Capture Comparisons - w/ and w/out ES-10, HDMI, SVHS and Combi VHS

[IamScrooge]

Hi,
Firstly many thanks to those who helped me with advice when I joined - and to the wider community for the wealth of knowledge I have been trawling through on this forum.

I had a Panasonic DMR-EZ45V which, for years, I thought I'd use when I eventually had time to take up a project of transferring some old VHS tapes - but after research I decided to pull the trigger on some new old kit from eBay for better results.

I made a video of comparisons between different combinations of the gear I have to help decide which method I'd ultimately use and, in the hope it will be helpful to others, have decided to share it here.

The two decks I'm using are:
Panasonic DMR-EZ45V which I've had since it was new
JVC HR_S5967 which isn't part of the 6000 series up as recommended but those units are rarer and expensive - and also a piece of advice on this forum said I'd need to disable the TBC anyway if I was using the ES-10
The capture device I'm using is:
Diamond VC500
And I also have a
Panasonic DMR-ES-10 which I use as a passthrough for it's "TBC-like" processing features as raved about in various posts on this forum. These are pretty cheap on eBay in my neck of the woods.
All devices are PAL.

The capture & deinterlace method I'm using is the one described here: https://www.youtube.com/watch?v=sn_TDa9zY1c although I'm not using one of his recommended video devices, as when I bought one matching the description in that video most of the configuration options in VirtualDub were greyed out and I couldn't switch out of 480 resolution mode. The Diamond VC500 which is recommended on this forum worked much better, although I found that I had to use a USB port directly mounted on the motherboard instead of one of the front ports on my computer or the device would repeatedly crash.

I also took a capture with a HDMI capture stick using the 4 different output resolutions of the DMR-EZ45V - these are unprocessed as either the machine itself or the capture stick appears to be taking care of the deinterlacing. I didn't find much visual improvement from 576i so I that's just what's used for the rest of the video.

The split screen comparisons at the end I think are the most useful part.

Anyway I hope that this is useful to someone!

[lordsmurf]

That's such a horrible Youtube video.

Those HDMI 576p captures are horrible, drop-field deinterlace, 50%+ of data is missing. Also dropped frames are obvious.
The 720p isn't drop-field, but still dropped frames.
The 1080p is stretched, and values are ghastly, lots of overexposure. Again, dropped frames are obvious.

Dropped frames are everywhere. Alias jaggies when not drop-field.

Overall, this is really lousy video.

The "SVHS" captures are more a statement about the recording VCR and the tape, not really the playback S-VHS deck. That model lacks line TBC, it's to be expected when the tape sucks.

I don't understand how you cannot see the motion flaws here. Some jerky ugly stuff in there.

Essentially, this is a perfect example of what NOT to do.

[IamScrooge]

That's such a horrible Youtube video.

Those HDMI 576p captures are horrible, drop-field deinterlace, 50%+ of data is missing. Also dropped frames are obvious.
The 720p isn't drop-field, but still dropped frames.
The 1080p is stretched, and values are ghastly, lots of overexposure. Again, dropped frames are obvious.

Dropped frames are everywhere. Alias jaggies when not drop-field.

Overall, this is really lousy video.

The "SVHS" captures are more a statement about the recording VCR and the tape, not really the playback S-VHS decl. That model lacks line TBC, it's to be expected when the tape sucks.

I don't understand how you cannot see the motion flaws here. Some jerky ugly stuff in there.

Essentially, this is a perfect example of what NOT to do.

Hi,

Yes I also thought the HDMI output was pretty bad.
The purpose of this exercise was to determine how to make the best of what I have - and to confirm for myself and illustrate some of the advice I have gathered from this forum.

The quality of the source material isn’t very good - it’s just what I was able to record off TV using the Panasonic, I’m not sure what better source I could use for these tests due to copy protection. It was good enough to illustrate quality differences in all the capture method permutations.

My intention at this point based on the outcomes from this exercise is to go ahead using the JVC deck and ES10 to capture my VHS memories using the Diamond VC500, and Virtualdub. Would you advise I do anything differently (that is, without investing a lot more money into this project)?

Thanks for your comments

[oln]

Most of the outputs of the EZ45 (and other panasonic dvd-recorder/vcr combos) go through the internal digitizer, so adding the ES10 after those won't be of any help. It's possible you would get the raw output from composite over the scart connectors judging by the manual. I don't know if the ES10 would be able to stabilize a whole lot better than the dvd-recorder part of the EZ45 though, the EZ45 is part of the 2006 lineups where the ES15 was the cheapest/simplest model. On the raw captures from the SVHS decks the stabilization of the ES10 is quite noticeable as that VCR doesn't have any TBC function of it's own.

It should be possible to get 576i from the HDMI out, though there are some hdmi capture thingies that do deinterlacing if they detect an interlaced signal.

The main difference playback wise with the higher end 7/8/9000 JVC SVHS decks and the cheaper ones is the TBC module, they otherwise use mostly the same parts for the video and system stuff. The dynamic drum system on the top of the line models doesn't do anything during normal playback. The TBC in the JVC decks isn't really able to stabilize any more than the ES10, at least in my experience (PAL) but YMMV.

