VHS capture quality comparisons?
 

Okay, I'm testing out some of the new equipment I purchased, and it's making a believer out of me. This first test compared capture cards: The Tevion USB capture card I recently purchased NOS against the Hauppauge HVR-2250 which I've had and used for not quite four years now.

The setup is consistent between both tests: The source video is a souvenir of the USS Missouri (BB-63) around-the-world shakedown cruise in 1986. It was produced by the Public Affairs Office in the ship's onboard TV studio using Betamax equipment and then professionally duplicated on VHS for distribution. I've actually tried capturing this before; see below. The full video runs 71 minutes; this is just one music video clip.

The equipment setup is JVC SR-VS30>>S-Video>>El Cheapo Switchbox>>S-Video>>TBC-1000. I had the two capture cards hooked to two of the four outputs of the TBC-1000 so each is seeing exactly the same video signal. The Hauppauge was capturing uncompressed through VDub (no compression) on a Windows 10 computer with modern (2014) specs and CPU; it's the first video. The Tevion was on a used Dell Dimension running Windows XP; it was capturing using the HuffyUV codec and it's the second video. Also included are two screenshots of VDub's status following each capture.

Finally, for reference I am including the same clip as I first captured it almost 4 years ago using my Mitsubishi HS-HD2000U direct to the Hauppauge and capturing through their WinTV7 software in .ts format. No TBC, although the Mitsu's line TBC and DNR was active. It's actually easier on the eyes than the Hauppauge video above, although my guess is that's due to lack of sharpness making things smoother. You see color washing out on this one, though, in ways that you don't see on either of the clips made with the TBC.

No filtering of any type done on these clips; just edited for length and then rendered to MP4 in Corel VideoStudio 2018. I do welcome suggestions as to what I can do to improve the quality of the video after I capture the full 71 minutes; this tape is a rarity and it's personally important to me. Thanks for comments and help.

Adding a note...the screenshots were .png files. It appears that your uploader converted them to .jpg format. Is it supposed to do that?

Editing to Add: I had some trial and error with this setup...more error than trial. At first I had difficulty getting the Tevion setup to capture at all, and the Hauppauge was capturing but with a lot of dropped frames. After a few test runs I saw that I had audio playback enabled, and apparently that was more overhead than the old XP's horsepower could cope with. After I turned audio playback off for both units capture performance and dropped frames improved significantly. I still like to hear what I'm capturing, but the cheapie gaming switchbox has a headphone jack. A wire from there to my speakers works just fine.

For reference, that specific Tevion is an ATI 600 clone, same capture chips.

Why? "S-Video>>El Cheapo Switchbox>>S-Video>>"
That just adds a potential vector for noise and signal problems, so the switchbox and a cable should be removed.

The site software converts PNG to JPG, and resizes anything over about 900 pixels wide. The reason is because it makes the webpages huge, perhaps look awkward, and harms loading times. So if you a PNG, and need it to stay PNG, and a large <900px image, then that's what archives (zip,7z,rar) are for. Attach special images in compressed archives.

A problem here is that the files are not labeled. So no idea which file is from which card/setup.

Also unsure where you're going with this. Perhaps point out some things, or ask some questions. Right now, it's just "I took some captures" without any real conversation starter.

You can hear while capturing if you dual-route audio input, or adjust OS audio settings, or split then route audio. Start a new thread on that, and I can explain more. You hear audio separate from whatever VirtualDub is doing (ie mute).

Are you capturing to mp4? Lossy encoding bitrates, GOP size, and other variances simply invalidate your test. Are we comparing your mp4 encoder? You should be capturing to unaltered lossless YUY2. Any popular lossless codec will suffice to reduce file size without altering content (huffyuv, Lagarith, UTVideo codec, etc).
:wink2:

The captures were made lossless, to uncompressed with the Hauppauge and to HuffyUV with the Tevion. They were edited (for length) and then rendered to MP4 with the same encoder/program (VideoStudio) using the same settings. I can post brief clips of the original AVIs if you let me know the time segment which would be best for evaluating.

The original AVIs may not be needed, just clogs up server space. If the H.264 has a really high bitrate, that may suffice for whatever it is you're trying to show us.

