises and sharpen in a significant way

I recently tried to set QTGMC to sharpen the video, with little success. Only unsharp (ffmpeg) on default settings gave the desired effect.

Strange. On analog captures I find the denoise and sharpening of QTGMC() quite evident. At the point that adding a better denoiser (TemporalDegrain2) and a better sharpener (LSFmod) after it, without taking precautions, introduces plastic look and enhaces halos. YMMV.

https://imgsli.com/MTc3NjE0

[dellsam34]

On fast setting it will be minimal, but still as I said it is a lossy process on its own.

[oln]

Is it a sony digital8 camcorder you were using or a hitachi one?. They 2008 caps do have signs of the typical chroma bug on the right side that the newer sony video8/hi8/digital8 camcorders had, though can't say for certain. The picture below the head switch line isn't straightened up with the TBC in these tends to be able to so maybe a variant without TBC or with TBC off. (Would make sense that they used one of the less featured ones for messing with alignment I guess.) Suppose DNR on/off and EDIT setting could have some impact even though I've never seen much difference with those on/off.

[lollo]

On fast setting it will be minimal, but still as I said it is a lossy process on its own.

The opposite. Sharpening will be slightly slightly higher, because is less accurate. Deinterlacing will be lower in quality for moving objects, and denoise will be less effective (the frame I used for comparison is quite static and not very noisy).

https://imgsli.com/MTc3NjYz



Or, at a first glance, not much difference for denoise and sharpening, just lower deinterlacing quality

edit: if you have samples where this is not true, it will be interesting and instructive to see them, please, if you can upload some

[oln]

You can adjust the degree of sharpening in QTGMC with the sharpness parameter as well.

[Bwaak]

https://imgsli.com/MTc3NjYz


[Attachment 70952 - Click to enlarge]

Or, at a first glance, not much difference for denoise and sharpening, just lower deinterlacing quality

I don't see much difference. I would shift chroma one or two pixels up and a couple of pixels to the right.

[Alwyn]

A proper capture does not need much of the destructive post software processing, Nothing you do in software is harmless, Sometimes it is needed to address an issue but for a clean analog source

Depends almost entirely on the tape. I use Neat Video (noise reduction) on most of my captures because it drastically improves the look without being plasticy. Sky, walls, consistent, even surfaces such as cars and aeroplanes all have zero noise in the real world and that's how they should be restored-to-original after capture.

[Marvolo]

Is it a sony digital8 camcorder you were using or a hitachi one?. .

The 2023 capture was made using a digital8 camcorder via FireWire. 2008 I don't know what kind of equipment they used. I thought, that black screen with the purple specks might have been a clue. As far as I know, for digital8 camcorders those screens were straight blue, not grey. Maybe it even was a stationary Video8 or Hi8 player they used??

[Marvolo]

Hi again,

OK, so maybe I did a dumb thing - I only just discovered some posts here about capturing 8mm lossless - but I purchased a Digital8 Camcorder on Ebay, the modell is DCR-TRV 730E (it has both DV-IN and Out as well as Analogue In and Out - so, could be used to capture from analoge devices using the camcorder as a converter).

However, I now read that the DV output isn't really so great for capturing analogue. I mean, even those 2008 files (that actually look great to me compared to the 2023 one) were captured in DV-AVI - but am I really going to get even better results recapturing those tapes NOT using a DV stream?

At this point, I'm really determined to go for the maximum / best quality outcome. Then again, it's been 15 years since the first capture and I don't know if I might ever get close again to the 2008 quality due to deterioration of the tapes over time.

But oh well.. That camcorder will arrive next week but actually, I was going to capture in DV-AVI again until I came across these threads saying that this is actually not the best way for analogue material.

Now I'm rather confused...

[lollo]

I don't see much difference. I would shift chroma one or two pixels up and a couple of pixels to the right.

Yes, using fast option in QTGMC compared to default settings does not reduce sharpening in this video.

You can adjust the degree of sharpening in QTGMC with the sharpness parameter as well.

