I think i must have been the first person on this forum to come across this problem last year and i was pulling my hair out to find a solution. I had just shot a job for an important client of around 30 tapes and on 1 or two tapes i had dozens and dozens of tc breaks and a loss of a few seconds of vision between each break .

ABSOLUTELY MADDENING!

The only work around the editor could find was to import thru imovie HD.

However......... I have tried to replicate this problem and cant. It's never happened since. I have a theory though........

Perhaps, just maybe perhaps, it has something to do with the voltage level of the camera. From memory on both tapes i encountered this problem i recall having a nearly flat battery.....i am running idx v locks and seem to remember changing batts around the time the tape was shot.

Any thoughts boffins?

Craig D

Isn't just as simple as native HDV capture has the problem, but AIC capture doesn't?

That seems to be the case, but AIC isn't CAPTURED.

HDV is captured and converted on the fly to AIC. So the HDV capture software used when one chooses "AIC" operates differently than the HDV capture software used when one chooses "HDV."

Specifically, to encode AIC each HDV frame must be decoded to YUV. If a frame(s) is screwed-up, the software creates the best YUV frame it can. The tape fault is automatically cleaned-up. Think of it as a DOC circuit. The each YUV frame is encoded to an I-frame. (Same thing must be done by CineForm and Canopus Intermediate codec.)

NOTE1: I'll bet HDV camcorder and decks also fix bad frames. Some Hi8 units had great DOC circuits while others didn't.

When you capture as "native HDV" -- when a flawed frame is encountered, FCP has a choice. It could/should "repair" the flawed-frame and put it into the disk file. But, it seems that FCP simply aborts the GOP and then starts again with a new GOP. The 4 second delay is the time needed to close the current file and open a new one for writing.

Over the next week I'll check Liquid and Canopus to see what they do.

NOTE2: I've been copying recorded tapes to a Focus HDD and finding it randomly aborts the transfer.

Apple won't "fix" anything native HDV since they are writing new code for 24p, 25p, 50p, and 60p. Who knows when we'll see this!?!

Frankly, FCE 3.5 seems an ideal way to do HDV capture to AIC.

I agree with Craig Donaldson about the battery power causing the breaks in the timecode. Two weeks ago i filmed a wedding and while capturing i noticed one tape was divided into 66 clips and most of them under 1 second. Not all of the clips though, all the shots at the beggining of the tape were perfect but after 20 minutes the timecode breaks started and at the end of the tape a couple of 10 minute shots were fine... the first shots i was handheld using a fully recharged battery and followed the couple around greeting people, after about 35 minutes or so i switched to the tripod and plugged the camera to the AC adapter. All the shots that were filmed handheld using battery when the battery was around 7.0V or less were splitted.
This was not the first time i noticed these breaks and each time i have noticed missing footage was when i was using battery.
To capture the footage i used iMovie HD after various attemps to capture with FCP.

I did more testing today with our Camera "2". The one I had tested in the studio and had no TC breaks. Later, a producer took this camera out and shot a time lapse...an hour of one shot of a sunset. He had lots of breaks during capture via his G5. So I tried importing the footage on my system. Here's what I found:

Captured via FCPro 5.1, "A" version HD100 HDV 720p30. BR-HD50 VTR , Quality 1 meter Firewire cable.

In about 40 minutes of one continous shot, I had 3 TC breaks and aborts and 1 "stream error" abort.

FCPro is set to stop and alert on TC break.

Each time, I would rewind a few minutes and start again. In each instance of TC break, it would recapture the same TC without error!!!! IOW, the 2nd attempt would work fine. No stop or break, no visual disruption of image.

Now 30 some minutes in there is a "stream error". This time, it's repeatable.
Each time I rewind and it hits this frame, the error happens. No visual disruption, when I just play it back past this break.

So, there you have some more pieces of the puzzle.

Dave Beaty

To capture the footage i used iMovie HD after various attemps to capture with FCP.

Hi!

Melvin, in which format did you capture in iMovie HD? Was it shot in 720p25?

Thanks!
Sergio Barbosa

i'm having sync issues with capturing with imovieHD, does anybody know a solution to the problem? i'm capturing from my HD-100 to a dual 2.3 G5 powermac.

yooch

Hi!

Melvin, in which format did you capture in iMovie HD? Was it shot in 720p25?

Thanks!
Sergio Barbosa

I captured in 29.97fps the footage was recorded at 720p - 29.97 fps

I then used the share function to export it to Quicktime using the HDV 720p30 codec, at the moment im doing a test to see when the timecode breaks come, i filmed 1 hour of footage using battery and with my second camera i filmed the lcd panel to see the battery level. If the timecode breaks when the voltage changes then i'll find out if it's the battery or not. I'll post about it once i have looked through the material.

Well as i said i filmed 1 hour continously using 2 cameras with the second camera filming at the lcd panel of the HD101 to see if the TC breaks occur when the battery level changes. The HD101 footage was shot at 720p30 and i captured using FCP capture now with HDV 720p 30 codec. I had 3 stream errors in this tape but no timecode breaks like before that TC would jump back several minutes, when i looked at the other cameras footage the stream errors were not when battery level dropped. Im using a 1.5 m FW cable (is this too long to cause some lag or something?).

Guess i'll try to capture using AIC next...

have you guys tried capturing on a pc using HDVSplit ?

With the hour long test I did with the footage of a sunset, I was using the Anton Bauer Dionics which power the camera for many hours - so I don't think power was an issue in our case.

Again the odd thing, was the TC error was not reproducible in FCPro. The TC break errors happened on the first capture, but rewinding and recapturing the same scene did not break...

BUT, when a "stream error" occurs, that is reproducible.

Dave Beaty

... the TC error was not reproducible in FCPro. The TC break errors happened on the first capture, but rewinding and recapturing the same scene did not break... BUT, when a "stream error" occurs, that is reproducible.


That tells me: either there is a weakness in the FCP capture software or the TC break was a READ error while the Stream break was a WRITE error. Both Read and Write errors can be caused by a tape/head problem. However, a Write error could also be caused if for some reason a bad bit-stream was sent to the head.

Since it seems it's only FCP folks using Native HDV capture reporting problems, it would seem to be a software problem.

However, if other native NLEs have the problem, my best guess, based upon years of similar problems with Hi8, is that various "types" of tape problems are occurring.

With evaporated (ME) Hi8 tape, it was typically a missing patch that caused a DO. These DOs were "forever." So with ME one avoided repeated playings and NEVER jog/shuttled.

With metal partical (MP), it was typically junk that built-up. If recording, the error would be permanent. With MP one never let the tape stay in Record-Pause very long. If the playing, you could rewind and play again.

I suspect Sony understood this issue which is why they introduced HDV tape as a step above DV tape. After all, it was Sony's release of Super Quality Master ME Hi8 tape that finally made Hi8 useable. It took about three years to develop.

We've been using Panasonic DVM63MQ Tape. We've never had any problems visually with the material dropping out. I'm open to trying anything. Perhaps another tape stock would help. Is anyone using the Sony HDV tape seeing the broken clips?

Dave Beaty