USB Firmware V 1.14 Release

The point is: misalignment of something in the neighborhood of 75 samples is both typical and pretty damn good.
Yeah agreed - you don't necessarily have to do any extra adjustments because the amount is very small now.


Sure, but nobody re-amps with an empty preset :).
But that's irrelevant in my opinion - let me elaborate.

A basic recording scenario for me is, say a backing track on Track A; I want to capture my performance against Track A on Track B, so that when I play all of the tracks together afterward, the alignment of Track B relative to Track A is the same as it was when I actually captured the performance.

Doing this with the Axe-Fx III USB interface, I monitor my performance through the Axe-Fx III itself.
  • DAW outputs Track A, which is subjected to the following latencies - ASIO Output buffer (A), Axe-Fx III USB input buffering (B), Axe-Fx III DAC (C), Monitoring system latency (D) (I use active monitors with DSP) and Time-of-flight (from speaker to ear) (E)
    • This is just straight playback so no Preset Latency.
  • I hear Track A, I react to it and play through a preset - I hear the output after the following latencies: the A/D (F), Preset latency (G), Monitoring system latency (D) and Time-of-Flight (E).
  • The USB capture of the performance however, is subjected to different latencies: A/D (F), Preset latency (G), Axe-Fx III USB output buffer (H), and ASIO input buffer (I)
So the data for Track B arrives at the DAW after A+B+C+D+E+F+G+H+I.

When the performer hears the start of Track A, that's "time 0" for the performer - there could have been a delay relative to when the DAW thinks it started playing, but the performer does not "hear that". The performer hears their own performance after F+G+D+E - in other words, as I capture my performance, I'm hearing Track A + my performance delayed by F+G+D+E, relative to track A.

So then when the DAW plays both Track A and B together, what I expect to hear is what I heard during capture, mainly Track B should be delayed by F+G+D+E from Track A.

So then, what we want is to turn the current delay of Track B at the DAW (A+B+C+D+E+F+G+H+I) into F+G+D+E - that means the DAW should compensate for A+B+C+H+I, which are the ASIO buffers, the Axe-FX III USB Buffers (the 32 samples we have mentioned) and the Axe-FX III DAC latency.

As you can see, the latency due to the preset/patch itself is not involved here in getting an accurate DAW capture of what you hear during the performance, in this scenario.
 
Last edited:
I get what you're saying and I agree that the empty preset RTL is the amount of latency you can't compensate for in your head while playing. However, there is no opportunity to perform that psychological compensation when re-amping. At that point, the block latency becomes relevant and the empty preset RTL doesn't tell you how much misalignment you'll incur.
 
Not sure if this is related to new firmware or new USB driver but since the update, my expression pedals are losing communication. It's an FCB1010 and all the buttons work and the CC number in the Axe-FX is set to the correct values but I have to manually "relearn" each pedal every time I start-up the Axe-FX. I've never had this issue before.
 
Not sure if this is related to new firmware or new USB driver but since the update, my expression pedals are losing communication. It's an FCB1010 and all the buttons work and the CC number in the Axe-FX is set to the correct values but I have to manually "relearn" each pedal every time I start-up the Axe-FX. I've never had this issue before.
Doubtful that is related to the USB driver.

Are other settings from the setup menu persisting across reboots?

It could be a battery issue...
 
Not sure if this is related to new firmware or new USB driver but since the update, my expression pedals are losing communication. It's an FCB1010 and all the buttons work and the CC number in the Axe-FX is set to the correct values but I have to manually "relearn" each pedal every time I start-up the Axe-FX. I've never had this issue before.

I use also a FCB1010 (with uno chip) and have no problem with the new firmware, pedals work perfect!
 
Just to throw more gasoline at the fire. I remember the old days when I recorded bass. When using a DI and an amp together I often had to flip the polarity, delay the DI track or offset the the amp track so that both tracks play in phase. The reason for this is that the di track reaches the console faster than the amped track. This is normal behavior, engineers still use this technique in live settings.

The same goes with everything that run thru cables at certain distances. It's of course more audible at lower frequencys and the fact that the two bass signals get blended together. But the delay is still there. This is why the saws today got the sample offset. To adjust for latency. (The time it takes for the signal to be processed and reach its destination.

With today's technology this is rarely noticable, because the signal travels fast. But it's physically impossible to get it to null completely. As it is with chain reactions in general. 32 samples is not much, and it's easy to adjust this to compensate.

Edit: 75 samples is not that much either
 
Last edited:
I do an awful lot of work for film. I deal in Frames/second... sometimes Feet/second. Anyway... unit of measure is irrelevant for the purpose of what I'm about to say.

Now that the latency situation has been brought into the realm of the real world... and even before it wasn't...

I ask you to visualize and old black and white film reel. It counts down... 4... 3... 2...POP... 1...

I'm simplifying...

But that... POP. (sound)... is there to ensure picture and sound are in sync... To the FRAME! So... that pop is essentially just a square wave. Starts at 0 (dB)... spikes immediately (85 dB)... for ONE frame... then disappears again... 0 dB. A square wave for 1/24 of a second (film). Really, really loosely speaking here. BUT...

If you want your re-amps COMPLETEY phase-aligned... and who doesn't? Record your dry track... with much better latency, thanks to the recent improvements. Then... put a POP in front of the track. Start your playback for the re-amp track prior to the POP. Record your re-amp. Align your POPs. Time-aligned... guaranteed! No Bruce Lee mouth on your guitars. Although... Bruce Lee mouth was a wonderfully spectacular part of my youth!

