Rate my Setup - Capturing Digital8 and Analog

[Bermuda1]

Hi everyone,
I am new to capturing and I wanted to show you the setup that I use to digitize old family Digital8, HI8 and Video8 cassettes. Since I am from Germany, most of the videos (and camcorder) should be PAL.

Digital 8 (and for the sake of comparing the quality of the captured video also HI8 and Video8):
- Camcorder: SONY TRV 230E
- ASHATA 2.5Gbps PCI-E PCI Express FireWire 1394a IEEE 1394 Controller Card with FireWire Cable 800Mbps (https://amzn.eu/d/eOVmvCN)
- ScenalzerLive 4.0

Note: albeit several different accounts and lists that I found, my version of the TRV 230E is able to convert the analog signal from the HI8 tapes to digital so that I can capture via FireWire.

Analog tapes (HI8 and Video8)
- Camcorder: SONY TRV 230E
- WIN TV Hauptauge WIN-TV-HVR-1200 TV Card
- Goobay 50057 S Video Cable (https://amzn.eu/d/dJXV3It)
- Composite / RCA cable that has the yellor, red and white connectors on one side and the 3,5mm on the other
- VirtualDub

Note: to connect the camcorder to the TV Card, I use the S-Video and the RCA cable (3,5mm goes into the camcorder, white and red into the TV card)

So my first question (and only question so far) is: is there any significant flaw in the hardware that I am using / is it garbage ?

Thanks!

Bermuda1

[Bermuda1]

No one ? Planning to start the s-Video capture soon and would be good to know if it’s worth the time with my hardware or I am missing something. Thanks!

[Alwyn]

If you're in PAL land, little will be gained by using a analogue-lossless process. One of the respected commercial digitising outfits, video99uk, allegedly uses the DV route: playing the V8/Hi8 tapes in the D8 camcorder and capturing the DV. My own experiment showed the two methods yield the same result.

I suggest you do a couple of test runs to assess for yourself.

For the analogue process, yes, connect as per your note.

[Sharc]

If you're in PAL land, little will be gained by using a analogue-lossless process. One of the respected commercial digitising outfits, video99uk, allegedly uses the DV route: playing the V8/Hi8 tapes in the D8 camcorder and capturing the DV. My own experiment showed the two methods yield the same result.

I suggest you do a couple of test runs to assess for yourself.

For the analogue process, yes, connect as per your note.

I agree with your recommendation, however I have some doubts with your example. Look at the white fence in the background: totally mushed for the analog capture, all details lost. I really wouldn't have expected this. Any explanation as to why this happened?

[Bermuda1]

If you're in PAL land, little will be gained by using a analogue-lossless process. One of the respected commercial digitising outfits, video99uk, allegedly uses the DV route: playing the V8/Hi8 tapes in the D8 camcorder and capturing the DV. My own experiment showed the two methods yield the same result.

I suggest you do a couple of test runs to assess for yourself.

For the analogue process, yes, connect as per your note.

I agree with your recommendation, however I have some doubts with your example. Look at the white fence in the background: totally mushed for the analog capture, all details lost. I really wouldn't have expected this. Any explanation as to why this happened?

Thanks for the replies so far!
So you are saying the DV capture route may be better than capturing via s-video?

[Sharc]

If you're in PAL land, little will be gained by using a analogue-lossless process. One of the respected commercial digitising outfits, video99uk, allegedly uses the DV route: playing the V8/Hi8 tapes in the D8 camcorder and capturing the DV. My own experiment showed the two methods yield the same result.

I suggest you do a couple of test runs to assess for yourself.

For the analogue process, yes, connect as per your note.

I agree with your recommendation, however I have some doubts with your example. Look at the white fence in the background: totally mushed for the analog capture, all details lost. I really wouldn't have expected this. Any explanation as to why this happened?

Thanks for the replies so far!
So you are saying the DV capture route may be better than capturing via s-video?

No. The DV route is easier though as it has much less pitfalls. In fact the analog (via S-video) route - if done correctly - should yield slightly better results. Most people won't notice the difference though.

[Sharc]

@Alwyn: apart from the fact that the brights are crushed and clipped (source probably overexposed) which destroys any details in those areas it seems to me that the 2 videos are not properly synched. They are 1 or 2 frames out of sync. So we don't compare the same frames side-by-side and one may easily draw wrong conclusions, e.g with the fence.
In order to demonstrate the subtle differences between the DV and the analog S-video route one should actually use a higher quality video (which has not been additionally butchered for YT upload.)
Anyway, as I said I agree with your statement for PAL. The DV route is at least much more fool proof for less experienced users than the S-video route with its many pitfalls.

