Ending up having a hell of a summer and wasn't able to work on any of this video stuff... but I've started the process up again tonight and did my first capture with the Hauppauge Live2 USB. Recording went smoothly with zero dropped frames. I did a recording using the JVC DR M100 DVD Recorder to compare against, both with pass through from the Panasonic DMR ES-10.

Just finished the Live2 USB recording and the main issue I've noticed so far is there is quite a lot of noise in the audio track. Typical VHS noise, but this tape is in good shape, my VCR heads are clean and I have not had this noise issue using any of my DVD recorders for the same tape.. that's 3 different DVD recorders and none have picked up this noise which is constant and distracting on the Live2 USB recording.

Will have to do some recordings of different tapes to see if its always present or what.. but I'll get some clips made of the different recordings to post for the sake of reference.

-- merged --

Ok, here are a couple comparison clips of transfers:

VCR: JVC HR-S3500U SVHS using S-Video connection

One transfer is using the JVC DR-M100SU DVD Recorder
Other transfer is using the Hauppauge USB-Live 2

Both of these recordings had the video signal passed through the Panasonic DMR-ES10 DVD recorder to apply its TBC function to the video

The only issue I had with the Hauppauge is the presence of audio noise in the right channel that sounds similar to audio noise during tracking, and its present through the whole recording and quite loud. It seems to be just this VHS tape but I have not gotten this audio noise problem when I record the tape off to 3 different DVD recorders... strange.

Since the lossless avi clips are quite large, I have encoded all clips to x264 to reduce size and also correctly display AR/cropped out the black. I haven't filtered the picture or anything otherwise and used sufficiently high bitrate to where quality shouldn't be affected. I don't know if anyone really cares to see these comparisons at this point or not, if anyone wants uncompressed clips for comparison I can provide them

How did you do this? Transfer via a DVD?

Yeah, its a DVD Recorder. From what I've read its supposed to be one of the best, or maybe the best DVD recorder for transferring VHS tapes because of the chip. That one is just recorded to DVDR, passed through the Panasonic as well.

-- merged --

I should also mention the clips are from a 1999 movie "The Apartment Complex" and the VHS release is a very cheap, shoddy release with a lot of problems in the video. So if anyone's looking at these clips for an idea on the recording quality of the equipment, that should be factored in.

I will be transferring some other tapes, if there's any interest I can post more example clips from different sources

The videos have been compressed, so I do not know if the comparison is fully appropriate. In addition the USB-Live 2 capture has been downsampled to YV12.

I can comment that Hauppauge capture is nice for details, colors and overall look; the levels are well distributed (and do not present the classic "spiky" histograms because the post-processing operation).
Attachment 15698

The levels of JVC capture are shifted-up compared to USB-Live 2, and extends in Y higher levels, if you need a 16-235 range for whatever reason you need to shrink them.
Attachment 15699

I did not equalize the levels to USB-Live 2 for the comparison, here is some sample:

https://imgsli.com/MTI5OTA3 use fullscreen and zoom

https://imgsli.com/MTI5OTQ3 use fullscreen and zoom

The is some ugly image distorsion in JVC capture in the lower part of the gate in the background, easibly seen when sliding the images in the above comparison; it should not be related to its original MPEG-2 compression, and I ignore the reason for it.
Attachment 15700

There are inserted and dropped frames in JVC capture in the second clip (I did not check intensively the first).

About your audio noise problem with a specific tape, I do not know what to say, I never had a significant problem with USB-Live 2 capturing the audio. If you experience that the audio recorded by the JVC is better, than you can try to mix it to the USB-Live 2 video, but be careful of the dropped/inserted frames in the first.

An AviSynth denoise/sharpening post-processing on the USB-Live 2 capture can improve a bit.

And finally, if if wish to redo the comparisons, do it on original captures, and eventually equalize the levels to have a better reference. Also an

I don't agree. It's just a decent capture, but has flaws. You won't really get a flawless capture, and have to choose what flaws are acceptable. The Live2 loses details in shadows, dark areas. The color are not rich, and may actually undersaturate if contrast is lowered. Not a bad capture, but certainly not perfect.

Yep, the IRE for most recorders is off a bit. The JVC is about +5 hot. But of all the recorders, only referring to tape transfer quality, it is the best ever made. (For off-air, RCA with Zoran chip is best. When chroma noise is fully removed, image not bad in the slightest, the Zoran RCA also great for tapes. But that's mostly for broadcast masters, not homemade anything.)

I doubt it's in the MPEG, just the compressed clip.

Note that it could have happened anywhere, including the software conversion from MPEG. Not all encoders read MPEG properly. So full info on workflow, hardware and software, is required. Cannot assume anything here.

