Using FM3 in loop of external device and phase issues

D

dschaaf

Experienced
Hi all,

My Axe FX III is my primary workhorse unit and go to but I had recently picked up a quad cortex and an FM3. I wanted to try out the FM3 in the loop of the quad cortex (to essentially provide all post amp effects) and have noticed even with all blocks disabled in the preset, it sounds as though there is a phaser on. I have the quad cortex set up so that the FX loop is assigned to a foot switch and so I can quickly go back-and-forth between the sound without the FM3 in the loop and with it, and as mentioned, when it is engaged, it almost sounds like there is a phaser present. Has anyone experienced this and should I be potentially changing the phase in the settings of the FM3?

Thanks very much,
Derek
 
dschaaf said:
Hi all,

My Axe FX III is my primary workhorse unit and go to but I had recently picked up a quad cortex and an FM3. I wanted to try out the FM3 in the loop of the quad cortex (to essentially provide all post amp effects) and have noticed even with all blocks disabled in the preset, it sounds as though there is a phaser on. I have the quad cortex set up so that the FX loop is assigned to a foot switch and so I can quickly go back-and-forth between the sound without the FM3 in the loop and with it, and as mentioned, when it is engaged, it almost sounds like there is a phaser present. Has anyone experienced this and should I be potentially changing the phase in the settings of the FM3?

Thanks very much,
Derek
Click to expand...
It sounds like there may be a direct path around the loop in your QC. That path will be fast and direct, while the loop involves analog-digital conversions that add latency. That delayed signal, when mixed with the direct signal, will cause phase cancellations
 
Last edited:
Rex said:
It sounds like there may be a direct path around the loop in your QC. That path will be fast and direct, while the loop involves analog-digital conversions that at add latency. That delayed signal, when mixed with the direct signal, will cause phase cancellations
Click to expand...
That is also the case with the HX Stomp.
 
I have HXStomp in my Ax3 loop for post fx and it works well without phase issues as long as the HX loop is routed in series after the Axfx amp/cab model ie:

... Ax Amp+Cab Model > AxLoopOut > HXIn > HX Post FX model (or HX bypass signal) > HX Out > AxLoopRtn > AxOutput.

The same routing logic should work in the context of FM3 in QC loop (make sure only QC is going out to monitoring and FM3 loop occurs serially after QC amp/Cab).
 
Rex said:
It sounds like there may be a direct path around the loop in your QC. That path will be fast and direct, while the loop involves analog-digital conversions that add latency. That delayed signal, when mixed with the direct signal, will cause phase cancellations
Click to expand...
Interesting, thanks very much!
 
sprint said:
I have HXStomp in my Ax3 loop for post fx and it works well without phase issues as long as the HX loop is routed in series after the Axfx amp/cab model ie:

... Ax Amp+Cab Model > AxLoopOut > HXIn > HX Post FX model (or HX bypass signal) > HX Out > AxLoopRtn > AxOutput.

The same routing logic should work in the context of FM3 in QC loop (make sure only QC is going out to monitoring and FM3 loop occurs serially after QC amp/Cab).
Click to expand...
Awesome, thanks very much.
 
dschaaf said:
Awesome, thanks very much.
Click to expand...
👍 - fyi the above assumes mono, - the QC cab could be placed after your FM3 post fx loop if preferred. For stereo post fx there are more considerations. In my case I want my HXStomp post fx running stereo between a mono Ax3 amp model and a stereo Ax3 cab block (hard panned left/right IRs), so my more detailed routing is:

... Ax Amp+Cab Model-L > AxLoopOut-L > HXIn-L > Stereo HX Post FX model (or stereo HX bypass signal) > HX Out-L/R > AxLoopRtn-L/R > AxOutput-L/R.

(the loop is
placed at the point where my signal changes from mono to stereo)
 
I think it might be solved… I inadvertently must have bumped the mix setting on the QC FX loop from 100% down to 89% and so it was letting some of the other signal through. When I pushed it up to 100% mix, it almost entirely went away :)
 
That will definitely cause it. Parallel loops don't work with digital gear except for some delay based effects due to the conversion latency. 100% should be the equivalent of a serial loop.
 
mr_fender said:
That will definitely cause it. Parallel loops don't work with digital gear except for some delay based effects due to the conversion latency. 100% should be the equivalent of a serial loop.
Click to expand...
This is not true.

All you need to do is to delay the direct path to match the FX Loop's latency (FX Loop's D/A and A/D, plus any other digital devices using their analog I/O connected in series within the FX Loop). You can use published specs, or measure the latency with any DAW.

In the Axe FX I use the Flanger Block as a micro-delay to do this (0ms to 10ms out to two decimal places, Mix 100% wet, Modulation 0%). It's likely the QC and L6 stuff can do something similar, but I'm guessing as I don't have any of their HW.
 
That can work IF the device in question has a fixed processing latency and you have a means to accurately match that latency. Many devices, including Fractal gear do not have a fixed latency. It can change depending on what you add to the grid.
 
mr_fender said:
That can work IF the device in question has a fixed processing latency and you have a means to accurately match that latency. Many devices, including Fractal gear do not have a fixed latency. It can change depending on what you add to the grid.
Click to expand...
Sure, though you measure as needed to account for the differing latencies due to changes to the grid, noting that the latency correction doesn't have to be perfect to the sample either, and just like when aligning mics, there are also creative aspects when playing with the phase response between paths (i.e. use your ears once it's ball-park'd...)

In the Fractal, you can easily compensate up to 10ms with the Flanger, out to two decimal points, which is way more granularity than what's needed IME.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom