Yikes & HOLY #$@&! - Z1 and timecode breaks!
 

I'm hoping this is peculiar to my Z1 but I doubt it. Therefore this is a huge red flag to anyone who logs their tapes and captures across camera stops. The timecode is not continuous and in Premiere Pro 1.51 it can be off by 9 frames or more (possibly compounded). I've seen it in the DV downconvert (60i) and a Cineform 3.2 AspectHD capture (1080/30i). It happens at points where the camera was turned off. When turned back on the code appears to be picked up by the camera. In fact it's not. There's a hole on the tape which when played back at normal speed shows as the picture freezing for a second. But if you play the tape back a slow speed there is nothing there. The code picks up on the other side of the hole 30 frames later but Premiere pro sees the hole as 39 frames long which is probably its correct temporal length.

Ive also seen a one frame jump back in PPro over some camera pauses using scene detect in DV downconvert mode. Not sure why and I can't confirm that's what's coming out the firewire port because slow play is a little too fast to see a one frame jump. But it would mean a 1 frame timecode inaccuracy over some camera pauses in batch mode.

So this appears to be a bug in the camera firmware in how it reconstitutes code to the ilink port over a broken GOP. I think the only reason it may not have been reported before is that the only time it will come into play is in an offline/online scenario which most people haven't been doing. But assuming that this is not peculiar to my camera it means that at present the only way to capture your material frame accurately in PPro is by avoiding capturing over camera stops. Using scene detect is one (the onle?) way this can be done.

I'm really curious to find out if this is just my camera or not. Can someone else confirm by doing a scene detect doing dv downconvert (I don't think AspectHD yet supports scene detect). Over a camera turn-off you should see a jump back in code of 9 frames. You should also see 1 frame jump backs over some camera pauses.

David

By "turning the camera off", do you mean recording some footage, turning the camera off then recording more footage? If so, I did about 4 hours of that on Friday and Saturday and experienced no problems whatsoever. However, I was shooting in DV SP mode, which could make a difference.

DV mode would make a huge difference because there's no GOP that gets truncated at camere stop and turn off. It's a completely different thing - outputting it from the camera is a tried and true process.

It seems the problem is timecode reproducibility from DV to HDV over hard scene breaks. This sounds like a camera limitation. Aspect HD does handle these scene breaks but it only runs in HD modes, your DV offline needs similar scene break detection. I don't have an answer for DV offlining, which is not a workflow we recommend -- CineForm is trying to promote on-line all the time.

I think it is your camera or you might have drop out or your lithium battery that keep timecode info when camera is off is dead because it is working perfect with me, I'm using Edius NX for HDV and timecode is dead on.

I did run a brief test with the AspectHD 3.2 trial that I had downloaded back then and not installed. I ingested two overlapping clips. The first crossed a spot where I had determined the camera stopped. The secong began two seconds after the stop. I lined them up on a timeline which was set to begin at the start code of the first clip and I dropped the second to begin at its matching code. The frames should have lined up. They were 9 frames off.

So it seems to be a camera issue (perhaps an issue with my camera - that has yet to be determined) that also would affect AspectHD if someone ingested over camera stops and wanted to re-ingest at a better resolution.

David

When the camera is off the Li battery is still supplied by the main battery. If it WAS dead I would lose my settings when changing the main battery but I don't. Have you used edius to ingest across camera stops? Perhaps edius is set up to compensate properly for the dropout. A way to test it is directly on the camera by playing in slowmo over a spot where the cmera was definitely turned off. If you don't see a "blue screen" gap then it may well be my camera.

David

Yes I frame my frame advance over the many spot where I know the camera was turn off completly and there are no timecode mess up or blue or black or whatever, one thing that came to mind is that if you have quick record turn on in the main menu, this might be the problem because the manual said you will not have continous timecode when this option is on, so check that and be sure that it is off.

Koi,

"Frame advance" sounds like you're looking at it in the NLE not the camera where you can only slow play it. If your In PPro the same thing - no blue. On the camera, blue.

David

You can advance it frame by frame with the remote control, you need to read the manual about quick record, I think that is your problem and not anything with the Z1.

David,

Could you tell me how you are setting the timecode on the Z1?

This may or may not apply to your problem or maybe completely unrelated. However, if you manually set the time code, let's say starting with timecode at 1:00:00:00:00, this should start on PRESET. After running tape for just a few seconds, I switch it immediately to REGENERATE. I do this for every tape I put in the Z1. Let's say you finish with tape one. On the second tape I set the timecode at 2:00:00:00, record a few seconds on PRESET and switch it to REGENERATE.

Why am I telling you this? The Z1 will lose timecode if you review your tape even after cuing properly, change batteries or switch it off if it is set on PRESET. The camera does not allow for setting timecode in REGEN, so you have to switch to REGEN soon after you have set timecode to avoid this t/c skipping issue.

Again, this may not apply to you and you maybe already be well aware of this procedure. I hate when people assume you do not know basic stuff and think we are not doing something basic the right way. That's why I say this may not apply to you at all or maybe irrelevant in your case. But, just in case ...

So it seems that it is a problem with Qick Record, then. I did read the manual section on Quick Record before I went out and shot. It states that "the transition from the last recorded scene may not be smooth" It says nothing about discontinuous timecode. When shooting, of course, the code reads continuously. If the manual had mentioned anything about code I wouldn't have ingested 20 hours of material without doing tests. A BOOOOOO for Sony. I'm getting tired of manufacturers that gloss over important information with euphemistic phrases that aren't explicit.

I agree with you on this. In any case, it is good to know that people are still going on with the good old habits: offlining/onlining. Good editing techniques never die.

"So it seems that it is a problem with Qick Record, then. I did read the manual section on Quick Record before I went out and shot. It states that "the transition from the last recorded scene may not be smooth" It says nothing about discontinuous timecode."

Yeah you are right, it did not say discontinous timecode, It was a long time ago that I read the manual but that I entepret "the transition from the last recorded scene may not be smooth" as discontious timecode in my mind. But let us know if you are still having the problem after you turn it off.

Timecode breaks
 

Unfortunately, I too have experienced this while in HDV mode... FCP HATES me when I try to capture... seems to happen when the record heads dislodge from the tape, such as after 10 minutes of "downtime" or a power down... odesn't seem to happen all the time but definately more often than not...

Not that it should make a difference but I should state that this is material shot in HDV and down converted in camera to 16:9 ana.