Analog hardware latency

[Peters86]

Hello guys,

Thanks for the additional info. I still can't really figure it out. I have used the Avid signal generator to generate a sine wave on track 1. I then add the i/o plugin on this track. I made a second track on which I selected the input that is receiving the audio through the hardware units, and pressed record. It seems that there is actually audio coming into Pro Tools BEFORE the sine wave. I hope this makes sense. See attachment for a screenshot. The difference is about 13 ms.

[chrismeraz]

That's perfect. Now you know there is a 13 ms delay caused by a round trip through your converters.

However, ProTools doesn't know this yet. So, go to the PT I/O panel, and tell PT there is a 13 ms delay for all inserts. Then go to Preferences and make sure AutoDelay Compensation is ON, and that should sort you out.

[Carl Kolchak]

Originally Posted by chrismeraz

That's perfect. Now you know there is a 13 ms delay caused by a round trip through your converters.

However, ProTools doesn't know this yet. So, go to the PT I/O panel, and tell PT there is a 13 ms delay for all inserts. Then go to Preferences and make sure AutoDelay Compensation is ON, and that should sort you out.

No, he's saying the recording is arriving 13 ms EARLY.

@ OP : you'll get far more accurate readings by measuring in Samples, which you can input exactly into any offset fields.

Rerun the test with all compensation turned off (show the delay at the bottom of the mixer, and right click to disable/enable).

Repeat at different hardware buffer sizes.


Sent from my iPhone using Tapatalk

[john1192]

bus the 1st track to the 2nd track .. DO NOT have the 2nd track Input set the same as the Insert I/O return ..

[Carl Kolchak]

Originally Posted by john1192

bus the 1st track to the 2nd track .. DO NOT have the 2nd track Input set the same as the Insert I/O return ..

Good catch - if that's what he's doing (I interpreted it the other way, but on re reading it, you may be right).


Sent from my iPhone using Tapatalk

[Peters86]

Got it guys, sorry for the delay () in responding. Please look at the attachments. Hope this is right? I disabled delay compensation and the recording through the analog gear seams to arrive back with a delay of 12 ms or 572 samples.

In the I/O window, if I now enter 12 ms as delay time, will PT compensate for the delay when I re-enable delay compensation?

And, is the delay only caused by the roundtrip through the converters? In other words: I could theoretically add a hundred more analog EQ's after the current chain and it would still be 12 ms?

Thanks a lot guys!!

[Peters86]

One last thing. When I enable delay compensation on the track, PT seems to automatically set the delay value to 578, which is very close to my own measurement. Could it be that PT is receiving the amount of delay from my Apogee Ensemble? I'm using PT 12 Native by the way. It also changes the amount of delay correctly when changing buffer size, this matches my measurements as well. I guess that when PT does get this info correctly, I should just leave the values in the I/O panel at 0 and let the ADC do its thing?

[JFreak]

Originally Posted by Peters86

And, is the delay only caused by the roundtrip through the converters? In other words: I could theoretically add a hundred more analog EQ's after the current chain and it would still be 12 ms?

Analog circuitry doesn't cause delays, so yes.

[Emi]

Originally Posted by Peters86

One last thing. When I enable delay compensation on the track, PT seems to automatically set the delay value to 578, which is very close to my own measurement. Could it be that PT is receiving the amount of delay from my Apogee Ensemble? I'm using PT 12 Native by the way. It also changes the amount of delay correctly when changing buffer size, this matches my measurements as well. I guess that when PT does get this info correctly, I should just leave the values in the I/O panel at 0 and let the ADC do its thing?

This is my experience too. I see that pt vanilla does compensate for hardware inserts.

This is how I tested it:
Duplicate track, one for reference. The other one I inserted the hw and than sent it, via a bus, to another track. Armed the track and record. After that i compared the reference track to the recorded one. Both tracks were perfectly in sync.