AVT-8710 TBC & Lack of RCA Audio Input/Output - Will It Matter?

[NY2LA]

I just received my AVT-8710 time base corrector by U.P.S. ground this afternoon, for putting between my Toshiba SD-V296 DVD/VCR Combo, and Canopus ADVC110. I haven't tried it yet.

I have to use its Composite Video Input and Output. I wasn't prepared for no RCA Input and Output for the audio.

Not that I want the AVT-8710 to do anything to the audio, mind you. Sending the audio straight to the Canopus ADVC100 for encoding makes sense.

But is the video and audio going to stay in synch? Considering the video passes through a buffer in the AVT-8710, isn't that going to take a few nanoseconds that will add up? (I don't know anything about this. Please explain.)

I'll find out whether I have synch problems soon enough, when I try it, I suppose. Nevertheless, if you will, please explain how the AVT-8710 will clock the video and not the audio, yet things will stay in synch.

[lordsmurf]

The offset is about 1/30th to 1/60th of a second. No, it won't lose sync.
Video and audio are never 100% perfectly in sync anyway, because light and sound travel at different speeds. There's always an offset, from the moment a camera captures it.

Nothing can "add up" -- offset is constant.

[NY2LA]

Thank you, lordsmurf. So far, I've been experiencing such incredible weirdness using the AVT-8710 that I haven't been able to really evaluate the video/audio synch.

When I have everything ready, including Composite Video going from the AVT-8710 to the Canopus ADVC110, the Canopus ADVC110 no longer defaults to Analog In. (It used to always immediately default to Analog In as soon as I hooked up the powered VCR to it, even before playing a tape.) Instead, it defaults to Digital In. Strange. And holding Input Select on the Canopus ADVC110 for the full duration of it producing the color bars only temporarily lights the Analog In while Digital In stays on; afterwards, it defaults back to Digital In.

I captured a problematic VHS tape of about ten minutes which previously resulted in skipped frames, using iMovie 6.0.4 (an old version I still like to use, because of its Full Quality DV export feature, but that's another story.) The ADVC110 did not switch to Analog In until I clicked import while the tape was already playing. I captured this tape twice using the AVT-8710. First time, it helped get through the section that had previously resulted in skipped frames, but was not a cure all. The problematic section is primarily during rolling credits, and the rolling credits still jump around in the resulting capture file. (I don't know how to describe it.) Second time, it led to double images where the skipped frames had previously occurred!

I have been unable to use the AVT-8710 with Final Cut Express at all so far, perhaps because the ADVC110 insists it is getting a digital signal. I can't get the FCE's DV-NTSC Converter setting preview interface to recognize an analog signal at all.

I tried holding the reset button on the AVT-8710 for over three seconds. All that did was make the "A" LED go off. Unplugging and plugging the device has not made the "A" LED go back on.

[NY2LA]

Correction:

I wrote: "I can't get the FCE's DV-NTSC Converter setting preview interface to recognize an analog signal at all."

I meant: "I can't get the FCE's DV-NTSC Converter setting preview interface to recognize any signal at all.", because it is over Firewire at that point.

[edDV]

ADVC-110 dip switch 3 on the bottom selects power on default. Off = Analog. On = Digital.

See P-16 of the manual.
http://www.pro-video.it/provideo-incorso/images/File/Canopus/ADVC110%20Manuale.pdf

[NY2LA]

ADVC-110 dip switch 3 on the bottom selects power on default. Off = Analog. On = Digital.

See P-16 of the manual.
http://www.pro-video.it/provideo-incorso/images/File/Canopus/ADVC110%20Manuale.pdf

The Power-on Input Mode DIP switch is No. 2 according to my manual. Anyway, I've always had that switch to Off (for Analog), and haven't changed it. I just checked, and it is still set to Off.

[NY2LA]

I finally got the "A" LED on the AVT-8710 to come back on by holding down the Reset button. The "N" LED has been on, and is still on. The System LED has always been on, and is still on. Is it supposed to be on when I am intending to use the default settings?

[edDV]

Sw 2 correct. For the ADVC 100 it is switch 5.

[NY2LA]

I finally got the Canopus ADVC110 to default to Analog In by shutting down my computer, and powering down the ADVC, while keeping everything else hooked up. Then, upon starting up the computer again, the ADVC110 did what it is supposed to do, default to Analog In.

(I use a MacBook Pro from 2006, with one Firewire 400 port. I use the Firewire 400 port exclusively for the ADVC, which powers it. I connect an external drive via a Firewire 800 ExpressCard. Canopus tech support advised against using the ADVC with a hub, and advised against chaining.)

