I've been lately playing around with VHS capture, nothing in professional terms but just an amateur distraction. As I recently learned VHS does degrade over time, I wanted to keep some tapes from my childhood like old Terrytoons cartoons and my beloved Star Wars trilogy.
My equipment is pretty basic, but it works fine for my intentions, as I don't really want to restore or improve quality, just keep it as near to the source as it can:
- Sony DCR-TRV33E paired with a PCI FireWire card as capturing device
- Daewoo ST847S VCR (6 heads)
I've succesfully transfered a couple of tapes. I first used sclive but now I'm using Premiere to remove (just a bit of) noise before encoding in h264 to archive.
But I'm having some trouble with the Return of the Jedi VHS: While it seems to capture properly, I've noticed that, in some specific parts of the movie, brightness goes up and down, concretely after scene changes.
At first, I thought it was an anty-copy problem, but using a Macrovision box I have didn't change anything. Then I tried with another VCR, an LG C20P, with no change at all except for lower image quality overall. Not to say that the problem does not occur when displaying the VCR through a TV (brightness keeps stable, with any of the VCRs).
The problem is already present in the Handycam (I can see the brightness fluctuating in its LCD screen), so I guess it has nothing to do with the FireWire cable or card. As others have suggested, it seems to be some kind of brightness 'auto-adjust' that is causing the ups and downs, but it's confusing that it only happens in some movies.
I've read threads of users with similar problems (like this and this) to no conclusion. Even read that people using superior devices such as SVHS with TBC can't get rid of the problem.
I can upload a portion of the capture if that helps. Thank you in advance!
+ Reply to Thread
Results 1 to 30 of 30
Last edited by trotskito; 18th Nov 2020 at 14:10.
As I'm using a DV Handycam, I thought DV capture was the thing to do... On the other hand, while Premiere bests in editing, it has a capturing option.
Anyway, I see the problem in the camera's LCD, so I guess using another program to capture won't change a lot.
You may have some image processing happening inside the VCR or the camcorder, Try to disable every image option you have set to ON on both the VCR and camcorder, I can see clearly some automatic gain control going on when there is a scene change. Alternatively you can hookup the VCR to the TV and see if it's there so you can at least eliminate one component.
Yeah, I already tried to play the movie on TV and there are no brightness changes.
I'm almost convinced the problem is on the camcorder, but I can't see any options and those that I've disabled have not had any effect.
One user at digitalfaq suggested it could be indeed a Macrovision problem. It's weird because the tape is from 1995 and I'm almost sure I made some VHS copy of it back in the 2000s without this problem.
Could it be that I'm using the Macrovision box not with its original scart cable, but with different RCA ones instead (in order to use the camcorder's RCA-to-minijack cable)?. I attach pictures of both.
I have the original cable, but as it ends in another scart, I cannot use it in the camcorder, unless there's some kind of female scart-to-rca adapter. But, would that make any sense?. Also, maybe the box is broken.
Last edited by trotskito; 19th Nov 2020 at 04:23.
This could be a cable problem...
Try with another cable set.
I've just tried with:
- Another RCA for the Scart adapter -> Box connection,
- An S-Video-to-RCA cable for the same purpose and
- Another direct Scart-to-RCA cable...
...but no luck. The brightness keeps crazy on the affected scenes. I forgot to say I've tried with two different RCA-to-minijack camcorder cables, resulting in the same problem.
I'm starting to seriously suspect the Macrovision box has broken with the years (or is not useful to avoid this sort of errors), though it was well preserved and never used in like 15 years. I'll try to confirm it tomorrow by trying to make a VCR-to-VCR copy.
Last edited by trotskito; 19th Nov 2020 at 06:31.
It's an automatic gain (brightness/contrast) problem. probably in the DV camcorder. Here you can see the background brightness darken every time the bright white lights in the background appear, lighten when the lights disappear. Video trimmed, slowed to 12.5 fps, waveform monitor added:
Sadly, I can't see any option in the camcorder to prevent this automatic gain setting. I guess I'll have to look for another capture device if I want to do this properly. Apart from those cheap RCA-to-USB sticks, is there any consensus on a cheap, but not so bad, option?.
What intrigues me is that the problem is not (at least, not visibly) present in every tape, just in some of them, like ROTJ.
Last edited by trotskito; 19th Nov 2020 at 08:55.
Yes, I found that option yesterday but only works in camera (record) mode. I haven't found a similar setting for VCR mode. It seems it is "hard-coded" in the camcorder behavior .
Update: The camera has an S-Video input/output, so I tried it both with an S-Video to RCA cable and a regular S-Video cable (through the scart adapter, which has also an S-Video output). Apart from getting lower quality (and B/W in the second case), the problem is still there.
