Axe-Fx III USB Audio Latency Troubleshooting

In Cubase you can also setup the Axe as an External FX in Studio > Audio Connections > External FX and you can have it measure it for you. This is useful for all outboard gear.

 
Last edited:
I'd run the offset test again (using a shot transient sound like a stick-click) after setting the offset amount... to be 100% sure new tracks are aligning properly.

Any time you switch audio interface (to a model you've not used before), I'd recommend checking the offset.
 
I'm very happy that I got it working now however, my mind is starting to play tricks on me.
Sometimes it now feels like my riffs end up a little too early on the timeline, as if I still haven't got the offset right.
It probably is right now but it would be sweet if the drivers did this for us.
Just so you know a 100% sure that when a part is off, it's you, not the gear.

@AlbertA shouldn't we be compensating for possible latency cause by the click going all the way back into the Axe fx after d/a and d/conversion from the 'reamping' of the click?
Because when we're playing guitar we're only going through one A/D, but when we're reamping the click through the loop we're getting another extra D/A.
Perhaps we should add -265 (or whatever number is was) to the result of the offset measuring test?
 
Last edited:
Guitarjon said:
@AlbertA shouldn't we be compensating for possible latency cause by the click going all the way back into the Axe fx after d/a and d/conversion from the 'reamping' of the click?
Because when we're playing guitar we're only going through one A/D, but when we're reamping the click through the loop we're getting another extra D/A.
Perhaps we should add -265 (or whatever number is was) to the result of the offset measuring test?
Click to expand...

When you record the loopback you are measuring all of the latency encountered, both playback and recording - that's why it's the most accurate thing you can do. This includes the DAC and ADC latency because you are using an external loopback cable.

If you change the ASIO settings (including checking/unchecking safe mode) or the USB Buffer size in the Axe-Fx III, you'll have to change the compensation.

Let's assume you have a backtrack in your DAW. The DAW will send sample timestamped 0.

You get to hear this sample after its been through some software latency (ASIO buffers + USB Buffer Size in the Axe-Fx III) and hardware latency (DAC), because you are monitoring through the Axe-Fx III either with headphones or speakers attached to the analog outputs of the Axe-Fx III.

There's some additional latency from sound getting from the speaker to your ears, but will ignore that for now (or always I guess :))

You then react to that sample and begin playing your part. Let's say you react instantaneously. The first sample of your playing has to go through an ADC conversion in the Axe-FX III, then through some software buffering (ASIO buffers + USB Buffer Size in the Axe-Fx III).

The DAW sees that first sample you played with your guitar arrive 1160 (or whatever your measured loopback latency is) samples later amount of samples later than the first sample it sent.
 
Last edited:
AlbertA said:
When you record the loopback you are measuring all of the latency encountered, both playback and recording - that's why it's the most accurate thing you can do. This includes the DAC and ADC latency because you are using an external loopback cable.

If you change the ASIO settings (including checking/unchecking safe mode) or the USB Buffer size in the Axe-Fx III, you'll have to change the compensation.
Click to expand...

Yes I understand, but if you compare the test to 'regular' recording, isn't there one stage of conversion less than during the test, thus making you need to compensate for that one more stage for recording?
Hope this makes sense... :)
 
Guitarjon said:
Yes I understand, but if you compare the test to 'regular' recording, isn't there one stage of conversion less than during the test, thus making you need to compensate for that one more stage for recording?
Hope this makes sense... :)
Click to expand...

I edited the post above while your wrote this :)
 
I was recording a riff today.
It has super fast but steady 16th notes in the high D string.
Each time I got a very satisfying take it sounded just a little too fast when I played back the take.
I know from experience that I usually tend to play slightly laid back but never really too fast so I decided to try 900 and I think that offset is more accurate for me despite the test result being 1160.
Strange...
 
Guitarjon said:
I was recording a riff today.
It has super fast but steady 16th notes in the high D string.
Each time I got a very satisfying take it sounded just a little too fast when I played back the take.
I know from experience that I usually tend to play slightly laid back but never really too fast so I decided to try 900 and I think that offset is more accurate for me despite the test result being 1160.
Strange...
Click to expand...

After setting the offset, re-test (with a short high-transient sound like a stick-click) to ensure that newly recorded audio lines up perfectly with existing audio. If not, you need to readjust the offset in samples. If it does, the timing issues are the performer/player.
 
I am so ignorant about this issue but this sounds like a major problem particularly for those of us without computer technical skills. Is this a similar problem with Pro Tools? I guess I am about to find out...
 
IDK. It seems like asking the Axe Fx III to serve as an audio interface is a bit much. I'd suggest getting a good audio interface. No? I don't know. What do I know? I've never used it as an audio interface. I have plenty of high end ones so I don't need to. But it's a speciality tool.
 
henryrobinett said:
IDK. It seems like asking the Axe Fx III to serve as an audio interface is a bit much. I'd suggest getting a good audio interface. No? I don't know. What do I know? I've never used it as an audio interface. I have plenty of high end ones so I don't need to. But it's a speciality tool.
Click to expand...

The Axe FX is a great interface actually and my old II and XL+ have served me very well.
It's supposed to be a good interface because it was designed to be.
It is probably a bug that needs to be ironed out.
 
Last edited:
I know but why wouldn’t the III be able to do correctly what the 2XL can do ?

It’s indeed a problem with the 3 or its driver that should be solved.
 
OldSnail said:
I know but why wouldn’t the III be able to do correctly what the 2XL can do ?

It’s indeed a problem with the 3 or its driver that should be solved.
Click to expand...

I don't know and I don't want to speculate further.
Perhaps it's best that we wait for an official statement from FAS.
 
FWIW, This offset issue is more of an "annoyance" than a show-stopper.
If using the Axe-FX III as your audio interface, measure the offset in samples... and enter that into your DAW software.

If you use a quality dedicated audio interface, the Axe-FX III has both S/PDIF and AES digital outputs.
 
OldSnail said:
I know but why wouldn’t the III be able to do correctly what the 2XL can do ?

It’s indeed a problem with the 3 or its driver that should be solved.
Click to expand...

It's the same for the Axe-Fx II. Perhaps you never noticed. But the default latency IIRC is greater than in the Axe-Fx II, but can easily be adjusted in the Axe-Fx III (just turn down "Setup->I/O->Audio->USB Buffer Size to say 32).
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom