Logic Pro: Recording Delay Setting

Axe-Fx III / Logic Pro:

Are you using Recording Delay in Prefs: Audio? I get the best timing with a setting of -128 samples, but I arrived at this by ear rather than precise measurement.

Please discuss.
I've always done it by recording 2 tracks.
Observe the misaligned tracks.
Zoom in a lot to find the transients at the beginning of the two tracks.
Change from bars/measures to samples in the transport
place the cursor on the transient of each track and subtract in samples.
Sample accurate old school manual method.

Im dead on at -307 samples, but every machine will be different
 
I've always done it by recording 2 tracks.
Observe the misaligned tracks.
Zoom in a lot to find the transients at the beginning of the two tracks.
Change from bars/measures to samples in the transport
place the cursor on the transient of each track and subtract in samples.
Sample accurate old school manual method.
Im dead on at -307 samples, but every machine will be different

This is what I did, except instead of playing twice (subject to human error) I used synchronized metronomes -- one live and one on playback -- panned hard left and hard right, then used a precision delay to align them dead center. I recorded the result and then applied the technique you describe. There's got to be a better way.

(Coincidentally, I'm answering two support tickets about this right now, one customer and one endorsing artist. I've asked Apple for their advice too).
 
This is what I did, except instead of playing twice (subject to human error) I used synchronized metronomes -- one live and one on playback -- panned hard left and hard right, then used a precision delay to align them dead center. I recorded the result and then applied the technique you describe. There's got to be a better way.

(Coincidentally, I'm answering two support tickets about this right now, one customer and one endorsing artist. I've asked Apple for their advice too).
2 sharp clicks, and math is all it took for me. Only have to do it once.
I don't need an easier method, but by now there's probably an app for that.
 
This is what I did, except instead of playing twice (subject to human error) I used synchronized metronomes -- one live and one on playback -- panned hard left and hard right, then used a precision delay to align them dead center. I recorded the result and then applied the technique you describe. There's got to be a better way.

(Coincidentally, I'm answering two support tickets about this right now, one customer and one endorsing artist. I've asked Apple for their advice too).
It's not the live to recorded ratio, as in lag, its the recording offset of the tracks actually printing. Using a live monitored source isn't what the daw actually hears, only what you do. After recording two tracks, you'll see the difference
 
I haven’t run into any issues and I’ve done nothing as far as setting the recording delay. The other night I recorded about 10 different tracks, replacing tracks I recorded with an actual amp, I had zero issues with any kind of lag or latency before or after recording.
 
Interesting... I only used that feature when reamping, but might check it out for tracking as well. Please, Matt, keep us informed with your findings.
 
I’m glad M@ posted this. I have been having fits with recorded latency in 2020. I tried to use the in/out utility but the issue is that that is sending a ping out of Logic, through the axe and back into Logic, which is maybe great for setting latency experienced when using re-amp tracks to record new guitar tones. However, it doesn’t account for the true delay time experienced when playing a guitar into the axe and printing that track to Logic.

I’m having to set -1335ms of latency and it is killing the sync between my wet axe stereo track and my dry Reamp track.

basically now, in 2020, I have to record what I can, and then move it into alignment with the pocket after the fact. And, if I do try to use a reamp track to re-record a part using a new amp tone, forget about it - That audio is waaay out of time.

it’s like the axe fx is taking a longer time to process the guitar signal and send it to Logic, then ever before.

ps - using the latest beta #5 (4) or whatever we’re calling it.
iMac 2018 Catalina 32gb ram
 
I set up a PreSonus Quantum and the AxeIII as an Aggregate device. I use the Axe-FX III as the sound card and make sure the Quantum is being clocked by the Axe-FX III. It doesn't matter what I/O Buffer Size I choose using this method (usually 64 Samples but it could be, say, 512 because I don't monitor the Axe-FX III via Logic and I don't monitor my voiceover vocals which are being shown the Quantum, due to feedback issues), the two interfaces will line up perfectly as long as the Axe-FX III is doing the clocking. If left to the Quantum, I have to realign after the recording regardless of the Sample rate.
 
I set up a PreSonus Quantum and the AxeIII as an Aggregate device. I use the Axe-FX III as the sound card and make sure the Quantum is being clocked by the Axe-FX III. It doesn't matter what I/O Buffer Size I choose using this method (usually 64 Samples but it could be, say, 512 because I don't monitor the Axe-FX III via Logic and I don't monitor my voiceover vocals which are being shown the Quantum, due to feedback issues), the two interfaces will line up perfectly as long as the Axe-FX III is doing the clocking. If left to the Quantum, I have to realign after the recording regardless of the Sample rate.

Does this clock rate thing solution work with Axe III being the clock, and my computer running Logic Pro reading the Axe III's clock?

Help me Tom Cruise, Help me Oprah!!!!
 
Does this clock rate thing solution work with Axe III being the clock, and my computer running Logic Pro reading the Axe III's clock?

Help me Tom Cruise, Help me Oprah!!!!

I am certainly no expert in this field and my iMac is a 2011 machine running 10.13, but it works for me with the 'clock rate thing', yes. A video popped up just this morning that kind of addresses this subject.

See:

 
Axe-Fx III / Logic Pro:

Are you using Recording Delay in Prefs: Audio? I get the best timing with a setting of -128 samples, but I arrived at this by ear rather than precise measurement.

Please discuss.

I use -220 in Logic. The correct value for you will depend on what is happening in the AxeFX though, most notably your USB buffer size.

FWIW, it might be a good idea if Fractal supplied a driver for the AxeFX so we wouldn't have to bother with this. I completely understand the tradeoffs with class compliance, but this adjustment can be a headache.
 
Last edited:
I think so too.

I have a call with a Logic team engineer at Apple. I'll get to the bottom of it and let you know.

I would ask if OSX USB 2 Audio driver supports TE_LATENCY_CONTROL queries and the OSX audio stack reports uses that and reports them to applications - then the missing piece would be support in the Axe-Fx III USB firmware for TE_LATENCY_CONTROL descriptor (bmControls should be 1) - the reported number should include DAC/ADC latency and the latency incurred internally which depends on "USB Buffer size".

Then if all those pieces are in place, there would be seamless latency compensation adjustments in the DAW - One can dream :)

For those that want to measure their current latency, you can do a loopback measurement roughly following this post:
https://forum.fractalaudio.com/thre...-latency-issues-with-usb.137958/#post-1637251
 
Should the AXE III USB buffer size internally, match the buffer size setting I'm running in LOGIC, or is there some other strategy where they are mismatched, to reduce latency, and if so, please elaborate?

Forgive me but I've had the Axe III for 18 months, and never had the latency issues I've been having for the past 3 weeks. It has just magically worked all this time for me. So, out of necessity, I'm having to delve into the menus and implement any tips you more knowledgeable people can provide.

Thank you!
 
Back
Top Bottom