digitalFAQ.com Forum

digitalFAQ.com Forum (https://www.digitalfaq.com/forum/)
-   Capture, Record, Transfer (https://www.digitalfaq.com/forum/video-capture/)
-   -   VCR TBC off if also using ES10 passthrough? (https://www.digitalfaq.com/forum/video-capture/10447-vcr-tbc-off.html)

benzio 03-21-2020 07:15 AM

VCR TBC off if also using ES10 passthrough?
 
1 Attachment(s)
Hello!

It's often said that if you intend to use the ES10 as a passthrough you have to turn off the TBC of the VCR.

I read in this forum that "ES10 needs bad signal to clean it up". What are the principles behind this statement?

I've tried to experiment some combinations that leads to different chroma noises (see attachment).
Top row: U channel Even - U channel Odd
Bottom row: V channel Even - V channel Odd

Frames 0-150 - JVC S9600EU VCR TBC on / NO ES10 PASSTHROUGH
Frames 151-300 - JVC S9600EU VCR TBC on / ES10 PASSTHROUGH
Frames 301-450 - JVC S9600EU VCR TBC off / ES10 PASSTHROUGH <---USUALLY RECOMMENDED COMBINATION

Despite that the third combination is the most recommended it seems to be the worst of the three.
The lack of VCR TBC leaves out a lot of noise.

I read that if the TBC of the vcr is on then the ES10 does nothing, but this is apparently not true: it's stabilizing the remaining noise on the side of the picture.

Is it true, instead, that if I use both the VCR TBC and the ES10 it happens that sometimes the picture "jumps" up and down for some reason.

What's the theory behind this?
Is there a way to take the benefits of the VCR TBC and the benefits of the ES10 without tradeoffs?

Thank you!

scharfis_brain 03-21-2020 08:01 AM

Also do one capture like this:
JVC S9600EU VCR TBC off / NO ES10 PASSTHROUGH
You will see that there is the same amount of noise as with the last portion of your sample.

JVC VCRs seem to apply some sort of temporal denoising when the TBC is enabled.
You certainly don't want that, because it temporally smears the video and limits post-processing in software be a fair amount.

The DMR-ES10 just does TBC but has a noise filter, which you can disable (which you probably did).

So essentially you're getting everything (including the noise) which is present on the tape.

I personally don't care so much on the noise. I prefer a very stable drop-free capture.
Noise can always be removed after the capture is done.

benzio 03-21-2020 08:22 AM

Quote:

Originally Posted by scharfis_brain (Post 67400)
Also do one capture like this:
JVC S9600EU VCR TBC off / NO ES10 PASSTHROUGH
You will see that there is the same amount of noise as with the last portion of your sample.

JVC VCRs seem to apply some sort of temporal denoising when the TBC is enabled.
You certainly don't want that, because it temporally smears the video and limits post-processing in software be a fair amount.

The DMR-ES10 just does TBC but has a noise filter, which you can disable (which you probably did).

So essentially you're getting everything (including the noise) which is present on the tape.

I personally don't care so much on the noise. I prefer a very stable drop-free capture.
Noise can always be removed after the capture is done.

Thank you! Very clear!
If I understood in a correct way the tbc of the JVC has two functions: 1) It gives a stable timecode to the lines of the video 2) It applies a noise reduction that can be obtained also in post production.

The Line Tbc of the ES10, on the other way (not its frame stabilization-syncronization feature) only gives a stable timecode.

Therefore is preferable the recommended setting: No TBC from the VCR when there is the passthrough with the ES10.

lordsmurf 03-21-2020 09:12 AM

Quote:

Originally Posted by benzio (Post 67399)
I read in this forum that "ES10 needs bad signal to clean it up". What are the principles behind this statement?

That makes no sense as stated. However, I would guess that the person refers to using the ES10 on an "as needed" basis. The ES10/15 is not supposed to replace a TBC at all times, and cannot do so anyway. It has weak+strong line TBC functions (weak at allowing errors, such as anti-copy; strong due to anti-tearing), but lacks frame level TBC. To make the ES10/15 a "TBC replacement" (poor man's TBC), create a TBC(ish) setup, you would been to add the DataVIdeo DVK-100/200 units, or the TBC-5000. But realize that the ES10/15 has drawbacks, mostly being posterization (color palette compression/loss) and NR that is always on (to some degree). The ES10/15+DVK/500 gives about 99% TBC function.

Quote:

Frames 301-450 - JVC S9600EU VCR TBC off / ES10 PASSTHROUGH <---USUALLY RECOMMENDED COMBINATION
Despite that the third combination is the most recommended it seems to be the worst of the three.
The lack of VCR TBC leaves out a lot of noise.
The ES10/15 is meant for "least-worst" scenarios, namely tearing. If you attempt to use it at all times, you will see NR/posterization related artifacts.