But I'll be posting again soon, with comments about the results I'm getting from using the AVT-8710.

[NY2LA]

I guess I'll always be encountering a new issue to learn about every day.

Frequently, I check captures by playing the media file with Quicktime Player, rather than through the non-linear editing program's interface. A moment ago, I tried to find the aforementioned double images, which I could swear I had observed were digitized into the capture.

Well, I didn't see the double images this time when examining the file with Quicktime Player. I thought that I must be going bonkers. Later, I discovered that I only see the double images when playing the same clip through the iMovie 6.0.4 interface.

My iMovie Playback preferences are set to Highest (field blending). I don't know what this means. I don't know which presentation mode to trust.

Also, the AVT-8710 has added -- I think you guys call it a hockey stick -- to the top of the image throughout the capture.

P.S. The video and audio are staying in synch.

[edDV]

I always use VLC on a Mac to check file playback. If the file is interlaced, set Video, Deinterlace, Yadif for deinterlaced playback on the Mac monitor.

iMovie will preview the timeline back to the ADVC (switch to digital input) for playback on an attached TV (composite or S-Video out). This way you are directly monitoring the interlace DV timeline. For details check "External Monitoring" in iMovie help.

I've never used an AVT TBC so can't help with that.

[NY2LA]

edDV, thanks for confirming -- it is possible to play DV in a non-linear editing program, through to the ADVC over Firewire, and from the ADVC's analog out to a TV, in order to watch full DV quality.

I've been intending to try this, and finally did so today. I hooked the ADVC analog out to a 720p HDTV, which does a swell job of upscaling. I went into iMovie 6.0.4's Preferences, and checked "Play DV project video through to DV camera."

I played a few old iMovie projects consisting of DV, and they looked good on the HDTV from the proper viewing distance. I should be checking all of my captures this way.

However, when I tried to play my recent captures done with the AVT-8710, I was only able to get audio to play on the HDTV. Same iMovie program. Same set-up. Something seems to be wrong with those captures; their presentation in the iMovie interface is not the only issue.

(By the way, I've already been hooking up the ADVC110 analog out to the HDTV, to monitor the video and audio during capture. I find this alerts me to capturing issues that I might not see if I hooked the VCR directly to the TV.)

Yesterday, I recaptured a one hour long problematic pre-recorded VHS tape, using the AVT-8710. Previously, without the AVT-8710, this videotape captured perfectly fine, until about 52 minutes into the video. At that point, I observed intermittent "No Signal" messages on the HDTV during capture, and skipped frames were in the capture file. When recapturing this videotape using the AVT-8710, it seemed to do fine until, once again, about 52 minutes into the video. Then, instead of "No Signal" messages on the HDTV, I observed artifacts, maybe interlacing artifacts.

Today, I watched the entire recapture, purposely within the iMovie interface. For myself, I confirmed that the double images do not occur until the same point in the video, at about 52 minutes.

The results I'm getting with AVT-8710 has led to a bizarre mystery from my perspective.

edDV, even though you've never used the AVT-8710, can you surmise what the heck is wrong with the aforementioned videotape? Whatever it is, the AVT-8710 doesn't fit it. Note: no problems observed when copying this tape to miniDV tape.

[edDV]

I can't add much. It could be tape damage or sticky tape. I always fast wind VHS tape to the end and back before capturing.

Try the bad parts with the TBC bypassed to see if it is causing the "no-signal" message.


PS: If you use an interlace CRT TV for monitoring it will show field detail and flag reversed field order errors. I use both an interlace TV and an HDTV for pre-encode timeline quality monitoring.

BTW: timeline DV monitoring is also possible with Avid, FCE, FCP, Premiere and Vegas.

[NY2LA]

Try the bad parts with the TBC bypassed to see if it is causing the "no-signal" message.

I'm not sure what you mean. As explained, the "no signal" message occurred before I started using the TBC.

P.S. Too bad I took my last Cathode Ray Tube television to the electronic waste dump. Possibly, I shouldn't have done that.

[NY2LA]

I can't add much. It could be tape damage or sticky tape. I always fast wind VHS tape .

But if it is tape damage or sticky tape, why does it play okay in the VCR connected directly to the TV, and copy okay to miniDV tape?

I've had this exact same problem with three or four pre-recorded VHS tapes, out of about thirty that I've captured with the Canopus ADVC110. They always: 1.) Play okay on the VCR directly connected to TV, 2.) Copy okay from one VCR to another VCR, 3.) copy okay to miniDV tape, 4.) During capture with the ADVC110 (no TBC), result in intermittent "no signal" messages on the HDTV connected to the ADVC analog out, and skipped frames in the file, in the problematic parts of the video.