Also, I think the Macrovision box is working: I tried to record the VHS signal directly on the camcorder, but unless I plug the box between it and the VCR, it displays a "Copy Inhibit" message on the LCD screen. So, it seems it's doing its duty there, even if that doesn't help in this case.
I just sent an email to Sony support. Curious what they will answer about a 2003 camcorder. By the way, I checked the previous VHS captures (the ones I thought were fine) and the problem is present in almost everyone, but is less noticeable.
No, Sony will not answer your question, Get a USB or PCI capture device and move on, DV is not a capture standard anyway, It was a compressed recording format for camcorders and served its purpose 20 years ago, it never meant to be a video file storage format for archival purposes.
I guess I'll have no more options than that. I see plenty of capture USB sticks, ranging in price from €10 to €100. Can't find any PCI card in stock. Could I ask for a quick recommendation?
Your ROTJ example is an extreme test case for such an AGC design error. You've got two heads framed in tight closeup, with a bright white background, one head is Caucasian with khaki colored garb while the other is almost a featureless gloss black helmet in its entire area. Each head/face just slightly dominates over the other in frame coverage as the editing alternates over-the-shoulder POV shots. So all the AGC "sees" in this instance is 60% of the frame alternating between stark black and white tonal emphasis (compounded by residual MV interference most filter boxes don't clear 100%). When the black helmet takes up more of the frame, the AGC scrambles to moderate it, then when the white face/white wall dominates, it has to catch up again. The prolonged visible adjustment may be traceable to a lag that gets triggered whenever the lighting balance hits an exact gap in the logic programming, then changes too rapidly for the circuit to keep up while concealing its actions.
Note shifting over from DV capture to the "standard" workflow most often discussed here will not be fun. For all its faults, DV capture is about as easy as it gets short of using a VHS>DVD combo recorder. Once you start in with direct VCR>PC capture via USB dongles or cards, all bets are off. More steps are involved, way more HDD space, more complex software adventures, and often the need for a several hundred € outboard TBC box (your old MV filter box might not suffice). You may get lucky, and obtain satisfactory results with just the dongle and simple software, but be prepared for a learning curve and further expenses if that proves ineffective or disappointing.
Last edited by orsetto; 19th Nov 2020 at 16:24.
I always say capture without an external TBC first, if there are timing or audio sync problems then try a DVD recorder in the workflow assuming a good capture card from the 2000's is used, 99% of modern capture devices will not work on their own and need some sort of timing correction unfortunately, from the $5 easycraps all the way to the $300 BM intensity shuttle.
On the field of capture cards, ages ago I had an AverMedia for TV reception, but now I'm lost. I assume the more expensive means the better, so... AverMedia again? Elgato seems to be have a lot of fans but is out of my budget. And I've read tons of positive comments on the €10-€20 ones, but then I assume they come from people whose standards are lower than mine.
Anyways, thanks a lot for the explanation!
I have the possibility of getting a Panasonic DMR E-50 DVD recorder. I've read very good opinions about its built-in TBC. My question is: Will that make any difference when capturing with the camcorder?. If not, is it a good option as an intermediate between the VCR and, say, a Diamond 600 USB?.
PS: I really have some allergy to USB plug-and-play thingies. I'd rather prefer a PCI/PCIe card but it seems they're no longer produced.
PCI cards can be had used and some still work well with Win 10 using Win 7 drivers. It doesn't have to be made in 2020, VHS was dead decades ago so there is no advancement made in the capturing technology ever since. As a matter fact a steep decline happened in making processing chips with very low video quality.
If not, is it a good option as an intermediate between the VCR and, say, a Diamond 600 USB?.
Making it more confusing, some tapes and capture setups can occasionally require BOTH a true TBC and a dvd recorder like ES10 used in tandem. A couple of Panasonic models can repair geometric distortions like bending verticals or washboarding seen in some really bad tapes, which a true TBC box cannot fix. Put another way: generally a true TBC repairs issues you can't see but "invisibly" irritate PC capture circuits and ruin the transfer, while certain DVD recorders and some premium SVHS vcr models with "TBC/DNR" circuits mostly fix distortions you can easily see when you play the tape. Still other DVD recorders like the E50 fit between these capabilities: their internal "TBC" doesn't do anything for visible distortion, but has a frame synchronizing ability somewhat similar to a true TBC that can be just capable enough to patch over some issues like lost lipsync that can occur with unaided VCR>PC connection.
Like I said, things can get real confusing real fast once you leave the comfort zone of capture via camcorder DV conversion. If you've been satisfied with your DV results aside from this AGC glitch, perhaps continue to use that method for ordinary tapes and reserve more complex methods for problematic examples like ROTJ.