[oln]

If you are capturing analog tapes via firewire on a d8 camcorder, make sure to change the audio mode on the camera to 16-bit/48000 khz, it's usually set to 32000 khz by default which is not optimal.

IMO Main thing you can gain from the analog method for analog tapes over firewire is avoiding the DV compression which can be beneficial on tapes with a lot of noise, e.g tapes filmed in low light conditions which can otherwise result in some noticeable compression artifacts. I think it's generally worth the effort but since you have the setup you can easily compare yourself. I also don't know for sure how well the video levels are handled by the internal analog->DV converter whether you risk clipping or not.

I don't think the halved vertical color resolution on DV is much to worry about, there is already a loss of vertical color resolution from the comb filtering during the vcr process, and unless it's from some absolute top of the line professional 3 CCD camera the resolution of the source material won't have full color resolution to begin with due to low resolution sensors and bayer filters.

Digital8 tapes should of course be ingested via firewire since those are already in DV format so by using firewire you getting bit for bit what is on the tape and using analog output would just lose quality.

[Bermuda1]

Many thanks to all of you - this has been really helpful. Now, I checked a few posts and googled a bit on what settings to use in VirtualDub. Since I am capturing PAL tapes, I am not sure if the recommended settings I found always apply. So there is a quick list of settings and I would be so very thankful, if you could take a look and let me know if this works (see screenshots attached).

Best regards,
Bermuda1

[lordsmurf]

Many thanks to all of you - this has been really helpful. Now, I checked a few posts and googled a bit on what settings to use in VirtualDub. Since I am capturing PAL tapes, I am not sure if the recommended settings I found always apply. So there is a quick list of settings and I would be so very thankful, if you could take a look and let me know if this works (see screenshots attached).

It seems fine, but I question editing the chunk size setting.

Timing settings also matter per-card. Never uncheck the top two boxes on the timing settings, all that does is disable dropped frames reporting (and NOT prevent dropped frames).

- WIN TV Hauptauge WIN-TV-HVR-1200 TV Card

This is a known-problem card for capturing Hi8/Video8 or VHS. Those PVR style HD cards were not made for SD analog capture.

I suggest you do a couple of test runs to assess for yourself.

But he came here to ask for our advice, not be told "learn for yourself".

No. The DV route is easier though as it has much less pitfalls. In fact the analog (via S-video) route - if done correctly - should yield slightly better results. Most people won't notice the difference though.

It's more than just slight. DV literally tosses out 50% of color data for NTSC, and 25% for PAL -- plus adds blocks. It's 1990s video tech, back when a Pentium III was the suggested specs (Pentium II minimum). It also has lots of odd tint changes to colors. This is all obvious on modern large HDTVs, only tiny preview windows or mobile/cell phones appear to be "fine".

[Bermuda1]

Many thanks to all of you - this has been really helpful. Now, I checked a few posts and googled a bit on what settings to use in VirtualDub. Since I am capturing PAL tapes, I am not sure if the recommended settings I found always apply. So there is a quick list of settings and I would be so very thankful, if you could take a look and let me know if this works (see screenshots attached).

It seems fine, but I question editing the chunk size setting.

Timing settings also matter per-card. Never uncheck the top two boxes on the timing settings, all that does is disable dropped frames reporting (and NOT prevent dropped frames).

I will change the timing settings. What chunk size setting would you recommend? I just took the value I found recommended in another thread.

- WIN TV Hauptauge WIN-TV-HVR-1200 TV Card

This is a known-problem card for capturing Hi8/Video8 or VHS. Those PVR style HD cards were not made for SD analog capture.

When you say "problem card" are you saying that there is an issue with the resulting quality or more of a problem with compatability (quality fine if you get it to work?). Or what is the problem you are referring to?

Since the alternative (until I find another tv card) is DV/FireWire capture, would you say use DV/FireWire or the problematic TV card to get the best results (given the circumstances)?

Thanks,
Bermuda1

[Bermuda1]

So my first results are in. See the screenshots. I find that the S-Video has better colors for sure and I feel like stuff in the background is sharper! But in some of the video clips there is some sort of "sway" - not sure how to descibe it - almost a little lag when in motion (see the close-up screenshots). I dont have that with the FireWire Capture.

Based on the screenshots, the FireWire capture seems a bit "broader" compared to the S-Video. I would actually say that that FireWire one seems correct when looking at faces. Any way I can verify that?

Other than that there are parts of the S-Video Capture where the sound is off: really slow, really sped up, or a high pitch noise occurs. Other parts are normal.

Any idea what the reason for this may be?

