Digital Beta timecode issues

Came across a problem yesterday where the digital betacam deck (DVW-A500) was seemingly unable to record a proper vitc track, but after a few frustrating hours, it mysteriously decided to work properly. Not sure exactly what happened to fix it. I still can't pin down the cause of it or even the fix for it, so I thought i'll throw this out here to see if anyone has any suggestions.

Basically, the deck couldn't edit to tape. It will scroll to the timecode position, then the timecode will jump by a few frames in either direction and freeze, and after a while, FCP will say that it's unable to lock deck to servo.

The setup is a Mac Pro (I think an '08 8-core), on FCP 6.0.6, Leopard 10.5.8, AJA LHi. The signal is routed through a JVC HD monitor (not sure if this is synced up with the in-house blackburst). I was working with DvcproHD 720p25, the AJA control panel is set to down convert in hardware. in FCP, video playback is set to AJA Kona 720p50 8 bit, control protocol is set to Sony RS422. On the deck, timecode is set to internal, preset, TC generator set to rec run, vitc on. Pretty much everything seems fine in FCP and on the deck.

The deck receives the signal just fine, and I am able to preview the picture off a composite feed from the deck, as well as off the HD monitor. I am also able to black the tape from FCP.

I went through these troubleshooting steps a few times (including the dinner):

- restarted FCP, restarted the deck, restarted the mac
- trashed FCP preferences
- Increased pre-roll to 5 seconds
- re-blacked the tape from the deck, switched tapes
- tried blacking a portion and doing an insert edit instead of an assemble
- exported a short clip as DV and switched AJA and FCP settings to SD (to eliminate down conversion issues)
- switched ref to free run to try eliminate in-house sync issues
- cleaned the tape head
- went out for dinner

Trying to force the deck to record by switching the deck TC reader from auto to LTC on edit to tape, will cause the deck to go to a position near the in point, and slowly progress frame by frame, and eventually it will give up and say it can't locate the timecode. Switching to VITC doesn't produce any tape movement on the deck, and FCP will just give up after a while. Toggling between VITC and LTC on the blacked portion of the tape produces random timecode readings on the vitc track. Brought the tape to another deck to check, and it read the recorded vitc track as a bunch of zeros.

I've tried routing the signal through a DvcproHD deck into the Digital beta and sending timecode,
and the resulting recording produces the same erratic VITC readings.

Then suddenly it worked.



www.strypesinpost.com

Could have been a tape load error / misalignment on the deck, Strypes, assuming it wasn't just that the tape was damaged and/or weak timecode signal ... did you try toggling the Time Source in your RS422 device control preset?

This would be a slight variation on the EPC (Engineer Presence Conundrum) whereby a problem fixes itself or goes away due to the presence in the room of someone who could fix it but hasn't had a chance to even look at it yet. A corollary to the VQP (Vocalized Query Principle), the answer that becomes obvious within moments of stating a question you have been working on for hours.

ak
Sleeplings, AWAKE!

Quote

a problem fixes itself or goes away due to the presence in the room of someone who could fix it

Among my circle of friends, we call those "heisenbugs."

I call it "voodoo". Same idea!


www.derekmok.com

I came up with EPC after working at a place where getting someone from Engineering to come up and look at a deck required a work requisition form. The importance of the issue was inversely proportional to the length of time it would take for someone to show up, multiplied by a factor derived from the length of time you waited to submit the request and tried to figure it out yourself. Sounds terribly bureaucratic but I learned huge amounts from watching those people take apart a deck and get in behind the racks and patch bays. Anything to avoid the wait next time.

Heisenbugs.... awesome, I am stealing that one. Somewhere a cat is dead/not dead.

ak
Sleeplings, AWAKE!

I wish I could remember where I first picked up the term "heisenbug," so I could properly attribute it. It was probably from one of those great old broadcast engineers; you know the kind, the guys who have thick white beards and glasses caked with drywall dust and who seem to have Tardis pockets.

In any case, it properly means "any malfunction which disappears whenever you're looking for it."

>did you try toggling the Time Source in your RS422 device control preset?

Yea, we did. Same result as toggling it on the deck. The only thing we didn't try was "timer", whatever that means. CTL?

Wasn't the tape, definitely. I swapped 3 tapes, just to be triple sure.

My guesses so far, are- (a) cleaning the tape head; (b) re-power cycling the monitor; (c) toggling the vitc switch in the timecode generator on the deck; (d) toggling the pb switch; (e) having dinner. Although, none of these, in practice seemed to have worked when we were trying them out.



www.strypesinpost.com