[dellsam34]

For comparaisons like these always start from a 4:3 interlaced material not 16:9, anamorphic or de-interlaced, Second there is no reason to compare composite to S-Video everybody knows capturing a VHS tape with a S-VHS machine via S-Video yields better results than composite output, It's a fact that has been already proven. With that your video would have been much shorter and very few comparaisions to look at.

[lordsmurf]

using the JVC deck and ES10 to capture my VHS memories using the Diamond VC500, and Virtualdub.
without investing .. more money ?

That would be the best path forward here.
But be sure your VirtualDub settings are correct: interlaced, 720x576, 25fps (50i).

[Alwyn]

Geez you guys are brutal.

Thanks, Iamscrooge, for going to the effort of doing that. I found it quite interesting.

I found HDMI captures from my EZ-48 with the Startech USB3HDCAP were almost identical to my S-Video captures.

[themaster1]

Same here using a "chinese" timeleak HD72B pci-e card, on par with traditional caps minus some slight A/V desync from the start, nothing that can't be handled later on

[oln]

I've also encountered the av desync thing with hdmi capture on my avermedia card in Virtualdub. AmarecTV didn't give it, though it was a bit more finicky to get the frame rate correct.

I found HDMI captures from my EZ-48 with the Startech USB3HDCAP were almost identical to my S-Video captures.

Yeah the S-Video output is gonna be the same signal just converted back to analog. I haven't used any of the combo ones, but on my standalone EH57 (and for that matter on my Sony RDR-HX750), the hdmi output does not have the 8 pixels on the left/rightmost pixels blanked the analog outputs (and TBCs for that matter) have. Usually there isn't much there, but it's nice in case there is, and avoids the very sharp transition from black to image.

[IamScrooge]

Most of the outputs of the EZ45 (and other panasonic dvd-recorder/vcr combos) go through the internal digitizer, so adding the ES10 after those won't be of any help. On the raw captures from the SVHS decks the stabilization of the ES10 is quite noticeable as that VCR doesn't have any TBC function of it's own.

That is exactly what I wanted to confirm and the results match up nicely to expectations.

It should be possible to get 576i from the HDMI out, though there are some hdmi capture thingies that do deinterlacing if they detect an interlaced signal.

It looks like that's what's happening here. To be honest, the HDMI capture was just to see on the off chance how good it was compared to the tried and tested methods. As the results aren't promising from HDMI I do not intend to perform any further investigation to get better results from the device.

The main difference playback wise with the higher end 7/8/9000 JVC SVHS decks and the cheaper ones is the TBC module, they otherwise use mostly the same parts for the video and system stuff. The dynamic drum system on the top of the line models doesn't do anything during normal playback. The TBC in the JVC decks isn't really able to stabilize any more than the ES10, at least in my experience (PAL) but YMMV.

That's good to know - anything from the 6000 series up would have been 3 times the price of the deck I got so I think I've optimised my spend on this project.
Thanks for the info!

[IamScrooge]

For comparaisons like these always start from a 4:3 interlaced material

Alright, good to know for future, thanks. I don't think the interlaced video would have helped me discern the differences, I find the interlacing too distracting. I guess that with more experienced eyes there is probably more value in your approach.

not 16:9, anamorphic

Alright I'm going to admit I dropped the ball on this one - I'm not entirely sure what Davinci is outputting on it's 576p preset.
I'm learning as I'm going and I think I've realised I need to pay close attention to the output encoding.

Second there is no reason to compare composite to S-Video everybody knows capturing a VHS tape with a S-VHS machine via S-Video yields better results than composite output, It's a fact that has been already proven. With that your video would have been much shorter and very few comparaisions to look at.

I'm actually of the opinion that it was worth comparing the composite output - I already knew that SVideo was superior but wouldn't have been able to quantify the gains without comparing them. I know from a lot of experience the kind of difference RGB gives compared to Composite from a clean signal from a sattelite receiver - I did not know what kind of difference it would make from a VHS tape.
Also - I think this is the first time I've ever actually used SVideo for anything. I'm in the UK we tend to use RGB over Scart here. I think in all the time I was installing TVs as a job I only ever used an s-video connector once when I ran out of inputs.

Thanks for your comments!

[IamScrooge]

using the JVC deck and ES10 to capture my VHS memories using the Diamond VC500, and Virtualdub.
without investing .. more money ?

That would be the best path forward here.
But be sure your VirtualDub settings are correct: interlaced, 720x576, 25fps (50i).

Smashing - thanks for your help!

[IamScrooge]

Geez you guys are brutal.

Haha thanks the response was slightly colder than I expected - but I'm sure it wasn't intended like that. Still - this is a niche subject and there are few folk with the expertise I'm looking for and the willingness to share it - I appreciate the time folk have spent helping me. I guess it's not like the photography forums where folk are jumping over each other to explain for the zillionith time what aperture is

Thanks, Iamscrooge, for going to the effort of doing that. I found it quite interesting.

I definitely found it interesting and learned a lot from the exercise, even though the results were what I was expecting. I'm glad someone else did as well!