[lordsmurf]

Interlacing.

[Bermuda1]

Interlacing.

Ok, thanks for the hint. Read up on that so it’s fine. Any chance you can help me out on or have an idea how to fix the audio issue and aspect ratio issue (seems like the S-Video is too cramed/condensed although both files are 720x576). Also what setting is better for the chunk size?
Thanks again, all of you and especially lordsmurf have been really helpful and I’m going up the learning curve

[Sharc]

.... or have an idea how to fix the audio issue and aspect ratio issue (seems like the S-Video is too cramed/condensed although both files are 720x576).

The analog S-video capture is not square pixel but has a PAR (Pixel Aspect Ratio) of approx. 12:11 (aka 59:54), given by the Rec.601 standard. Crop the width to 704 (~cropping the side bars off) and encode/display the 704x576 frame as 4:3.

On a sidenote: Turn sharpeners down to avoid halos

[Bermuda1]

.... or have an idea how to fix the audio issue and aspect ratio issue (seems like the S-Video is too cramed/condensed although both files are 720x576).

The analog S-video capture is not square pixel but has a PAR (Pixel Aspect Ratio) of approx. 12:11 (aka 59:54), given by the Rec.601 standard. Crop the width to 704 (~cropping the side bars off) and encode/display the 704x576 frame as 4:3.

On a sidenote: Turn sharpeners down to avoid halos

Sharpness setting is only at 3 in VirtualDub.
Other than that, can the changes suggested by you be done before capturing in VirtualDub or do I need other software? I also understand that cropping in VirtualDub before capturing is not recommended.

[Alwyn]

I'm not seeing much difference between the two apart from the wonky S-Video frame. The face shots are a bit hard to compare.

I suggest you do a couple of test runs to assess for yourself.

But he came here to ask for our advice, not be told "learn for yourself".

Yet another really helpful comment. I gave them advice so they could learn and now they are assessing. Grow up.

[lordsmurf]

Sharpness setting is only at 3 in VirtualDub..

Sharpness best at 0 or less, never more. Depends on card.

aspect ratio issue (seems like the S-Video is too cramed/condensed although both files are 720x576)

720x576 is proper PAL (Full D1) capture resolution, aspect is PAR (not DAR 4:3).

I'm not seeing much difference b

It has lots of obvious overexposure. You need to calibrate your monitor.

[Sharc]

Sharpness setting is only at 3 in VirtualDub.

Set it to minimum. If sharpening is desired better do it in post processing.

Other than that, can the changes suggested by you be done before capturing in VirtualDub or do I need other software? I also understand that cropping in VirtualDub before capturing is not recommended.

For PAL capture always as 720x576 to allow for some left/right headroom for not precisely centered active pictures, as you did correctly. Crop in post processing only - if at all. The exact 4:3 picture is represented by the inner 704x576 pixels of the 720x576 capture.

The S-Video .avi in your link has clipped Blue (and Green). Maybe it's on the tape like this, then you can't do much. Indoor/artificial light sceenes can be difficult. You could try to reduce the contrast and saturation a little (Proc-Amp settings of the GV USB2, or in post processing).

[Bermuda1]

Sharpness setting is only at 3 in VirtualDub.

Set it to minimum. If sharpening is desired better do it in post processing.

I cannot set it below three, its the lowest I can select. Any way to fix this?

Other than that, can the changes suggested by you be done before capturing in VirtualDub or do I need other software? I also understand that cropping in VirtualDub before capturing is not recommended.

For PAL capture always as 720x576 to allow for some left/right headroom for not precisely centered active pictures, as you did correctly. Crop in post processing only - if at all. The exact 4:3 picture is represented by the inner 704x576 pixels of the 720x576 capture.

I don't really understand - I captured correctly, but when viewing on a PC monitor, it still seems condensed. Is it only on a TV where I can change i.e. to 4:3 that it will look correct?

The S-Video .avi in your link has clipped Blue (and Green). Maybe it's on the tape like this, then you can't do much. Indoor/artificial light sceenes can be difficult. You could try to reduce the contrast and saturation a little (Proc-Amp settings of the GV USB2, or in post processing).

I will try your suggestions! Thanks! However, any idea on how to fix the audio issue in example I uploaded?

Best regards,
Bermuda1

[The_Doman]

Ok, thanks for the hint. Read up on that so it’s fine. Any chance you can help me out on or have an idea how to fix the audio issue and aspect ratio issue (seems like the S-Video is too crammed/condensed although both files are 720x576).

I don't really understand - I captured correctly, but when viewing on a PC monitor, it still seems condensed. Is it only on a TV where I can change i.e. to 4:3 that it will look correct?

