TBC and the ADVC-55

[pgoelz]

I find that enabling the TBC in my AG-1970 makes the captured frame wiggle back and forth noticeably. This starts about 3 seconds after turning the TBC on and continues for as long as the TBC is on. It is visible during capture and in the captured file. The frame is stable when connected directly to an analog monitor.

Is there some sort of oddity between a TBC and the ADVC-55 that would cause this or is there perhaps something wrong with the TBC in the AG-55? It is as if the TBC was outputting stable synch but not clocking the video out correctly.

Paul

[lordsmurf]

There is no TBC in the ADVC-55.

[pgoelz]

The TBC I refer to is in the AG-1970, not the ADVC-55. I know there is no TBC in the ADVC-55. Given that, I am trying to figure out why the captured video wiggles back and forth when the TBC in the AG-1970 is on.

Can anyone with an AG-1970 verify that video captured by an ADVC-55 or ADVC-110 is stable with the AG-1970's TBC on?

Is this perhaps because the AG-1970 has FIELD TBC, not FRAME TBC? Maybe an unstable tape where the two fields are not at the same synch rate confuses the ADVC-55 and it hunts? Funny thing is that captures are 100% stable with the TBC off.

Maybe this is an excuse to look for an AG-1980 with full frame TBC.

Paul

[lordsmurf]

Is it just this tape or all tapes?

[edDV]

The AG-1980 would have a line TBC not field or frame. I don't have one so can't help.

A monitor has a fast h-sync for each line so the signal will appear stable. A capture device clocks at a stable rate so faithfully reproduces all the h-sync jitter.

[pgoelz]

It is all tapes. They all capture about the same with the TBC on.

I thought the AG-1970 had field TBC and the 1980 did the whole frame. I'll have to look again.

Paul

[lordsmurf]

The AG-1980 has a full-field (multi-line) TBC.

[pgoelz]

OK, so help me out here. If my 1970 has LINE TBC, what exactly does that accomplish? Does it attempt to find an average clock rate for the line synch and then clock each line at that rate? If so, I would imagine it would have to adjust from time to time and that would explain the captured frame moving back and forth if the ADVC-55 didn't follow the varying clock rate as fast as the TBC changed it. ??

Paul

[edDV]

A line TBC samples a line into memory with the sampling clock reset from each off tape sync h pulse then clocks out of memory with a stable sync. I.E. the output clock is fixed, the input sample clock re-syncs each tape line. This corrects line to line horizontal displacement but doesn't correct jitter that happens within the line (aka velocity jitter).

If it has a field TBC I'd need to research how this works on a free run (non-genlocked) VCR.

The ADVC (or any other capture device) has a slow sync stable sample clock that settles over a frame or two then samples at a stable rate. Any time base jitter is captured.

[pgoelz]

OK, that makes sense. The line TBC removes line to line displacement but only after the sync pulse. The left edge of the frame could still jitter because it is following the recorded sync line to line.

In fact, the side to side movement I am seeing is more consistent with frame to frame (or field to field?) sync changing and the ADVC not following it as fast as it is changing. The left edge of the frame is completely vertical and straight.... the entire frame moves side to side.

I never thought about it in depth but any TBC HAS to change its output clock rate from time to time, right? If it didn't, it would potentially need an unlimited buffer. My issue seems to be that the ADVC adapts slower than the AG-1970 TBC changes. Fortunately, the non-TBC captures look very good, with only very slight line to line jitter.

Paul

[edDV]

I suspect it is an artifact of a field TBC. A frame sync (which is differs from a frame TBC) can genlock to an external sync generator.

[JQK18]

Line TBC
Multi-line TBC
Field TBC
Full-field TBC
Frame TBC
Full-frame TBC
Frame sync

I'm confused.

[edDV]

Yup. Some kid should write a term paper.

[JQK18]

Looks like someone already did:

https://www.videohelp.com/forum/viewtopic.php?p=1115672#1115672

[JQK18]

Although I'm curious about what happens when you try to use more than one time-base corrector at a time.

For example, doing video capture using a VCR with a full-field TBC (like the AG-1980), hooked up to an analog-digital conversion device with a line TBC (like the ADVC-300). Does that make the picture better, or do the two interfere with each other?