If you could somehow rent or borrow a full-on, VHS-optimized TBC (like the DataVideo TBC-1000) and connect it between your VCR and Sony camcorder, testing it with this problem ROTJ scene would definitively tell you whether the issue was the camcorder AGC or more localized to out-of-control MV protection on the tape. Most MV filter boxes merely masked the MV: they didn't completely eliminate it. MV works specifically by tricking the AGC circuits in consumer gear, and some video hardware brands like Sony were notorious for being hyper-sensitive to the slightest whiff of MV. If your filter is leaking enough MV remnant to fool the Sony AGC, that could be your narrowly specific problem. True TBCs completely annihilate MV by rebuilding the baseline VHS signal. OTOH, if the problem remains even with a true TBC, its almost certainly localized in the Sony AGC circuit lagging during high-contrast rapid edits. Trying different camcorders is unlikely to change anything: most DV-converting models were Sony, all probably containing the same AGC design.
Most likely, you're simply stuck with an AGC issue in the Sony thats sluggish to react when certain extreme contrast ratios occur during certain scene edits. As long as the Sony is part of your capture chain, you'll occasionally encounter the issue.
Last edited by orsetto; 21st Nov 2020 at 13:22.
Years and years ago(more than 3 decades ago) I used to use an old Betamax('81 Sony SL-5800) as a passthru device between my VHS VCRs, for CP'd tapes. As it was the Beta format of the early 80s ignored MV not only for recording to the Betamax but even if the Betamax was put inline between the two VHS machines it would allow you to copy the original without the bright and dark flashes associated with VHS MV. By the late 80s I purchased a cheap MV remover and stopped using the Sony for this but I wonder how something like my original Betamax would compare to say an ES-10? note the Betamax lacks S-video which I always use for DVDs so if you use S-VHS then maybe an ES-10 would be better anyway.
Note I've heard the ES-15 also works for this?? and I have several ES-15's but it's a year older than the ES-10 and notably a lower build quality so I wonder just how good the ES-15 works for this?? I believe the ES-10 had a MSRP of $230 while the ES-15 dropped to $170 and I purchased several new for $99 towards the end of their run, with a $60 lower MSRP on something as cheap as a DVDR they really have to start cutting back on quality to achieve a drop like that in one year......the E50 for comparison is 2 years older than the ES10 and had a MSRP of $600! really build like a tank but I wonder how good the video quality was? sometimes first models while high build quality didn't have as good video quality......I personally wouldn't go older than the ES-10 but that's just me.
Last edited by jjeff; 21st Nov 2020 at 16:25.
For the "pass thru" functionality we're discussing here, the build quality of any individual Panasonic DVR model is not as important as its video circuit topology. Over the years the topic has been hashed here and at other related forums, the Panasonics have been whittled down to one "All-Star" (ES10) and two "Runner-Ups" (ES15, ES20). These have consistently proved to have some definite helpful effect when a VHS signal is passed thru them into a separate capture system. The ES10 is one of a kind, more dramatic and more brute force, able to repair some truly terrible defects (at the cost of some residual artifacts of its own, just as you sometimes see with dedicated TBCs). The ES15 and ES20 dial back some of that pass thru ability to a more mild, less intrusive but also less effective form. They can suffice as frame synchronizers, and have some degree of visible defect correction, but aren't nearly as powerful as the E10 (nothing is, really).
The E50 and later Panasonics would fall somewhere in the vicinity of the ES20 as passthru units: little to no visible defect repair, but depending on your capture device, these might provide basic frame synchronizer functionality that may or may not be able to sub for a "true" TBC. The only recorders that have been extensively tested and vetted for passthru versatility are the ES10, ES15, and ES20: any other recorder model from Panasonic (or any other brand) is a question mark. If you search hard enough, you'll find clusters of favorable anecdotal reports for many late-period (2005-2009) recorder models re frame synchronizing ability, but it varies widely depending on what is connected to what and the specific tapes involved. At best, these random units will minimize the common, irritating lip-sync audio drift problem that plagues many simple VCR>PC setups, but thats about it.
Some tapes and/or capture devices won't fully cooperate with anything less than a true TBC from DataVideo, AVT or CBT. If the tape shows significant visible distortion like flagging, you may also need to add a Panasonic ES10 to the TBC. Then theres the whole sub-topic of optimized VCR models...
Last edited by orsetto; 22nd Nov 2020 at 18:35.
No one has mentioned the fact that this is telecined film; pulldown has been added. If you walk through the scene transitions frame-by-frame, some of them don't have any brightness surge, but all of them have ghosting from the field duplication. Doing IVTC will help quite a bit at the transition points. You should ALWAYS do IVTC on any telecined film material. If you don't, you'll end up with much worse encoding artifacts. The difference is not subtle.
It sounds like your other tapes transferred OK, and only this one has a problem. Were the other tapes movies, or were they 29.97 interlaced video?
Firewire has nothing to do with the problem.