Quote:

I read that if the TBC of the vcr is on then the ES10 does nothing, but this is apparently not true: it's stabilizing the remaining noise on the side of the picture.
That cannotbe true. It's not possible. The 1st line TBC owns the signal, the 2nd can do nothing. The ES10/15 does have a frame sync (not frame sync TBC), which is often useless. But I have no doubts that it will sometimes do something, because statistics would suggest it. You may be that 1 in a 100 person. Congrats. (Buy a lottery ticket, too! :laugh:)

Quote:

the picture "jumps" up and down for some reason.
This is usually caused from start line mismatch. In theory, I guess the JVC and ES10/15 cold be competing to own the signal?

Quote:

Originally Posted by scharfis_brain (Post 67400)
JVC VCRs seem to apply some sort of temporal denoising when the TBC is enabled.
You certainly don't want that, because it temporally smears the video and limits post-processing in software be a fair amount.

No. :no2:

The main source is JVC NR is the picture settings, not the TBC. Yes, the TBC does emit some NR, as does a Panasonic, but not using the TBC will result in a far inferior and degraded image. The minor NR artifacts would be far better than an untamed wiggly mess of a signal.

Quote:

The DMR-ES10 just does TBC but has a noise filter, which you can disable (which you probably did).
No. :no2:

The ES10/15 "off" is never off, just reduced. This is a DVD recorder, and DVD-spec MPEG is ruined by artifacts when noise is present. This was a dummy-friendly consumer machine, and NR is always there to assist in the DVD creation.

Quote:

Noise can always be removed after the capture is done.
No. :no2:

Many noises cannot be removed post-capture, and that especially includes line timebase correction. Do not turn off the line TBC in the VCR unless it is causing significant problems to the point where the image is unviewable. Or in cases when the line TBC duty is instead being passed downstream, such as ES10/15 units.

Quote:

Originally Posted by benzio (Post 67401)
If I understood in a correct way the tbc of the JVC has two functions: 1) It gives a stable timecode to the lines of the video 2) It applies a noise reduction that can be obtained also in post production.

No. :no2:

A "timecode" is completely different jargon, the TBC corrects signal timing. I'm sure this what you meant, but terms matter, especially when trying to learn (and others are learning from reading these threads). And as per above, not everything can be address post-production, it gets "baked in" to the signal.

Quote:

The Line Tbc of the ES10, on the other way (not its frame stabilization-syncronization feature) only gives a stable timecode.
It's not a true "line TBC" as it allows some errors to pass. And then again with "timecode", as per above.

Quote:

Therefore is preferable the recommended setting: No TBC from the VCR when there is the passthrough with the ES10.
Yes -- but ONLY when the ES10/15 is in the chain. The VCR line TBC will prevent the ES10/15 from working at all, or give some sort of weird conflict as you've observed. When not using ES10/15, always use the TBC (aside from rare cases where the signal is made worse, and most people will never see that happen, it usually has to be a really screwed-up nth-gen tape for it to happen).

hodgey 03-21-2020 10:05 AM

The TBC in the VCR digitizes and buffers some amount of the video signal (not sure how much). It tries to locate the start/end of each line, and resizes it to the correct length and replaces the horizontal synchronization signal (that tells where a line starts) with a clean newly generated one, and converts it back to an analog signal.
additionally, in the JVCs, the same chip also runs digital noise reduction on the digitized signal, which especially helps with chroma noise. This comes in addition to the more rudimentary analog noise reduction in the VCR.
Since the horizontal synchronization is replaced, the ES10 (or any other device) will use that to determine the line start, so if the VCR TBC is not able to fix things fully, the errors will be baked into the video and the ES10 can't do anything more about it.
What it does not do from what I've seen is to guarantee a completely stable 50/59.97 fields per second output signal with all timing correct, hence why it's suggested to put a a full-frame TBC or DVR after it in the capture chain.

The ES10 can buffer two (or more) frames (don't know how much exactly.) It will digitize the video, try to resize the video lines based on detected line starting points in the process (like the VCR TBC). It can also apply noise reduction if enabled (the PAL ES10 doesn't have the permanent-on posterization effect that the US one as far as I know.), and show the on-screen display over the video.

It will then convert the digitized frames/fields back to a stable analog signal with new vertical and horizontal synchronization signals. Later DVRs could also output digitally over HDMI. I can't say for 100% sure with the NTSC models, but at least the on the PAL ones the output is a fully stable video signal (though it will have baked in any errors that the ES10 could not deal with from the original video from the VCR.) This means that it can be helpful even though the VCR TBC is turned on.