These mp4s were rendered at 2.5 Mbps:

However, VideoStudio offers me a choice going up to 65 Mbps for 4K. Which would be most effective to demonstrate the performance and quality differences between the capture cards?

-- merged --

So, I'm still trying to get familiar with the new equipment. My next test tape is a cheapie cartoon collection which I haven't looked at in 20 years, "100 of the Greatest Cartoons (none of them would make MY top 100!)", recorded in EP mode and sold as a 2-pack of 6 hour tapes.

I tried making this capture on the Panasonic AG-1980, which is supposed to have a good reputation with EP tapes. However, I had a nasty issue which seems to be tracking...vertical lines across the screen. You'll see them degenerating into pure hash at times; that's me as I'm attempting to run the manual tracking through its cycle in order to find a sweet spot. I never succeeded. Workflow is AG-1980>>S-Video>>TBC-1000>>Tevion/WinXP capture card...Lordsmurf, I'm only using the cheapie switchbox when trying different combinations of equipment; for the real capture I plug the VCR direct into the TBC and just use the switchbox to hold the extra cables.

After giving up on this capture, I decided to try my old Mitsubishi HS-HD2000U...and 20 minutes into the tape it's working perfectly with the auto tracking; no picture issues at all. It looks so good, in fact, that I've elected to go on and capture the full 6 hour tape (Which means that I won't have the comparison video available until late tonight).

Is there some setting I could try to get better results out of the Panasonic? Or is this just a case of hoping you have enough wrenches in your toolbox to find the one which fits just right?

Well, I was having dinner with my parents and the capture ran an hour too long. I'm using VDub to trim off the excess...and it's taking longer to trim the capture than it did to capture it!

Quite a bit longer, actually...it looks to be processing at about 18 fps. That's slow enough that I can see some fine detail on the images which escaped me during capture. Turns out the Mitsubishi wasn't "perfect"...at slow speed, I can see the ghost of the tracking lines on the capture image. But it's a night and day difference from the Panasonic, as I should be able to show you tomorrow (late tomorrow...we're going to a friend's church 100 miles away early early in the morning).

I do know that the Mitsubishi's DNR was functioning, along with its internal TBC. The Panasonic's TBC was also turned on during its capture, but I don't know how to check whether its noise reduction was functioning or if the Mitsu just coped better with this particular tape. If there's some setting that I can check or change, though, I'd like to know about it.

-- merged --

And here's the capture of the same video from the Mitsubishi; all the rest of the workflow path unchanged.

Obviously, I'm satisfied with this result; it's just that for future reference I'd like to make sure that I hadn't overlooked something when setting or operating the Panasonic. So, thanks for any help.

Hi here is some pictures from Panasonic FS200 and Panasonic NV-HS930 and JVC HR-S8500

i did use Ati All In Wonder 9000 and Datavideo TBC-5000

Panasonic FS200 = Internal TBC Off & Edit On (Can´t remember if i did use ES10 but i do not think so)
Panasonic NV-HS930 = Internal TBC On & 3D DNR On Sharpen
JVC HR-S8500 = Internal TBC On & Edit On & all other setting Off

Hi here is some more pictures from Panasonic FS200 and Panasonic NV-HS930 and JVC HR-S8500

i did use Ati All In Wonder 9000 and Datavideo TBC-5000

Panasonic FS200 = Internal TBC Off & Edit On
Panasonic NV-HS930 = Internal TBC On & 3D DNR On Sharpen (i did add a second picture with 3D DNR Auto)
JVC HR-S8500 = Internal TBC On & Edit On & all other setting Off

If an ES10 is in the chain you'll want to turn off the internal TBC on your VCR. Its passthrough ability works better when it gets an inferior signal. I would think you would want to turn off any editing functions or picture settings on your VCR. They usually just add noise. However, the difference between noise and detail is highly subjective, but each tape and deck are unique.