It can happen on Hauppauge cards. (As I always state, Live2 is a long lived model, with production changes both in-card and on-chip. Conexant hasn't existed for many years now, and yet the Live2 is still sold. It's not all NOS. So some buyers of that card, seeking quality, will be SOL.) Conexant was never known for quality audio processing, a reason Realtek overtook it for embedded motherboard audio.

I don't agree. The final quality depends on the tape, not on the USB-Live 2, which is an excellent devices. And for this tape the winner between the Hauppauge and the best ever dvd recorder JVC is clear.

Let's see the uncompressed capture to understand.

Let's see the uncompressed capture to understand.

Once more, there are no evidence of production changes both in-card and on-chip. And BTW, the OP problem only relates to 1 single tape, which can be marginal.

I will post up some different, very short clips that are uncompressed - they will have to be extremely short because the clips I posted here all exceed the size limit for the uncompressed lossless avi capture from the USB Live2. The uncompressed mpeg isn't a problem for size, but the lossless avi is quite large.

Yes, the audio problem so far is limited to just this tape - at least based on my *limited* testing. I have only popped in a couple other tapes to check and they so far have not had this audio issue. But I can't get rid of it for this particular tape. I will sync the DVDR recording audio instead, which isn't a problem for me as I do audio syncs regularly for years... I just don't want to have to do this every time I make a recording. I've also tried moving the VCR and changing power outlets in case its some kind of interference, no luck. The audio from the USB Live2 is actually quite good aside from this noise - it is much more clear and crisp than the DVDR. I haven't tried to rerecord the tape to DVDR again, so perhaps the noise would appear on DVDR now if I tried (maybe the tape has somehow degraded, although seems unlikely)

This movie varies in how bad certain issues are - outside daylight scenes look awful and I didn't post any of those clips. After comparing many different scenes, the USB-Live 2 wins for me by quite a lot in overall perceived quality

I wasn't sure how much people would still be interested in these clips - I can post some different tapes if there is interest. I also have this same tape recorded by the ES10 if that's of interest for comparison

Also, regarding inserted frames - it could be that when I encoded the clips to x264 I left the framerate at 29 instead of doing an IVTC to get it back to 23.976, which has worked very well for this recording.

Maybe user hodgey can help you there, he was referring to a "audio/control head that picks up the tracking signal from the tape, or something electronically with the servo circuitry" in this post: https://www.digitalfaq.com/forum/vcr...html#post87137, and maybe the better defined audio of the USB-Live 2 is revealing it. Just a guess.

Ok here's a variety of clips that are untouched aside from clipping them out. I haven't done anything to them. I tried to get a decent assortment of different light/dark levels

I didn't realize originally that you are only limited to the size upload per upload, I thought it was the total attachment size. So if some of the clips are too short to use properly, let me know

The new raw samples confirm the previous impressions, and the differences are even more evident.

I quickly equalize the brightness, and the details are different in the JVC capture:
Attachment 15711

The image comparison here: https://imgsli.com/MTMwMDc4

The distortion is not present in the new JVC sample, it was probably due to a bad postprocessing: https://imgsli.com/MTMwMDM0

Thanks for your samples, and good luck with your captures!

Very keen eye, lollo2. Thanks for sharing the comparison. The USB Live2 looks better to me just on initial glance overall, but I hadn't noticed those levels of difference when zoomed in on the detail

-- merged --

I've done a little more testing and I think the audio noise issue I was having is mostly fixed by fine adjustments on the manual tracking. Although the tracking seemed to be set well, hitting the tracking button a few times seems to find a magic spot for the audio where it mostly avoids the noise. So it seems it was not an issue with the USB Live2, just for reference

-- merged --

Ok, I'm capturing a new tape and its very dark so I have to make some levels adjustments. I've been experimenting with the best way to do this and wanted to try just using the Proc Amp settings available in the USB Live2 Input Device Settings (it has brightness, contrast, hue, saturation, sharpness in the proc amp settings)

When I adjust in the settings, it clearly shows the changes in the preview window but when I hit capture to start recording, the capture display resets to the default values and it ignores the proc amp settings entirely. It doesn't actually reset the settings, because they remain the same, it just captures at the default settings regardless.

I'm using AmarecTV 3.10 and after a little searching, it seems I'm not the only one to encounter this:
https://www.digitalfaq.com/forum/vid...ive2-proc.html
https://forum.videohelp.com/threads/...Dub-Don-t-Work

Doesn't look like a solution was ever figured out.

Any chance you know what is going on here, lollo2?

I know levels can be adjusted in Avisynth, but I find it easier to do it via the Proc Amp in the initial capture, if possible (or at least get it in the ballpark)

That's different. The procamp setting prior to capturing are to avoid to crush the blacks, because the USB-Live can only capture 16-254 range.
With AmarecTV it works for me: Attachment 15725

If you experience a reset of your settings when you start recording, try to use GraphStudioNext: open the program, load the filter "Hauppauge Cx23100 Video Capture", go to the procamp menu and set the values. Then start the capture.