I've worked with the problematic tapes again and again. (The Canopus ADVC110 captured the majority of the tapes very well, without a standalone TBC.) The symptoms are very unique to those particular tapes, and the test results have become very predictable. I can't believe I'm the only person having this problem.

[edDV]

I can't add much. It could be tape damage or sticky tape. I always fast wind VHS tape .

But if it is tape damage or sticky tape, why does it play okay in the VCR connected directly to the TV, and copy okay to miniDV tape?

I've had this exact same problem with three or four pre-recorded VHS tapes, out of about thirty that I've captured with the Canopus ADVC110. They always: 1.) Play okay on the VCR directly connected to TV, 2.) Copy okay from one VCR to another VCR, 3.) copy okay to miniDV tape, 4.) During capture with the ADVC110 (no TBC), result in intermittent "no signal" messages on the HDTV connected to the ADVC analog out, and skipped frames in the file, in the problematic parts of the video.

I've worked with the problematic tapes again and again. (The Canopus ADVC110 captured the majority of the tapes very well, without a standalone TBC.) The symptoms are very unique to those particular tapes, and the test results have become very predictable. I can't believe I'm the only person having this problem.

I don't capture that much VHS with my ADVC100 (yet). I hope some other user joins in.

Somehow the ADVC is losing sync. A TBC with wide sync tolerance is a good strategy. A frame sync/TBC would be the next step.

[NY2LA]

I don't capture that much VHS with my ADVC100 (yet). I hope some other user joins in.

Somehow the ADVC is losing sync. A TBC with wide sync tolerance is a good strategy. A frame sync/TBC would be the next step.

I hope some others join in too.

Here's a screen shot of the Final Cut Express capture interface, showing the VCR's signal going through the AVT-8710 and through the ADVC, prior to inserting a tape. I wonder what the horizontal pattern is at the top. Whatever it is, it is more easily discernible on the HDTV that I use for monitoring than in this screen shot.

[NY2LA]

Re: my above screen shot. This phenomenon happened again, only this time, a recognizable image from the most recent videotape I had attempted to capture remained in the upper part of the screen, rather than an unrecognized pattern, after aborting the capture and removing the videotape from the VCR. A horizontal slice of the top of the last captured frame from the videotape continued to be generated by the AVT-8710, apparently still in its buffer.

(When I posted the screen capture, I apparently hadn't noticed it until attempting to capture the next videotape.)

This happens with the problematic videotapes, so far not with videotapes with no known problems. Does this tell us anything about what is wrong with those videotapes?

[lordsmurf]

That's an issue with the Canopus box or maybe even the Mac hardware/software.
Mac is a crappy system to capture video with.
Crap for capturing, great for editing, so-so for encoding, excellent for authoring.
Mac is really a DV-centric setup, made for working with cameras and not analog inputs.

[NY2LA]

That's an issue with the Canopus box or maybe even the Mac hardware/software.
Mac is a crappy system to capture video with.
Crap for capturing, great for editing, so-so for encoding, excellent for authoring.
Mac is really a DV-centric setup, made for working with cameras and not analog inputs.

The problem has nothing to do with the Mac hardware/software. Please notice that I wrote:

"I've . . . been hooking up the ADVC110 analog out to the HDTV, to monitor the video and audio during capture."

"I observed intermittent 'No Signal' messages on the HDTV during capture, . . . " when not using a TBC. And "instead of 'No Signal' messages on the HDTV, I observed artifacts, maybe interlacing artifacts" when using the AV-8710.

All that is before any signal gets to my computer over Firewire. Using the Mac software video out only applies to playing a capture after it's done.

[NY2LA]

That's an issue with the Canopus box or maybe even the Mac hardware/software.
Mac is a crappy system to capture video with.
Crap for capturing, great for editing, so-so for encoding, excellent for authoring.
Mac is really a DV-centric setup, made for working with cameras and not analog inputs.

I see the phenomenon shown in the screen shot not only in the Final Cut Express capture interface, but also on the HDTV getting a signal straight from the ADVC.

Re ". . .so-so for encoding". The Canopus ADVC110 is doing the encoding, not the Mac software. Final Cut Express is putting the DV data into a Quicktime wrapper, but the DV data is the same bit-for-bit.

Reminds me, elsewhere, I've heard people say the Canopus ADVC encodes to DV-AVI. That can't be. The Canopus encodes to raw DV Stream. Then the editing program being used either wraps the DV data in an AVI wrapper (which I know little about), Quicktime wrapper (as Final Cut does), or leaves it as raw DV Stream (as I know the version of iMovie that I have been using does.)

