Breaks in "continuous" HDV capture

PeterWright wrote on 11/9/2008, 3:05 AM
I recently persuaded a friend to move to Vegas Movie Studio from Pinnacle, and overall he is extremely happy and impressed.

He recently captured an hour performance video shot on HDV, with scene detection switched OFF, but as I remember experiencing, Vegas still split the capture into 2 or 3 clips, with audio breaks where the split occurred. (One was right on the punch line of a joke)

The irony is that when he followed my advice and recaptured from just before to just after each split, he was able to get continuous footage and use this to "splint" the gap.

This shows that there was nothing inherent in the footage to cause a split - just as it was when I was using HDV over a year ago. (Not a FAT issue - NTFS throughout)

There must be something in Vegas causing this - and I am really surprised that this known "feature" hasn't been eliminated. Are others still getting similar experiences?

Comments

farss wrote on 11/9/2008, 3:34 AM
"Are others still getting similar experiences?"

In short yes.
However I've never had a break occur in the middle of a recording. It does seem to occur when the camera has been powered down, the tape lost tension or was removed and therefore there was a break in the time code. From that I guess what happens makes some sense as Vegas doesn't really read the time code from tape, if it did read TC correctly we'd have even more issues being complained about.

It's possible in the instance you've seen there was a simple read error during capture. Possibly a bit of dirt in the heads or some such. Perhaps one reason I and a few others seem to have less trouble in this area than others is we're using good VCRs to capture. I recently captured 3 HDV tapes that my wife shot as B roll footage of a concert on our HC5. Not a single break even though she'd buttoned off and on a few times during each tape. Another job we captured 28 hours of HDV from a mixed bag of cameras and tapes. Vegas averaged around 1 break per tape, that was obviously where there was a significant time gap in when the footage was shot.

Bob.
nolonemo wrote on 11/9/2008, 8:00 AM
I've done some 55 min takes of concerts. On one I had to capture twice before I could capture without a break. That told me the tape was fine and there had been a hiccup of some kind during capture. On another I captured three times, and got a break in the exact spot each time. That told me I had a dropout on the tape and there was nothing I could do about it (fortunately it happened between numbers).

So if you have a break in a bad place, always try recapturing first - from what the OP said, I bet he can get the whole thing without a break.
Mahesh wrote on 11/10/2008, 1:54 AM
I use HC3 for capture. Continuous running for 60 mins. Always get atleast 3 random breaks.
John_Cline wrote on 11/10/2008, 4:02 AM
I rarely capture HDV from tape, I do, however, capture hours and hours of HDV live out of the camera via Firewire to my laptop using Vegas Vidcap. Most of the time, I'm capturing at least a couple of hours continuously. I have never had a single hiccup of any kind. I'm can only conclude that the breaks people are experiencing are due to tape dropouts.

What do we want Vegas to do when it encounters a dropout? Do we want it to just continue capturing and have glitches which may go unnoticed until much later in the edit process or do we want it to start a new file, which is what it currently does?
PeterWright wrote on 11/10/2008, 4:18 AM
The point from my original post was that there was no dropout - going back and recapturing from just before to just after a break produced continuous footage which could then be used to bridge the gap - so it appears that there needn't have been a break first time round.
John_Cline wrote on 11/10/2008, 5:03 AM
This is certainly a curious thing. In theory, there should be no difference in the Firewire stream coming from the camera whether it's from tape or live. Breaks in the capture seem to always be from tape.

It's conceivable that a certain "weak" spot on the tape is right at the threshold of the error correction capabilities of the camcorder. This could be due to some "schmutz" on the tape which gets scrubbed off and upon subsequent playback, it falls just under the threshold at which the error correction will fail and it plays and captures "fine." The error correction capabilities of HDV recording and playback is quite a bit more robust than DV, but that only means that it can correct "larger" errors from tape. I've also had a couple of tapes that won't capture a specific spot no matter how many times I try.

I don't have the answer, all I know is that I have occasionally experienced exactly what you have described when capturing from tape and each time I've captured live video from the camera, I haven't had a single problem, ever. Like I said, what comes out of the Firewire port on the camera should be the same whether it's live or coming off tape (assuming that the tape is free of dropouts too large to be corrected.) It shouldn't be a timing issue either, since, either way, the data stream is being clocked by a very stable crystal reference.
PeterWright wrote on 11/10/2008, 5:24 AM
Yes, it's certainly mysterious.
The other thing of course is - do all capture programs behave the same in this respect?
daryl wrote on 11/10/2008, 5:56 AM
I get a dropout like described using HDVSplit on lengthy captures (an hour or more). I use Canon's best HDV tapes, always new.