You can also use the AviSynth histograms before capturing to check the setting on a significant segment of the video, which is a wonderful feature: https://forum.videohelp.com/threads/...TV#post2614699

And you can also capture with GraphStudioNext building a dedicated graph, but that's another story, and the audio/video synch only relies on the presence of the coming audio and video frames at the right time, so if there is any dropped frame, the capture will be asynch:
Attachment 15726

Ah, that worked great. Thanks! I'm not sure why I can't use the settings in AmarecTV but as soon as I close the settings menu, the capture window goes black for a second and the picture resets to the default values. The Procamp settings are still where I set them, but its not applied. In any case, looks like GraphStudioNext will do the job

I'll also have to play with viewing the Histogram levels in Avisynth for the captures to get everything right. Thanks for the tip

Lollo, is it valid to use the Virtual Dub histogram to set the levels, then capture with AmarecTV (if necessary, using Graphstudio)?

Sure, you can fix the levels with Virtualdub Histogram utilty, and apply the values later in Amarec.

The AviSynth Histogram approach should also work with VirtualDub (I never tried), allowing a more detailed analysis; but it is only a nice to have feature, not strictly necessary.

You need to understand (or remember) certain aspects of AmaRecTV

It was newer software, only around for about 6 years (late 00s to 2013/4). Far fewer people used it. So it never had the bugs sorted out like the VirtualDub capturing function. And it has MANY issues. When it works, it (usually) works well. But more often it doesn't. It hates quite a few capture cards too, so the scope was always narrow, probably mostly focused on cards available in Japan in the late 2000s (ie, not too many).

The dropped frames issue needs further study. I'm not convinced that it magically doesn't drop frames, as compared to VirtualDub. The more likely issue is in how drops are reported, at best how badly timed frames are handled. It's not magic, something is going on there, and to date nobody has explained it (because nobody probably knows how). Remember, you can easily make VirtualDub "not drop frames" by simply ticking those two upper timing settings, which essentially just disables the reporting.

AmaRecTV mostly works with several Hauppage cards, Japanese cards, and maybe something like the VC500 (which is a crappy card, AGC issues).

So if it works for the basic functions as needed, and you're not seeing dropped/duped frames (LOOK FOR THEM!), then don't push your luck. Those "other features" of the software are not suggested. Proc amp controls in-card are already digital, so you can just do it in software later anyway (and probably much better than the weak AmaRecTV controls), it's not an actual analog domain proc amp.

Well, I think that you have to understand AmaRecTV. I won't comment any aspect of your reply, complete non sense :huh1:

No, not at all. I'm just acknowledging weaknesses, flaws, and letting others know.

When it comes to this software, do try it if needed, but proceed with caution. It's not the best software, but has narrow uses, not too different from other aspects of video hardware and software.

What I don't want to read is cheerleader-like glowing gushing praises, as it doesn't deserve that whatsoever. Realism needed. I do this with anything, be it JVC VCRs, ATI AIW cards, or various TBCs. It ain't all roses. When things get a bit too rosy, time for a douse of cold water, back to reality. That's what keeps you grounded, when it comes to ... well, anything, really. But for sure SD analog video capture. I get tired of seeing undeserved over-the-top praises for JVC 9800 decks, TBC-1000, etc. These are tools, not lovers you're trying to bed.

Understanding is knowing.

^ and I surely stole that from a 1980s cartoon! :laugh:

Nobody did about this topic, in this thread or others.

What I read, on the other hand, are non sense comments about unknown issues, dropped frames assumptions never rised and out of contest, confusion about procamp setting of the card mixed with the capture software capabilities, AmarecTV nice for Hauppauge cards while in the last days we had 3 ATI USB users solving problems with it, and so on. There is enough.

Today is not your best "posting" day, LS, it happens sometimes (not stolen from any cartoon :))

Perhaps. But I feel a vibe of it, as interpreted by novices, even if not stated outright by older members. Video newbies dart about too much at times, seeking quick answers, not fully vetting everything from top to bottom. As mentioned, douse of cold water, slow down, understand, read, proceed carefully, not magic.

Not nonsense. I was very clear. If you can explain the "no drops with AmaRecTV, but drops with VirtualDub", then please feel free to do so. I've yet to read anything where this has been addressed. Drops happen for a reason, generally in hardware. Software more often fails to report drops, not somehow prevent or avoid drops. We know this fact. What we don't know is what AmaRecTV is doing.

Perhaps.
But I don't know about the toons thing. I'm not sure Megatron never uttered "Today is not your best Prime". :laugh:

Are you serious? Nobody ever said that sillness, because it would be false.