[BrainStorm69]

Not sure about the answers to some of the questions being asked (I've been away too long), but here is what my AG-1980P manual says about the TBC function (it says nothing about the TBC's specs):

[edDV]

Looks like someone already did:

https://www.videohelp.com/forum/viewtopic.php?p=1115672#1115672

A bit long on problem definition (but not defining Macrovision or MV) and short on defining TBC correction methods.

[JQK18]

Looks like someone already did:

https://www.videohelp.com/forum/viewtopic.php?p=1115672#1115672

A bit long on problem definition (but not defining Macrovision or MV) and short on defining TBC correction methods.

Here's the relevant part:

"Enter the TBC. As the signal enters a full-frame TBC/Frame Synchronizer like the TBC-1000, the video information is digitized and stored in a buffer memory, one field at a time until an entire frame is stored. The timing information is discarded. Then the TBC replaces each and every horizontal line timing with its own. The timing information is determined by the length of the "retrace" area of the signal. This is also where the color information resides. It's the part of the signal you don't see on the screen.

The TBC also takes each completed field and does the same with the vertical "retrace" portion. This is where the vertical timing is stored, and is also where Macrovision pulses appear in the original source material. Line TBCs, like those found in camcorders and VCRs, do nothing to help vertical timing ...only horizontal."

[lordsmurf]

Although I'm curious about what happens when you try to use more than one time-base corrector at a time. For example, doing video capture using a VCR with a full-field TBC (like the AG-1980), hooked up to an analog-digital conversion device with a line TBC (like the ADVC-300). Does that make the picture better, or do the two interfere with each other?

Read this, too: http://www.digitalfaq.com/forum/showthread.php/alternative-avt-8710-1853.html?p=9889#post9889

[JQK18]

I think I understand this now, but please tell me if I am mistaken:

A "frame" is made up of two "fields."

Keep in mind VHS and analog TV video is interlaced. This means the picture is divided into horizontal lines, and the video switches off showing the odd-numbered lines and the even-numbered lines.

So get a complete frame you have to combine the "field" made of the odd lines and the "field" made of the even lines. That must be what the TBC does -- it waits for all the odd-numbered, and then all the even-numbered, lines to come in, and puts them together in its internal memory.

[microflyer]

Hello,
I am also experiencing this exact same issue.

I am planning to transfer some 10-15 year old PAL VHS tapes to DVD. Original capture was on VHS-C camcorder then transfered to VHS tape.
I have a Panasonic NV-HS1000B (which I understand may be similar to the AG-1970 mentioned previously) VCR playing the VHS tape. I am feeding the S-Video output into a Canopus ADVC 110 then Firewire into my PC.

When I view the VCR output (S-Video or Composite) on a CRT TV vertical parts of the image show as wavy. This is corrected by turning on the internal TBC in the VCR. Good so far.

However, when the image is digitised in the Canopus the wavy lines are straightened but the whole image is unstable in the horizontal plane. The image shakes intermittently about 1/4 inch horizontally. As the previous persion mentioned it takes about 3 seconds for the instability to start.

So, it looks like the Canopus cannot process the image correctly being fed from the VCR when the TBC is on but the TV can cope with it. I appreciate TV's are less sensitive.

All the tapes I have tested exhibit the same behaviour including commercial ones. It is the same if I feed S-Video or Composite into the Canopus. It cannot be the capture software in the PC as I have taken both the S-Video and Composite output from the Canopus into the TV and it exhibits the same shaking image.

I read elsewhere:

S-VHS VCR Line TBC: Unlike the standalone TBC, this one will NOT give a continual clean signal out from the VCR
What it will do, however, is clean the visual quality, by: removing geometric distortions from the image, such as the wiggling appearance of older video, as if viewed through a rippling pond or bathtub

Has anyone else observed this phenomenon? Any way to fix it (I guess an external TBC). Thanks.

[pgoelz]

I'm the original poster. I fixed it by turning the TBC off and accepting the slightly wiggly vertical lines. Not the best solution but I suspect the only one available without further purchases.

I suspect the problem is the fact that the TBC has to follow the sync frequency of the incoming video to some extent and if it does this too fast, the ADVC moves the frame sideways. Just a guess.

Paul