Offline Bounce and Automation

[DJ Hellfire]

Originally Posted by ExpressMix

Here's a single string note with XPand using audid mute automation. This time the mute automation rendered.

Ordinary mix renders, green is offline, purple is realtime. The third red one is a realtime recording of the original instrument track back to another audio track through a bus. Also notice - with no gain changes anywhere, that the purple is actually louder than the other two.

Look at the timing differences of the three, they're all in different places.



Here's an up close view of the largest timing differences between the 3 prints, over (if I read this right) 500ms.



Win 7 64, MOTU PCI-424 Drivers

Is Delay Compensation on? I would imagine that would explain why your in-session bounce (red clip) would be that far back. It's recording after the audio has been Delay Compensated. No?

[tj_davies]

Oh by the way. Dear Duc members, Avid has just forgotten to thank you for beta-testing PT11.

Unfortunately they forgot to beta test themselves before releasing PT11.

It is obviously not worth mentioning, that you will not be paid for finding those huge errors, that are just insane.

Timothy

[DetroitT]

Fab Filter Twin is working fine.
Same mod wheel'd midi file with Vacuum not working. 11.0.1
MIDI Mod to db-33 (FX on Audio) cab-speed also not working in offline.

[ExpressMix]

Originally Posted by DJ Hellfire

Is Delay Compensation on? I would imagine that would explain why your in-session bounce (red clip) would be that far back. It's recording after the audio has been Delay Compensated. No?

Yes, it was on.

The only point there is that bounce should never result in a positioning being at a different place than the realtime playback, in any case. That the timing you hear on playback should be what you get on bounce, in any case. That's what you're asking for... "Print what I hear"... timing wise.

The third track is the print of the realtime playback, the timing of what I actually hear when I play that midi note. The potentially problematic thing there is that neither bounce is at the same start position as the realtime playback.

Anyway, I only posted it so AVID could see it. I assume they'll test it for themselves (with non-AVID ASIO cards at multiple different latencies) and see if there is an issue that needs addressing. As you can see, even though I've owned PT since version 7, I only have 9 posts here, so I'm not generally inclined to look for and report issues with PT... I kinda only recently joined the DUC and won't be posting much.

[DetroitT]

UVI workstation (2.1) working here. (the Mod Wheel - offline is working).
Seems to be an Air issue.

[DigiTechSupt]

Originally Posted by ExpressMix

...with no gain changes anywhere, that the purple is actually louder than the other two.
Look at the timing differences of the three, they're all in different places...

Thanks for the detailed posts. Eng is investigating and asked for a copy of your session file. I started a support case so you can reply to that with a zipped copy (if willing).
-jc

[DigiTechSupt]

Originally Posted by DetroitT

UVI workstation (2.1) working here. (the Mod Wheel - offline is working).
Seems to be an Air issue.

Thanks for that info. It does seem to be specific plug-ins so we are continuing to investigate.
-jc

[Zarabozo]

Originally Posted by DJ Hellfire

Is Delay Compensation on? I would imagine that would explain why your in-session bounce (red clip) would be that far back. It's recording after the audio has been Delay Compensated. No?

No, no... Don't confuse such an error with such a concept. ADC is not supposed to record at a different position or be the cause for a missplaced clip, unless we are talking about a big bug on ADC.

The only thing ADC is supposed to do is playback each track with an adjusted delay so each track sounds exactly when its supposed to sound - but at the same time, it's supposed to place any recording where it was supposed to be. If anything, ADC should ensure that your recording is precisely on spot.

What you suggest would make sense only if he was switching ADC on and off between takes - if you record something with latency while ADC is off, then it will be recorded late, not the other way around.

So, if sending something to a bus and recording it while having ADC on gives you a missplaced clip, you are looking at a big bug.

Francisco

[DJ Hellfire]

Originally Posted by ExpressMix

Yes, it was on.

The only point there is that bounce should never result in a positioning being at a different place than the realtime playback, in any case. That the timing you hear on playback should be what you get on bounce, in any case. That's what you're asking for... "Print what I hear"... timing wise.

The third track is the print of the realtime playback, the timing of what I actually hear when I play that midi note. The potentially problematic thing there is that neither bounce is at the same start position as the realtime playback.

Anyway, I only posted it so AVID could see it. I assume they'll test it for themselves (with non-AVID ASIO cards at multiple different latencies) and see if there is an issue that needs addressing. As you can see, even though I've owned PT since version 7, I only have 9 posts here, so I'm not generally inclined to look for and report issues with PT... I kinda only recently joined the DUC and won't be posting much.

Originally Posted by Zarabozo

No, no... Don't confuse such an error with such a concept. ADC is not supposed to record at a different position or be the cause for a missplaced clip, unless we are talking about a big bug on ADC.

The only thing ADC is supposed to do is playback each track with an adjusted delay so each track sounds exactly when its supposed to sound - but at the same time, it's supposed to place any recording where it was supposed to be. If anything, ADC should ensure that your recording is precisely on spot.

What you suggest would make sense only if he was switching ADC on and off between takes - if you record something with latency while ADC is off, then it will be recorded late, not the other way around.

So, if sending something to a bus and recording it while having ADC on gives you a missplaced clip, you are looking at a big bug.

Francisco

Yes, but we are not talking about recording an external input here. The bus you are using to route this audio to an audio track is effectively a Master channel. You are recording the output of that Master channel back into the session. The audio that's being output on that Master channel is not playing back at the same location as the original audio clips. It's being delayed. So if I'm outputting that Master signal to a new audio track, then that new audio is going to also reflect that delay compensation since Pro Tools only records audio where it actually "hears" audio. For Pro Tools to be able to have recorded delay compensated audio line up with the original audio, Pro Tools would have to actually physically move that newly recorded audio on the timeline, which is not the nature of Pro Tools since Pro Tools is supposed to place audio exactly where it sees the audio being input. This is how I've always understood it.

But now that I think about it, was there even any delay compensation being made in this case? Being that there's only an instrument plugin, the latency should be 0. Or did you have plugins on the original track?

[scottgreiner]

Originally Posted by Zarabozo

So, if sending something to a bus and recording it while having ADC on gives you a missplaced clip, you are looking at a big bug.

Francisco

You need to remember that the original poster was bouncing a midi track -> virtual synth, not an audio track. Midi is not the same thing, and behaves differently with respect to ADC. Midi and ADC still need some "ironing-out" in my opinion.