What we know is that AmarecTV was able to keep audio/video in synch where VirtualDub failed in the examples that have been posted. The rest is "your" crusade.

For sure!

But how, why? Again, lack of drops can be a mere lack of reporting drops. It needs detailed analysis, both the captured clips, and the software itself. At this moment, I see zero reason why VirtualDub would drop, and AmaRecTV not. I would run these tests myself, but simply lack time. Hence my warnings that it's not magic, and that it needs further vetting and understanding -- aka, proceed with caution. Test, verify, don't just capture blindly, assuming all is well. That's not nonsense, mere good science.

If you see no drops on your captures, great! :congrats:

But others may not have that same experience. So again, test, verify.

Same here; again, nobody say that.

The results from others users in the last days have been properly reported.

And on my side, you can be sure that on my capture I check frame by frame, step by step, what's going on in all different test cases, with VirtualDub, AmarecTV or GraphStudioNext. And remember that for some of them I can also compare with a dvb-s dump of the same program, so I know exactly how many frames should be in the capture, without relying on false VirtualDub reporting. Yes, VirtualDub sometimes does not report dropped frames, even with the proper setting, known issue, the same for other capture software; but that's another story

I've never seen that with v1.9.x, not once (that I can recall) ever, in 10 years of use. Other versions, sure, especially the FM/2 forks and 1.10 official. VirtualDub has issues, but that's not one of them.

Now, drivers, hardware, for various cards, would not doubt it.

Because you have nothing to compare with

I'm clearly no professional and its unlikely I will ever dive as deep into understanding all the nuances of video that you guys have a solid grasp on, but I have been working with video on a amature/hobby level for a while - mainly assembling rare old movies from various sources. I've only just started out really trying to do my own captures from VHS (although I have done some basic DVDR recordings in the past) and the USB Live2 has given me drastically better results than any other device I've tried.

There are a lot of barriers in 2022 to having a proper setup to capture video from VHS. Finding the necessary gear is quite hard and quite expensive if you can find it... I certainly do not have the income to justify spending what it would cost to get a TBC, high end VCR, a good hardware proc amp, etc..

So for someone such as myself, the USB Live2 is really quite the find, even if it may have some potential flaws and AmarecTV is working great for capturing for my needs. Although I'm not examining my captures to same levels as you all are in regards to dropped frames and whatnot, my eye is a bit obsessive about more noticeable issues with bad frame motion, and low quality issues that are pretty common to bad captures. The captures I'm getting from the USB Live2 w/ AmarecTV are excellent in this regard - I can watch the full recording and not have any issues that really distract me from the movie. Comparing the recordings to other captures I have of the same VHS source (done by others, not by me), the detail improvement is quite noticeable..

Anyway, just chiming in with my perspective which maybe isn't so useful to professionals or more high end users, but to a user like me, this is really quite an appreciable setup for what I have access to

No. :rolleyes:

Dropped frames really have nothing to do with "being professional" or needing a "deep dive". It's essentially what you state: noticeable issues, even to grandmas and amateurs.

However, details matter.
- Are the frames being dropped, and not reported?
- Are the frames being duped instead of dropped?

Those issues are not visible at a glance, but can be with an actual watch, or edit.

So it's why you must proceed carefully, as this software is still somewhat of a mystery, even to those who suggest it.

Do you know how many times people have hastily transferred lots of video, without properly vetting the results? Too many to count. The quality sucks on watch, and people often opt to redo their mistakes. I try to save others from that fate. Do it once, do it right.

I've been consistent in my posting here. Proceed carefully, test, verify, vet. Do not assume, and realize software isn't magic. I don't think that's controversial whatsoever.

If you've truly have gotten good results, and do not later have regrets, great! :)

For what its worth in regards to my comments here, I spend quite a lot of time looking at the results of my captures and experimenting with applying filters in Avisynth/VDub for any issues... including taking many smaller clips to test settings on. I generally spend many hours just checking my settings and trying to get everything right, and always watch the entire movie after doing the final processing to check for problems.

Sometimes I kinda hate the fact that I obsess about these things so much because I end up looking at the scenes so many times that it kind of ruins the movie for more, or at least changes the viewing experience...

Excellent. :congrats:

It's that sort of experience that makes it even more curious, and important, to understand why AmaRecTV is working when VirtualDub is failing. Because that would mean VirtualDub can fixed, as both are just software.

Yep, that's the downside of this. :(

Imagine working on something like Marvel or Star Wars movie (or tv series). You'd be ruined to its enjoyment. There are some actors, like Tom Hanks and Johnny Depp, that never even see their own flicks, don't care, enjoyment was already ruined for them, just a job. When I worked for studios, I saw some neat stuff, but watching it after working on it almost felt like a chore.

So for home movies, you often end up doing it for the enjoyment of family, not necessarily yourself.