[edDV]

...
Reminds me, elsewhere, I've heard people say the Canopus ADVC encodes to DV-AVI. That can't be. The Canopus encodes to raw DV Stream. Then the editing program being used either wraps the DV data in an AVI wrapper (which I know little about), Quicktime wrapper (as Final Cut does), or leaves it as raw DV Stream (as I know the version of iMovie that I have been using does.)

Yes, the "capture" software adds the wrapper and creates the file. The DV data is the same in all cases. Type 2 AVI creates a separate copy of the PCM audio to save demux overhead.

[lordsmurf]

Re ". . .so-so for encoding". The Canopus ADVC110 is doing the encoding, not the Mac software. Final Cut Express is putting the DV data into a Quicktime wrapper, but the DV data is the same bit-for-bit.

No. The Canopus is part of the capture. I refer to encoding to final files (MPEG-2, H.264, etc). Mac options for encoding software is only so-so, for the most part. MainConcept Reference 2.1 is now on Mac, so that's helped give Mac some good options finally.

"I observed intermittent 'No Signal' messages on the HDTV during capture, . . . " when not using a TBC. And "instead of 'No Signal' messages on the HDTV, I observed artifacts, maybe interlacing artifacts" when using the AV-8710.

Your use of video jargon may not be accurate. I'm sure you're trying your best to describe what you're seeing, but it's just not coming out as anything that makes any sense. As far as I can tell, either the Canopus box is broken, or the TBC is broken. A piece of hardware is obviously screwed up. I'd bet on it being the Canopus box. And the Mac does play a part in this -- it can badly collect the data being sent by the Canopus box. The computer can affect the DV stream.

This isn't typical by any means.

[NY2LA]

I've used the Canopus ADVC110 to capture over thirty pre-recorded VHS tapes, without the help of any TBC. The Canopus ADVC110 has done a fine job of capturing around 90 percent of those tapes, and I'm glad that I did not introduce a TBC in order to capture them. The problem seems to be unique to "certain tapes" (a phrase that has been used by others on this forum when addressing a similar problem), and capturing with the Canopus ADVC110.

The AVT-8710 did not fix the problem. Instead, it substituted one set of symptoms (skipped frames) with a different set of symptoms (including tearing).

No. The Canopus is part of the capture. I refer to encoding to final files (MPEG-2, H.264, etc). Mac options for encoding software is only so-so, for the most part. MainConcept Reference 2.1 is now on Mac, so that's helped give Mac some good options finally.

My understanding is that the reason Apple created Firewire was so that DV data can be captured bit-for-bit.

Encoding from a non-linear video editing program has nothing to do with the problem that I've been addressing in this thread. I have no interest in compressing to MPEG-2 or H.264 at the moment. I've been talking about what happens during the capturing process, and what I see when I examine the capture files. I am addressing "pre-encode timeline quality monitoring", to use edDV's phrase. And when I use Final Cut Express, I check the Quicktime files in the capture scratch folder. When I use iMovie, I check the DV files in the Media folder (revealed by using "Show Package Contents").

This isn't typical by any means.

Perhaps not typical because most people don't feed an analog signal out of the Canopus ADVC to an HDTV during the capturing process??? Most people feed the signal from the VCR??? Feeding the signal out of the VCR won't reveal the capturing issues that I see when I use the Canopus ADVC110's d-to-a converter during the capturing process.

I don't know how to be more clear about describing an intermittent "No Signal" message on the HDTV. It happens during the capturing process where the Canopus ADVC is skipping frames.

The other symptoms that I see on the HDTV, when using the AVT-8710, may have something to do with the TV upscaling to 720p. But that phenomenon did not occur during the 90% of the captures that went well; again, it is unique to capturing these "certain tapes".

When using the AVT-8710, I don't see the "tearing" or "hockey stick" on the HDTV because it is obscured by the overscan. But I can see it in the capture file. I compare captures side-by-side on my computer monitor. The AVT-8710 definitely added "tearing" to the problematic tapes.

I think others have discussed a similar problem, such as on this thread. Note the original poster uses the phrase "certain tapes":
https://forum.videohelp.com/threads/280939-TBC-Help-Worse-results-with-than-without-%21

[NY2LA]

Somehow the ADVC is losing sync. A TBC with wide sync tolerance is a good strategy. A frame sync/TBC would be the next step.

edDV, will you please elaborate on your suggestion, a "frame sync/TBC would be the next step"? What equipment? Do you mean in addition to, or instead of, the AVT-8710?