Correct. I was just comparing the outputs without acting on sharpness parameters.
In general I use something like:

QTGMC(preset="slow", matchpreset="slow", matchpreset2="slow", sourcematch=3, tr1=2, tr2=1, NoiseTR=2, sharpness=0.1)

or some variant of it, based on some ideas I used from a discussion with Skiller.

Now I'm rather confused...

For best quality, Analog should be captured YUV 4:2:2 lossless. No compression and best fitting color space, ideal for further restoration.

The DV route is inferior to lossless for intrinsic quality (DV compressio, 4:2:0 color space), but it is ok if no restoration is planned and is much easier to implement.
Here a comparison from highly knowledgeable user Brad, many others across the forums:
https://forum.videohelp.com/threads/376837-DV-conversion-vs-lossless-capture-comparison-(Hi8)-[WARNING-auto-load!]

[Marvolo]

[QUOTE=lollo;2690115]

For best quality, Analog should be captured YUV 4:2:2 lossless. No compression and best fitting color space, ideal for further restoration.

This means, I must not use the DV out of my digital8 camcorder, but connect it via S-Video to some sort of not yet existing capturing device / card?

[Alwyn]

This screenshot here shows what can be seen at the beginning of each of his recordings - perhaps this allows conclusions to be drawn about the equipment he must have used. Before each recording I have a grey screen with purple spots for a few seconds until the tape starts and plays.

Arrh Ha. I am doing some Video8 to DV on my Digital8 camcorder and that is the pattern I see at the start of my captures as well.

[Marvolo]

This screenshot here shows what can be seen at the beginning of each of his recordings - perhaps this allows conclusions to be drawn about the equipment he must have used. Before each recording I have a grey screen with purple spots for a few seconds until the tape starts and plays.

Arrh Ha. I am doing some Video8 to DV on my Digital8 camcorder and that is the pattern I see at the start of my captures as well.

Interesting. I thought they were always blue for Digital8 camcorders. Well, whatever..

[Alwyn]

Yes, its starts off with that pattern then changes to blue (with a message "D8>8/Hi8" or similar' that'd be the mode change. I suspect that it only occurs when it's not a D8 tape.

[lollo]

This means, I must not use the DV out of my digital8 camcorder, but connect it via S-Video to some sort of not yet existing capturing device / card?

Yes, if you want to capture from Analog Y/C signal, you connect via S-Video your player and your (future) capture card.

In post #5 latreche34 already gave you indications. (basically good player with Y/C output and TBC and a good capture card; if needed add an external TBC or a specific DVD-R Recorder in pass-through mode)

[Marvolo]

Hmm, all these additional costs for capture card, TBCs and whatelse not. I thought, I could just run them through my D8 camcorder and call it a day. The camcorder itself wasn't cheap even. They still make good money these days on ebay with them.

[lollo]

I would shift chroma one or two pixels up and a couple of pixels to the right.

I see a degradation doing that (also on the green logo). What am I missing?

https://imgsli.com/MTc4MzQ2


[Attachment 71002 - Click to enlarge]

That red bleeding can be removed with L=-4, but the rest of the image will be badly impacted (my conclusion was that is not a full chroma shift, but just a bleeding). I may be wrong.

Here a small sample of the raw capture, if you wish to experiment:

spots_end_uons_amtv_v2_cut.avi

[lollo]

Hmm, all these additional costs for capture card, TBCs and whatelse not. I thought, I could just run them through my D8 camcorder and call it a day. The camcorder itself wasn't cheap even. They still make good money these days on ebay with them.

Start with a good player with integrated TBC and one of the recommended capture cards (I-O Data GV-USB2, Hauppauge USB-Live 2, Pinnacle 710 USB, ATI 600 USB, few others).

One step at the time.

Then add other devices only if needed. A specific DVD-R Recorder is not expensive, an external TBC is rare and very expensive.

Add results of your captures here once ready, to receive support from many experienced users