Your "condensed" captures are captured in Huffy AVI.
Huffy AVI video has no aspect ratio flags, thus is usually displayed as 1:1 in media players.
So when playing back your Huffy captures on the PC you have to set the correct aspect ratio manually in your mediaplayer.

Your DV-AVI captures DO support aspect ratio (and interlace) flags so most good players should display it correctly.
See this DV sample i uploaded some time ago here:

[Attachment 78850 - Click to enlarge]
Captured DV 720x576 format, should play correctly in 4:3
https://forum.videohelp.com/attachments/78851-1714910879/VHSDV_BRTN_Testcard.avi

The same file converted to Huffy AVI will play "condensed" in 1:1 aspect ratio.

[Sharc]

I don't really understand - I captured correctly, but when viewing on a PC monitor, it still seems condensed. Is it only on a TV where I can change i.e. to 4:3 that it will look correct?

Check you PC player settings. It should have a setting for 4:3 forced playback.

[Sharc]

See this DV sample i uploaded some time ago here:

[Attachment 78850 - Click to enlarge]

Your DV sample picture is horizontally squashed as well. The circle has 506/464 pixels which is not circular but has a PAR of 1.09 (mpeg4 4:3 PAR). It has to be stretched by the player accordingly. For undistorted view crop the small black sidebars and force the player to play it as 4:3.

[The_Doman]

See this DV sample i uploaded some time ago here:

[Attachment 78850 - Click to enlarge]

Your DV sample picture is horizontally squashed as well. The circle has 506/464 pixels which is not circular but has a PAR of 1.09 (mpeg4 4:3 PAR). It has to be stretched by the player accordingly. For undistorted view crop the small black sidebars and force the player to play it as 4:3.

Lol, we have had discussion already before eh.
The picture is not AR resized.
Just as example, it was captured from VHS AS IS through the D8 DCR-TRV120E camcorder.

[Sharc]

Lol, we have had discussion already before eh.
The picture is not AR resized.
Just as example, it was captured from VHS AS IS through the D8 DCR-TRV120E camcorder.

@Bermuda1's S-video huffyuv .avi capture which he has linked to has a PAR of 1.09, much the same as your DV capture
These discussions are sometimes like pissing into the wind, it seems

[Bermuda1]

@Sharc @The_Doman @lordsmurf: Thanks so far for the input! Changing the Aspect Ratio in VLC to 4:3 helped!! Now both the S-Video Avi and DV Avi look the same (in terms of aspect ratio).

I am currently capturing the same video again with altered settings in VirtualDub (i.e. Contrast and saturation) + made some other changes that I hope get rid of the audio issue. Will post an update soon and hope I can continue to rely on your great feedback !

[Bermuda1]

So the audio issues has disappeared - in VirtualDub I changed the chunk size to 4MB and under "capture timing options" selected "Do not resync between audio and video streams". That seems to have done the trick.

The S-Video .avi in your link has clipped Blue (and Green). Maybe it's on the tape like this, then you can't do much. Indoor/artificial light sceenes can be difficult. You could try to reduce the contrast and saturation a little (Proc-Amp settings of the GV USB2, or in post processing).

So I reduced contrast and saturation from 128 to 100. Not having any experience, I dont really notice any greens and blues being clipped. The whole clip now just looks less bright.

I also retook the screenshots from above for comparison. I feel like the benefit of the S-Video (better colors) is not visible anymore as this has now become significantly darker than the DV-capture.

Thanks!

[Sharc]

So I reduced contrast and saturation from 128 to 100. Not having any experience, I dont really notice any greens and blues being clipped. The whole clip now just looks less bright. Please let me know your feedback whether this goes into the right direction / whether Blue and Green are still clipped:
https://1drv.ms/v/s!AogvgivcX7rEgvEh1azDOuAiqc9gVw?e=yYVLcw

You reduced too much, maybe 120 would be a compromise for this tape/scene. Blue is still strong and clipping is still there, I guess it's on the source.
I prefer the S-video, alone by the somewhat lesser line wiggle and more details in the brights. Just don't reduce contrast and saturation so much.

[Alwyn]

Set your brightness and contrast correctly using the VDub histogram, as per my guide here.

[Bermuda1]

So I captured another version using 120 for contrast and saturation. Colors look better, however, if I understand clipping by now, I would find that its still there.

I also caputed the same screenshots again as before (S-Video vs. DV). Any more suggestions on the Levels I should consider?

Why so ever, my video and audio dont align anymore - I did not alter any other changes. Any clue as to why that can occur from one capture to the other?

@Alwyn: will try setting brightness according to your guide next.