What it can do though I think is to add in a macrovision signal if it thinks there is one on the input side which can upset some capture cards. (Though in that case the video tends to get messed up by the ES10 in any case.) and it can turn off the output if it thinks there is nothing connected to the input. I've never had any issues with just sending the output from it straight to a capture card, but YMMV.

The "jumping" effect is something caused by the TBC in the JVCs. I don't know what causes it exactly, but I have noticed it a lot with all the JVCs I've used, especially on camcorder tapes where the camera is moved around a lot. So much so that I usually use them with the built-in TBC turned off and use an ES10 or similar to correct wiggle instead (even though the built-in TBC can usually stabilize horizontal jitter slightly better.) The TBCs in our Panasonic NV-HS1000 and JVC SVHS camera doesn't seem to have the same issue. YMMV though.

scharfis_brain 03-21-2020 12:22 PM

I second hodgeys post.
The European ES10 has a noisefilter that can be turned off completely. Also it does NOT do any MPEG-Preprocessing. I've done extensive tests regarding this matter.

It just passes through the signal and removes any tearing/jitter from the instable signal. There is no posterization.
The signal (regarding timing) is completely pristine no matter what has been sent to the input. The capture card will always see a proper signal to lock on. No Framedrops will occur. Of course in tough situations this signal may contain garbage image, because the ES-10 itself wasn't able to lock onto its input.

The ES-10 completely outperforms the line-TBC of the Panasonic NV-1000 regarding image stabilization and luma-range. The TBC of the NV-1000 clips blacks and whites and often misses vertical frame sync.

The only drawback on my ES-10 I was able to see is an additional vertical chroma shift by one line due to the necessary PAL-chroma decoding (delayline) of the S-Video input. But that is easily fixed in post processing.

I once owned a JVC S-VHS VCR with TBC. It was completely useless to me, because it ALWAYS applied noise reduction. After years in the shelf I put it to the bin.

Since I always deshake my digitized home videos, a temporal noise reduction before capturing (like JVC does) will completely ruin the result: Deshaking temporally denoised video will make denoised areas of the video dance around!

I really prefer a higher image noise level (not jitter/tearing) over a temporally smeared output of JVC recorders. I always can remove temporal noise in AVIsynth quite effectively.

benzio 03-22-2020 07:23 PM

Quote:

Originally Posted by lordsmurf (Post 67404)
The ES10/15 is meant for "least-worst" scenarios, namely tearing. If you attempt to use it at all times, you will see NR/posterization related artifacts.

I'm always using it because I've a lot of tapes to do and I'm quite lazy... I didn't want to change the wiring of the cables in relation of the vhs I'm seeing... I've discovered that the JVC + ES10 goes well 95-98% of the time, and the other times I use a Panasonic NVHS1000.

I don't have a real tbc... Yes I've a AVT 8710... but it's a faulty one! I've tried to use it and it messes up fields in a bad way.
I've seen one TBC1000 on ebay... it's €1400. There are no tbc in the marketplace here, and even if they were probably would be at very high prices anyway. I know that it's an investment, and I can resell it, but I don't have so much liquidity.

So you are recommending that in most cases is better to just capture by wiring the VCR directly to the capture card, without pass-trough to the ES10, and use the ES10 only if there is evident tearing -IF YOU DON'T HAVE A REAL TBC-?


Quote:

Originally Posted by lordsmurf (Post 67404)
That cannotbe true. It's not possible. The 1st line TBC owns the signal, the 2nd can do nothing. The ES10/15 does have a frame sync (not frame sync TBC), which is often useless. But I have no doubts that it will sometimes do something, because statistics would suggest it. You may be that 1 in a 100 person. Congrats. (Buy a lottery ticket, too! :laugh:)

But the sample I've posted shows that... What's that strip of blurring of the chroma to the side of the frame, that there is in the sample with pass-through but not without pass-through?

Quote:

Originally Posted by lordsmurf (Post 67404)
No. :no2:

A "timecode" is completely different jargon, the TBC corrects signal timing. I'm sure this what you meant, but terms matter, especially when trying to learn (and others are learning from reading these threads). And as per above, not everything can be address post-production, it gets "baked in" to the signal.
It's not a true "line TBC" as it allows some errors to pass. And then again with "timecode", as per above.

Yes I was misusing the term.
Timecode -> Something to track down the point on the tape you are reading
Signal Timing -> Something to give stability to each line of a field.
Right?