I've captured a LOT of VHS using various DV decks and camcorders and have not seen this. The closest camcorder I have to yours is my Sony TRV-11. I suspect that the problem is that your camcorder is slightly out of adjustment. I have the service manual for my TRV-11 and it does show AGC adjustments and various other service settings that might cause the problem you are seeing. Unfortunately, changing these requires equipment that pretty much no one has.
My recommendation would be to find/borrow/buy another camcorder or, if you want, use a capture card in order to avoid DV (although for VHS I think DV is just fine -- VHS is already pretty awful).
I just shut down the editing program I was using to walk through your clip and noted that some of the scenes have a 100% bright object in the background. I also note that there are not brightness problems at scene changes when both scenes have 100% bright objects.
Are you seeing this brief exposure adjustment in all scene changes across the entire movie, or only those with bright objects? If you look at various posts over in doom9.org about AVISynth plugins designed to auto-correct exposure, one trick to make the algorithms work better is to provide pure bright and pure black in each scene by temporarily adding a border. You can't do that here, but I bring it up only to point out that if you don't have the problem on other scenes in the same movie, this may explain what is going on here.
Last edited by johnmeyer; 22nd Nov 2020 at 20:58.
"VHS is pretty awful" is a pretty good reason not to capture it into DV, Two wrongs don't make it right.
The color limitations don't matter much when capturing NTSC VHS because the colors are already limited. Therefore, your logic that bad color will be made worse is flawed logic: if the colors are already within DV's limited color space, they can actually be captured "perfectly" (in quotes because no capture is perfect).
As for the DCT artifacts, once again those aren't going to show up as much because of the very limited detail that VHS can record. Having said that, I won't try to defend that any further because the artifacts are definitely more pronounced than more modern codecs, but I'm guessing that 99% of the people doing these transfers will never notice.
One thing I learned early on was to always look at the results on a big screen, with good equipment, but do so in real time. I emphasize that because if I look at the video, frame-by-frame and magnified in my NLE, I can find and see every single flaw in each frame. I do this all the time when trying to optimize a particular workflow, but if I can't see the difference when playing the video, I don't obsess about it any further. Don't get me wrong, I don't intentionally make something worse, nor do I take shortcuts that ruin the result, but if I can do something in 1/10 the time and not see the difference on the screen, then I'm wasting my time and the client's money to try to use a more complex workflow.
The reason I still recommend DV is that is completely avoids the problems of dropped frames, bad proc amp settings, and a dozen other mistakes and errors that relatively inexperienced users always make when trying to use more advanced, "better" capture systems. I could write a book about all the problems I have had with various more sophisticated capture systems. And, all you have to do is read these forums for the past twenty years to see thousands of posts from newbies and intermediate users who have completely screwed up their videos by trying to use equipment that is beyond their pay grade.
It's like it was when I was growing up and most people used Instamatics to take photos. I would never have tried to encourage them to go out and buy a Leica M3 (I still have my 1958 M3) because they would probably get nothing but out of focus, badly exposed pictures because of all the settings they'd have to master before it would work for them.
One final note: not all DV is created equal. The worst example was the DV codec supplied with Windows, back in the day, which was absolutely awful. By contrast, the MainConcept DV codec (included in Vegas Pro) was rather spectacular. Canoupus also produced a top notch encoder. I own them both.
I have several VHS tapes that I had captured previously in DV with the Edirol VMC-1 which a consumer camcorder can't even come close to its capabilities and the tapes are not actually NTSC, they are PAL and I will post some samples when I find them between DV 4:2:0 and H.264 4:2:2 and you will be the judge.
Originally Posted by johnmeyer
I just got the E50 for less than €20, and in the next days I'll have a Canon MiniDV camcorder borrowed (which I don't know whether or not has A/V input), but I won't be able to test them until next weekend.
I'll be posting results. Thanks a lot everyone for your contributions.
Last edited by trotskito; 23rd Nov 2020 at 07:55.
Orsetto and anyone else interested,
The '06 Panasonic ES-25 is basically an '06 ES-15 with HDMI output and SD card input, just an FYI if someone runs across one. The '05 ES-20 is kind of an odd duck, it uses LSI silicon(the first Panasonic along with the VHS/DVDR combo ES-40v) to do so. I have several ES-20s but prefer to not use them due to their quirks and limitations like not being able to pause a recording using FR, a speed I use all the time but basically none of these limitations would hinder using it for a pass-thru device and maybe the LSI silicon might help?? I know some members prefer it although maybe that would only come into play if you actually recorded to the build-in DVD burner.
The ES-25 is kind of hard to fine as the vast majority of those models sold were the cheaper ES-15 but I have one I record to almost daily lately and it still keeps chugging along, I have had several ES-15s fail with laser issues but they were used a ton back in the day.