Horizontal Shift during long playback

[Jesse]

[Frank Nolan] “Wow! That is the first time I have heard of a facility producing multiple programs for broadcast that doesn’t have a house sync generator. What the hell are they thinking? Or not?
Horizontal shifts can also be attributed to incorrect timing callibration of the component video signals so if their engineers aren’t smart enough to have a blackburst generator in the faclilty they may not be smart enought to check the callibration of their machines. Tektronics website has a great downloadable booklet called NTSC video Measurements, that would be worthwhile taking a look at and maybe passing it on to their engineers.”

I think the way it works is that this show is autonomous although located in the facility. Depending on the day, a different show will be taping on the stage, but its not like there is coordination between the various shows housed within. I am still a little unclear on the relationship between the facility and this show. There may very well be House sync available somewhere in which case these guiys don’t need to buy the BBgen. This show I’m dealing with would’ve brought their own people to configure the suite so its more likely a reflection on whoever that was than the larger facility.
Thanks for the tip on the booklet.
-JG

[Kerry Soloway]

Hi Bob et al,

I have inherited this gig and problem from Jesse. We were at the facility today together, so that he could transfer things to me and I can tell you the following:

There is house black coming into the edit room. It goes first to the BVW-75, loops to the AJA Io, from the Io, to the WFM and from there to the vectorscope. The Io control panel was set to genlock not free-run.

At AJA’s recommendation, we moved the ATTO card from slot 3 to slot 4. Once again we ran a test, and there was a hit 20 minutes in. This is the most maddening part of it. It can take so long before this manifests that there are only so many tests one can run in a day, especially while production is taking place.

One thing that we did notice, with our two sets of eyes, is that the shift can be seen on the computer monitor as well as the video monitor. Sometimes the shift will manifest as an entire frame moving to the right and then back. Sometimes it will manifest as a half-frame shift and back. Sometimes one-half frame followed by the other half and then back. Doing a video edit to cover the shift doesn’t seem to be a problem.

I have posted a short (5 frame) example on my server at: https://207.234.170.33/clips/IoShift.mov.

Note that this doesn’t happen during playback while editing, only during outputting.

The fact that we see this on the computer monitor as well, would that rule out an AJA problem?

It will be another week before I am back there, but if there are any other ideas, I’m going to try to get in early enough to try some of them out. By the way, I’ve worked on a lot of FCP/AJA systems and have never seen anything like this.

Kerry

[Jesse Gordon] “The shift is present on the tape.
INSERT EDITS are clean.
Let me check the AJA IO control panel settings. Maybe it is not set to FREE RUN.
thx,
Jesse

[Bob Zelin] “i Lee – I respectfully disagree. If this gentlemen is doing insert edits, and he did not black his tape in 4 field to record the color framing info, he has the potential AT INSERT EDIT POINTS to get a color frame h shift. But it appears (and I could be wrong) that he is getting random shifting from simply playing back his material. Now certainly, he should not have this issue (even with no black generator) if he has selected FREE RUN in the AJA IO control panel, but becuase he is having this issue, we must eliminate as many possibilities as possible. “

[Bob Zelin]

you state –
One thing that we did notice, is that the shift can be seen on the computer monitor as well as the video monitor. Sometimes the shift will manifest as an entire frame moving to the right and then back. Sometimes it will manifest as a half-frame shift and back. Sometimes one-half frame followed by the other half and then back.

REPLY – well, as you observed yourself the very fact that you see it on the computer monitor, means that it’s not happening in the AJA product. If I wanted to make up a story, I would say “see – it’s your VTR – you recorded it this way” – but I recall you saying that you could back up the timeline, and you don’t see the shift – and you only see it when you edit to tape, and not during playback.

So what is the answer – I DONT KNOW – if you see it in the computer, it ain’t an AJA issue, but if you dont’ see it during normal playback on the computer screen – well, I don’t know. Normally I would say “reinstall everything for this crazy problem” – but it’s hard for me to say this, when you say that you don’t see any shifting during normal playback – only during output. If you didn’t see the problem on your computer monitor, and only on output , I would still blame your house reference, your VTR, etc. but this is not the case.

I don’t know. You don’t see this type of issue with other FCP systems. I almost think that something else is happening (like it IS shifting during playback, but you dont’ sit there during normal playback to watch for a shift).

I will keep thinking –

Bob Zelin

[Jesse]

[Bob Zelin] ” if you see it in the computer, it ain’t an AJA issue, but if you dont’ see it during normal playback on the computer screen – well, I don’t know. Normally I would say “reinstall everything for this crazy problem””

I don’t think we’ve ruled out the possibility that this problem occurs on “normal playback”. It is unlikely that we would playback from the timeline for 20 + minutes. If it did happen, It might’ve been missed if the editor was focusing on making an edit, etc.

– jesse