Quote:

Originally Posted by hodgey (Post 67407)
The "jumping" effect is something caused by the TBC in the JVCs. I don't know what causes it exactly, but I have noticed it a lot with all the JVCs I've used, especially on camcorder tapes where the camera is moved around a lot. So much so that I usually use them with the built-in TBC turned off and use an ES10 or similar to correct wiggle instead (even though the built-in TBC can usually stabilize horizontal jitter slightly better.) The TBCs in our Panasonic NV-HS1000 and JVC SVHS camera doesn't seem to have the same issue. YMMV though.

I noticed same issue as well... That's a pity that the JVC TBC has this problem! What could be the cause of that?

benzio 03-22-2020 07:38 PM

Quote:

Originally Posted by scharfis_brain (Post 67410)
I second hodgeys post.
The European ES10 has a noisefilter that can be turned off completely. Also it does NOT do any MPEG-Preprocessing. I've done extensive tests regarding this matter.

It just passes through the signal and removes any tearing/jitter from the instable signal. There is no posterization.
The signal (regarding timing) is completely pristine no matter what has been sent to the input. The capture card will always see a proper signal to lock on. No Framedrops will occur. Of course in tough situations this signal may contain garbage image, because the ES-10 itself wasn't able to lock onto its input.

The ES-10 completely outperforms the line-TBC of the Panasonic NV-1000 regarding image stabilization and luma-range. The TBC of the NV-1000 clips blacks and whites and often misses vertical frame sync.

The only drawback on my ES-10 I was able to see is an additional vertical chroma shift by one line due to the necessary PAL-chroma decoding (delayline) of the S-Video input. But that is easily fixed in post processing.

I once owned a JVC S-VHS VCR with TBC. It was completely useless to me, because it ALWAYS applied noise reduction. After years in the shelf I put it to the bin.

Since I always deshake my digitized home videos, a temporal noise reduction before capturing (like JVC does) will completely ruin the result: Deshaking temporally denoised video will make denoised areas of the video dance around!

I really prefer a higher image noise level (not jitter/tearing) over a temporally smeared output of JVC recorders. I always can remove temporal noise in AVIsynth quite effectively.

Very interesting! Thanks!

hodgey 03-22-2020 09:31 PM

Quote:

Originally Posted by benzio (Post 67445)
So you are recommending that in most cases is better to just capture by wiring the VCR directly to the capture card, without pass-trough to the ES10, and use the ES10 only if there is evident tearing -IF YOU DON'T HAVE A REAL TBC-?

If you don't have a TBC I would use it, the posterization issues that LS is talking about aren't present on the PAL models. It will help prevent frame drops similar to what a TBC would (in most cases it works as well). Though be aware that the ES10 can clip very bright spots if the video signal level is a bit high.

Quote:

But the sample I've posted shows that... What's that strip of blurring of the chroma to the side of the frame, that there is in the sample with pass-through but not without pass-through?
The ES10 will blank 8 pixels (I think) on each side (the VCR TBCs blanks a bit of the sides as well for that matter to re-create the horizontal synchronization bits). There usually isn't anything there anyway, but occasionally there can some visible signal, e.g with an off-TV recording where there was ghosting from the antenna. Maybe that's what you're seeing here, some noise that is being blanked out.

benzio 04-01-2020 08:09 PM

I found a bizarre counterexample to the rule "Line Tbc of the VCR when ES10 is in the setup".

I'm able to play properly an LP tape ONLY if I use Line TBC of the Panasonic HS1000 + Passthrough ES10.
Just the Line Tbc, Just the Passthrough or Just Vcr without Line TBC does not work (too many "jumping-frame" errors).

How is this behaviour possible?

msgohan 04-04-2020 06:03 AM

"Jumping frame" errors = (field or whole frame) vertical jitter / bounce / hopping. Right?

In order to perform line TBC, the VCR has to do sync separation. Apparently in this case, its sync separator is recovering the VSYNC and identifying the start of fields/frames more accurately than the passthrough DMR-ES10. And your capture card is choking on the VCR output regardless for some reason.

benzio 04-04-2020 08:08 AM

1 Attachment(s)
This is the "jumping FIELD" (not frame) I was talking about.

What's the name of this behaviour?

msgohan 04-04-2020 08:25 AM

I listed the names I've seen. The most common term is vertical jitter but I personally dislike this because "jitter" has at least three other definitions in video/audio. I think Lordsmurf shares this feeling, since he always types out clarifications of vertical & horizontal jitter whenever he's responding to a user about either one.


All times are GMT -5. The time now is 06:35 PM

Site design, images and content © 2002-2024 The Digital FAQ, www.digitalFAQ.com
Forum Software by vBulletin · Copyright © 2024 Jelsoft Enterprises Ltd.