Hi i did not use ES10 on does captures at all

-- merged --

Hi i did get a ATI TV Wonder HD 600 USB today here is some pictures

i did use ATI TV Wonder HD 600 USB and Datavideo TBC-5000

Panasonic NV-HS930 = Internal TBC On & 3D DNR On Sharpen

When I compare the captures from the ATI9000 & the 600 I see you have used the wrong capture resolution for the ATI9000.
The native resolution should be 704x576 as you can see the captures from the 600 has black borders on the right,left side.

My favourite would be the FS200 captures.

How well does the TBC-5000 work?

Bogilein yes i did notice the black borders on the right,left side on the tv wonder 600 usb

hmm i did use 720x576 on both ATI9000 & the tv wonder 600 usb

hmm i did try to find any setting to get full frame on tv wonder 600 but there is no setting for it maby it´s the drivers

i do get little black borders left and right with ATI9000 to on some other vhs movies like "In the line of fire" & "Magnum Force" but no black borders on the VHS movie "Back to the future II" & "Titanic"

i did notice that ATI9000 have much finer steps on the contrast slider so you can fine tune contrast much better on the ATI9000 then on tv wonder 600 usb maby that has to do with the 12bit vs 9bit video encoder

so on the virtual dub capture histogram i did get the contrast level just below the clipping point with ATI9000 but with tv wonder 600 usb it was either to high or little to low but it´s better to be on the little to low side

maby a external proc amp like Elite Video BVP4 or SignVideo PA-100 can help to get it perfect

-- merged --

hodgey yes it works very good does not blur the image at all
i did buy it directly from Datavideo maby 7 years ago

i did buy it because of blackmagic intensity shuttle USB 3.0

hehe my old Matrox marvel G400 TV + macrovision hack does not need any full frame TBC it´s better to not use Datavideo TBC-5000 with Matrox Marvel G400 TV but it needs windows 2000 SP2 or lower

For Matrox Marvel G400 TV To install Video tools w2kvt204.015 you must have windows 2000 SP2 or lower that have Windows Installer version 1 if you do not have Windows Installer 1 then you get this "Installshield Wizard Interrupted" message

i did not get the capture histogram to work in VirtualDub with Matrox Marvel G400 TV so i have to use my own eyes to fine tune the brightness and contrast sliders

Made another capture with the ATI9000 with the resolution 704x576 and compare it to the other captures. If you capture with 720x576 your capture will be streched.

Bogilein i have to see maby i do it :)

and i like FS200 captures best to

ATI TV Wonder 600 USB can not do 704x576 only 720x576

but ATI All in wonder 9000 i think it can do 704x576

While it wasn't designed for consumer formats like VHS, in conjunction with an ES10/15 as passthrough it's a pretty solid set up.

cbehr91 ok :)

Here is some capture with 704x576 on Ati9000 and i did a Test on a Test Pattern for 1.33:1 that i did find and i did capture it in 704x576 and 720x576 with ati all in wonder 9000 and it look that it´s best to capture in 720x576 with ati9000

i did use Ati9000 and Datavideo TBC-5000

Panasonic NV-HS930 = Internal TBC On & 3D DNR On Sharpen

one more thing i did a capture with Matrox Marvel G400 tv but the capture is more interlaced than the capture with ati 9000

on the ati9000 i see only interlace lines on the sweden subtitle but not on the titanic movie vhs

but on the Matrox marvel capture i see everywhere interlace lines

but on some other vhs capture it was the opposite the ati9000 did have interlaced lines but not the matrox
but mediainfo do say they are interlaced

here is Matrox Marvel G400 tv capture pictures

hmm here i test my Blackmagic intensity shuttle usb 3.0 and my DVDO iScanDuo
i did use VirtualDub to capture not Blackmagic Media Express

when i did use DVDO iScanDuo i did then use HDMI in 576P on Blackmagic intensity shuttle usb 3.0
when i did not use DVDO iScanDuo i did use S-Video in 576i on Blackmagic intensity shuttle usb 3.0

Test1
HS930>TBC5000>DVDO iScanDuo>Blackmagic intensity shuttle

Test2
HS930>TBC5000>Blackmagic intensity shuttle

The DVDO iScanDuo appears to add noise and reduce image resolution with that setup.