Per VCR Troubleshooting & Repair by Robert C. Brenner and Gregory R. Capelo, Second Edition, 1992: "The TV horizontal automatic frequency control circuit plays an important part in reducing the effects of horizontal time-base errors. . . . With the introduction of VCRs, American TVs needed HAFC circuits that provided faster compensation time."

Are you referring to a piece of equipment with this sort of circuit? Does the AVT-8710 lack this?

Bottom line: I've returned the AVT-8710. Keep in mind, 90 percent of my captures of pre-recorded VHS tapes went well without any TBC.

[NY2LA]

Here's a side-by-side comparison of two captures of one of the problematic tapes. At left: Canopus ADVC110 only, no TBC. At right: AVT-8710 between the VCR and the ADVC110. Notice the "tearing" at the top of the capture done using the AVT-8710.

[NY2LA]

PS: If you use an interlace CRT TV for monitoring it will show field detail and flag reversed field order errors. I use both an interlace TV and an HDTV for pre-encode timeline quality monitoring.

I've been meaning to ask you, edDV, to please explain this further. What do "field detail and flag reversed field order errors" look like on an interlace TV? Are these types of errors possibly buried in the DV data of a capture?

[lordsmurf]

This one is easy to figure out!

The tape is the problem. It has bad timing.
Or it may be the VCR that made it -- or even the VCR now playing it. And not the tape.
Read this for detailed explanation: VHS Tearing Even WITH TBC and JVC SVHS Player

Now this error needs to be fixed by an embedded TBC. Not a standalone.
Understand the different types of TBCs at: What is a TBC? Time Base Correction for Videotapes

Your non-TBC Canopus capture has very definite horizontal jitters (squiggling of linearity).
It needs correction.

Your with-TBC Canopus capture has the same distortions. Again, missing a TBC.
This comes back to the VCR. What VCR?
Without reading the whole thread again, I'm guessing it's low-grade consumer junk, yes?
And not a high-end JVC or Panasonic VCR?
Learn more: Best VCRs for Capturing Video a.k.a. VCR Buying Guide (S-VHS, D-VHS, Professional)

Another option is an ES10 between the cheap VCR (or a pro VCR) and the AVT TBC.
More details on what an ES10 is: Fixing VHS Tearing During Capture
Also goes over JVC D-VHS units, which can have similar abilities.

Your problem comes from the fact that the AVT-8710 is making the timing errors permanent.
The Canopus box was mildly correcting the problem, but it still was not fixed.
Honestly, the Canopus version is not much better than the non-Canopus version you show above.
Both are pretty screwed up, in terms of the image geometry.
Look at how wiggly the straight lines are in the architecture. That's terrible.
The uncorrected signal is also causing fluctuating luma values. It's crushed into blacks.

There is excessive chroma noise in the video.
You really do need a better VCR.
Or even a chroma-filtering capture device (JVC DVD recorder).

Pre-recorded tapes are no more special than homemade tapes.
They suffer the same chroma errors, grain noise and time base errors.
In fact, retail tapes are often worse, injected with false anti-copy errors.

The "no content" error is because the Canopus box is unable to lock a signal.
Same for skipped/dropped frames, and possible audio sync issues.
This is why the AVT TBC is required.

So the AVT-8710 has not "added" anything -- it's just not removing what's already there,
either because the VCR is doing it, or because it's present on the tape.

The computer can still be a cause for dropped frames.
Most TBC-corrected dropped frames errors are actually related to the computing hardware.

This was easy to solve, seeing the images, and getting more details from you.

[davideck]

These images demonstrate that the TBC performance of the Canopus is better than the AVT for this type of timing error.

If the AVT is not adding anything, then the Canopus must be doing a better job of removing the timing errors.
If the Canopus is not removing any timing errors, then the AVT must be making things worse.

I suspect that a Philips HDD/DVD Recorder used as a TBC in passthru would remove these errors, similar to the ES10.
It would also provide frame synchronization, so the AVT would not be necessary.

[lordsmurf]

But the Canopus is still missing MANY sync/timing errors, based on the geometric distortions in the architecture. While it does seem to help more with the one specific tearing issue, it fails at everything else. I would argue that it's "fixing" or "removing" the tearing -- it's just reducing it from what I can tell.

Indeed, the Philips may have some degree of corrections. But again, it won't be able to account for all errors. That's why multiple TBCs and frame syncs may be needed. And that starts at having a good VCR.

In fact, there's a chance all of this is the fault of the VCR.