[James Cohen]

We swapped our Sony XBR monitor with a Panasonic Plasma (that we have been using for years).

The Plasma has a direct HD-SDI input.

The sync looks good using the older professional monitor. So it must have been the latency introduced by the AJA Hi5-3D HDMI converter and the consumer monitor doing all its upscaling algorithms or whatever it does.

[James Cohen]

I talked to Blackmagic support, and they thought our signal flow was introducing some latencies.

They looked up our Sony XBR manual and found a “game” mode that is a low latency mode that bypasses a bunch of the algorithms the consumer monitor does. It went in through the options and turned off any other options like “Reality Creation” or “Smooth Gradation” or the “CineMotion” thing that makes everything look like a soap opera.

In our case, the HD-SDI output, through the HDMI converter carries video and audio. So we can play audio out of the TV speakers on the client monitor and the sync is spot on.

The problem is that the editor is used to hearing the zero latency analog audio at their desktop through Genelecs. If the analog editor speakers are up, and the client monitor TV speakers are up, there is a echo (that represents the latency between the two signals.)

So the Premiere GUI and the editors speakers are still fairly out of sync from the client monitor.

The Preferences / Playback / Video Offset seems like it would help, but doesn’t seem to be working.

[James Cohen]

We have another suite that is identical except that it has a Kona 3G instead of a Blackmagic card.

In that suite the preferences -> playback -> Kona Video offset actually effects the playback.

So my guess is that something is gummed up in the interface between Premiere and the Blackmagic card.

I’m not sure whether Adobe or Blackmagic should address the bug.

[James Cohen]

I’m making the “self contained” using the command-e export in FCP (this is a second export after the previous reference file.)

I see a shift when viewed in QT 10, but I dropped the self contained file back into FCP and you’re right, it is a perfect copy.

I’ll mess around with it a bit more, but I think this does the trick for us!

[James Cohen]

I wasn’t really aware of the stereo link (option-l) command. Exporting a QT ref after stereo link has been applied gets us a version of the file that has a stereo pair. Perfect for the H.264 exports.

If we choose to make the file self-contained instead, we get a stand alone ProRes with a stereo pair… however there is a slight color shift (a little brighter). Maybe there’s no way around that.