traal yes you are right it only get worse

-- merged --

i did get a ATI Tv Wonder USB 2.0 it works very good on WinXP SP2

i did use this driver cd 180-V01100-200 it have driver version 6.14.10.6317 and MMC 9.08

but the driver version 6.14.10.6317 capture did look more interlace like my matrox g400 marvel capture

but i did find driver version 6.14.10.6314 and now i do not see any interlace lines on the Titanic vhs capture like with my ATI all in wonder 9000 and ATI TV Wonder 600 USB

and i did some testing S-video VS Composite and it looks like i maybe get better quality with Composite

so here is some pictures of driver 6.14.10.6317 VS 6.14.10.6314 and S-video VS Composite

The S-Video screencaps show some weird banding. That's not normal.

traal ok have to do more testing and see if i can get it better maybe bad s-video cables

Looks like the fields 6.14.10.6317 captures are out of phase compared to the other setups, no idea what could cause that.

The TV wonder seems to crop more of the the left and less of the right.

The banding looks a bit like hannover bars, maybe the cards handle the chroma decoding a bit differently, it could be something else though. Won't be as noticeable on composite as the comb filter may cancel out the effect a bit more.

hodgey ok but 6.14.10.6317 works good on other vhs tape that have interlace lines like with 6.14.10.6314 and my ati all in wonder 9000 and ATI TV Wonder 600 USB and i do have ati all in wonder 9200 256mb but it´s the same as my 9000 64mb

on does vhs tapes 6.14.10.6317 looks progressive no interlace lines

but i think 6.14.10.6314 is the right way maybe to capture because it works then the same as my other capture card exept matrox g400 marvel

maybe it does something to do with Pal 25fps and Film 24 fps


i did find 6.14.10.6321 but it was the same as 6.14.10.6317

Regarding the "interlacing" (really phase-shifted progressive 25 fps). The 6317 has an extra pure-black line at the top, which reverses the field order.

msgohan oh ok hmm i did try to crop 1 line at the top did not get any better i did try field swap filter on virtual dub but did not get any better

hmm i upload 2 very small clips of titanic pal vhs im not sure how ok it is to do it but admin can delete it if necessary

the clips is with Composite HS930>TBC5000>ATI TV WONDER USB 2.0

6.14.10.6314 vs 6.14.10.6317 drivers

Thanks for the lossless video samples. They take a lot of guesswork out of it, compared to screenshots.

It would be better to test with a full-frame tape instead of letterboxed, as then you would immediately see the black lines being added to certain captures. In fact, there are 3 lines "added" to the top of the 6317. (Any odd number of lines would reverse the field order, and only 1 of these lines is pure black.)

It's a lot easier to work with fields in Avisynth than VirtualDub. Field Swap is definitely the wrong filter to use.

I decided that it was easier to show what I'm talking about by making a little screen capture video rather than trying to type up a long explanation with a bunch of attached screenshots. No audio though. In the video, I demonstrate using Avisynth viewed via AvsPmod to match the 6317 capture to the 6314 one.

https://youtu.be/BQZKRtI3x2M (view at 1080p)

There is a diagonal interference pattern in these composite captures. It isn't visible when looked at in weaved "frame" mode, because it changes between fields, and looking at the 50Hz video as 25fps frames is effectively a 50% temporal average. I demonstrate it from about the middle of the video until the end, but YouTube's compression smoothed it over a lot so it isn't as noticeable.

These are the Avisynth scripts I used. Lines that begin with # are "comments" that are ignored when the script is read. An Avisynth filter line can be "commented-out" so that it is ignored. I toggle these commented-out lines with AvsPmod's ALT+Q command in the video. They begin with #~ specifically.

msgohan Thanks :D

i have to look this youtube clip many times to fully understund it i do know little about Avisynth and AvsPmod
done some Years ago little super 8MM frame by frame and HDR capture here is
my 8mm capture example https://www.youtube.com/watch?v=nCcKXjlwgRw

the youtube clip you did was very good i think i start to understund it :)

-- merged --

Now my Matrox Marvel G400 TV capture i good to no interlace lines so matrox capture card did have the same problem but now it´s fixed :D

Thanks again msgohan :)