It's late where I am. I can probably explain better tomorrow... include a couple screen grabs.

Night
 
I do an awful lot of work for film. I deal in Frames/second... sometimes Feet/second. Anyway... unit of measure is irrelevant for the purpose of what I'm about to say.

Now that the latency situation has been brought into the realm of the real world... and even before it wasn't...

I ask you to visualize and old black and white film reel. It counts down... 4... 3... 2...POP... 1...

I'm simplifying...

But that... POP. (sound)... is there to ensure picture and sound are in sync... To the FRAME! So... that pop is essentially just a square wave. Starts at 0 (dB)... spikes immediately (85 dB)... for ONE frame... then disappears again... 0 dB. A square wave for 1/24 of a second (film). Really, really loosely speaking here. BUT...

If you want your re-amps COMPLETEY phase-aligned... and who doesn't? Record your dry track... with much better latency, thanks to the recent improvements. Then... put a POP in front of the track. Start your playback for the re-amp track prior to the POP. Record your re-amp. Align your POPs. Time-aligned... guaranteed! No Bruce Lee mouth on your guitars. Although... Bruce Lee mouth was a wonderfully spectacular part of my youth!

It's late where I am. I can probably explain better tomorrow... include a couple screen grabs.

Night
Yes, this is a good way to ensure that tracks end up aligned in time. But it's still manual. The sample offset makes this problem obsolete. (Unless you are switching between several different audio interfaces that all got different amounts of sample offsets. I use to measure all my gear and put notes on them. This is the reason why, and i rarely got anny issues with this. Even when reamping.
 
My clients would never expect anything but.

The machines are mine. I'll drive.
I understand! And of course that is the most reliable way of doing it. But when there is a single piece of gear involved, and there is a way to just plot in a number and then make the manual work less, I tend to do just that. Not that blindly trust all technology, but I think of it this way. I could do one job one time, or I could not do the job and have the job all the time. To me it's about spending my time where I think it's worth using it.

When it comes to sound and picture there are no way of doing this reliable and automatic, then I would most certainly do it manually and "drive my self"

But when it comes to just music, the small differences we are talking about are not audible in that sort. In example if you delay a signal with 4 ms, it is nearly impossible to distinguish between 0 and 4 Ms. But if there are 7 Ms or more, then it's noticable.

But I tend to use the automatic solution. I just want to play guitar.
 
So when 75 samples at 48, this means how many ms?

Just following the discussion as in future i wanna do some recording again.

Of course i wanna try-out also the 1 2 3 pop 😉
 
Samples divided by the sample rate = milliseconds

Edit: so 75 samples devided by 48.1 = 1,5ish ms
 
Last edited:
Hesitant to share this because I am not that bright, but I felt it might be useful to know. Maybe? I'll say, "Hopefully."

I have had my Turbo since last October or so, and I've actually never had an issue with the driver using Reaper and Win10. Not once!
I installed this USB driver and within minutes had problems in Windows 10. Playback from Firefox, Foobar, and Reaper all had dropouts. Most often it would be the audio dropping completely, while the Axe audio was unaffected. I did have a scare regarding my headphones when only the left channel dropped out, which I would estimate happens about 1 of 4 instances of drop-out, but always just the left channel and never the right. Duration seems random, and the audio returns sometimes on its own, and sometimes on switching presets/scenes, which could be entirely coincidental.

Since I need to pipe audio from my laptop to a PA for an audition on Sunday (yay!) I reverted to 1.11 and the issue is gone. Which works perfect for me because I've never had any USB issues until trying 1.14. And for that I apologize, hate to be the bearer of bad news but if it helps troubleshoot I figure I owed it to you all to share.

Gotta get back and try out FW20.04B1, I can't wait to play with the AT drive and the Aurora!

If anyone has tech or setup questions about the issue I'll try to answer them to the best of my limited understanding.
 
I noticed some latency with this firmware, just MIDI latency, audio latency is perfect for me. I noticed that with a Mark 2 and a Turbo. I switch my presets with PCs from 2 different DAWs and I have the same issue. Not a big deal but just wanted to report that.
 
Hesitant to share this because I am not that bright, but I felt it might be useful to know. Maybe? I'll say, "Hopefully."

I have had my Turbo since last October or so, and I've actually never had an issue with the driver using Reaper and Win10. Not once!
I installed this USB driver and within minutes had problems in Windows 10. Playback from Firefox, Foobar, and Reaper all had dropouts. Most often it would be the audio dropping completely, while the Axe audio was unaffected. I did have a scare regarding my headphones when only the left channel dropped out, which I would estimate happens about 1 of 4 instances of drop-out, but always just the left channel and never the right. Duration seems random, and the audio returns sometimes on its own, and sometimes on switching presets/scenes, which could be entirely coincidental.

Since I need to pipe audio from my laptop to a PA for an audition on Sunday (yay!) I reverted to 1.11 and the issue is gone. Which works perfect for me because I've never had any USB issues until trying 1.14. And for that I apologize, hate to be the bearer of bad news but if it helps troubleshoot I figure I owed it to you all to share.

Gotta get back and try out FW20.04B1, I can't wait to play with the AT drive and the Aurora!

If anyone has tech or setup questions about the issue I'll try to answer them to the best of my limited understanding.
I had nearly the same experience with my mark 2 Axe III. Rolled back to previous driver and all is well.
 
Back
Top Bottom