Rear Input Quieter than the Front Input?

Jason Scott

Jason Scott

Fractal Fanatic
I was discussing this issue in the Q9.01 Firmware Release thread but that's not the place for it, so I I'm creating a dedicated thread here.

After trying the rear input, I found it has approx. 1dB less noise at the same volume than the front input. It has me wondering if that's normal?

Bakerman, iaresee and Moke offered some feedback and helpful suggestions, so I'm posting this here to continue that discussion.
 
Jason Scott said: ↑
I compared the response in FabFilter and it appears that there's more low-end energy (44Hz - 115Hz) in the rear input sample and more energy at 265Hz and 697Hz in the front input sample. Here are a couple of samples:


My ears heard more Low on the Rear input and a bit less mids.. and a bit less low,but more mids on the Front. Just listening while playing. Wierd.. but i now like the Rear input
 
Bakerman said:
Was there anything in the preset during this test, like a high-gain amp block?
Click to expand...

It's a high gain preset with amp, cab and a few filter blocks.

Bakerman said:
Did you adjust input levels for identical LED response?
Click to expand...

In I/O, the front input is set to 36%. Input 1 is set to 50%.

Bakerman said:
Playing guitar during this step might not be accurate enough for this test.
Click to expand...

I recorded samples while playing and idling.

Bakerman said:
The front input processing reduces noise at higher frequencies. That might not affect a basic level meter very much. It's better to view the spectrum of the noise to see the difference.
Click to expand...

I compared the response in FabFilter and it appears that there's more low-end energy (44Hz - 115Hz) in the rear input sample and more energy at 265Hz and 697Hz in the front input sample. Here are a couple of samples:

Front Input
Rear Input
 
SeeD said:
My ears heard more Low on the Rear input and a bit less mids.. and a bit less low,but more mids on the Front. Just listening while playing. Wierd.. but i now like the Rear input
Click to expand...

I hear more low-end using the rear input, as well.
 
Those two .wav files are the same recording.

By "Playing guitar during this step might not be accurate enough for this test," I meant when you're adjusting input levels. It's better to find exactly where a test tone gives a solid red LED than to play notes on a guitar, where you can't be sure you're repeating the previous output level. An Ebow can make a guitar your tone generator, but you could also just run a synth tone from Out 2 to each input.
 
Last edited:
Bakerman said:
Those two .wav files are the same recording.
Click to expand...

Crank up the output. The difference is subtle, definitely not night and day. The RMS values in my DAW show about a 1dB difference, and I can see a difference using FabFilter's EQ Match. One track isn't really a problem, but layering 3 or more can add up.
 
Last edited:
Why do the files have the exact same thump around 0.6s? These waveforms match perfectly and cancel to leave much quieter white noise. I think you made an error somewhere between recording & uploading these.

BUrZSp6.png
 
Did you have a guitar connected with volume knob turned up? The biggest difference seems to be more 60 Hz hum in the front input clip. Front vs. rear input shouldn't affect that very much.

Here's a clip of front/rear/front/rear on a Mark II with input levels matched carefully. First half is input noise with gtr. volume off, ~90 dB of gain added but no other processing. Second half is a preset (037 FAS Modern w/ more gain) with guitar volume at maximum.

https://www.dropbox.com/s/celwxzvd6wj5fn7/front-rear-noise.mp3?dl=0
 
Bakerman said:
Did you have a guitar connected with volume knob turned up?
Click to expand...

Yes, all the way up, correct. The only thing that changed between the two samples was the input.

Bakerman said:
Here's a clip of front/rear/front/rear on a Mark II with input levels matched carefully. First half is input noise with gtr. volume off, ~90 dB of gain added but no other processing. Second half is a preset (037 FAS Modern w/ more gain) with guitar volume at maximum.

https://www.dropbox.com/s/celwxzvd6wj5fn7/front-rear-noise.mp3?dl=0
Click to expand...

The rear input being louder in those clips, I'm not sure what to make of this issue. I mean, it's not entirely significant, however it's not completely trivial to me, either.
 
I haven't checked the difference between front and rear input.µBut I noticed an increase in noise with the power connected to my laptop when the Axe-Fx was connected by USB.
 
so here's something interesting, in doing some noise tests today on the front and rear inputs. I attached a 1M ohm resistor between the tip and ring of a 6 inch guitar cable and plugged this into the input. Using the Plexi 100 High amp block and a speaker block.

The noise at the rear input has a definite pattern to it, but the front input noise is aperiodic

Have a look at the waveforms here. Front input is the bottom wave form.
https://www.dropbox.com/s/8r52jclayapvzsk/axefx front vs rear noise floor.png?dl=0

Also by my measurements the front input has a noise floor about 10dB higher than the rear on my unit.
 
Last edited:
groovenut said:
so here's something interesting, in doing some noise tests today on the front and rear inputs. I attached a 1M ohm resistor between the tip and ring of a 6 inch guitar cable and plugged this into the input. Using the Plexi 100 High amp block and a speaker block.

The noise at the rear input has a definite pattern to it, but the front input noise is aperiodic

Have a look at the waveforms here
https://www.dropbox.com/s/8r52jclayapvzsk/axefx front vs rear noise floor.png?dl=0

Also by my measurements the front input has a noise floor about 10dB higher than the rear on my unit.
Click to expand...
Location, location, location. :)

The interference being picked up in that location will be different from what's picked up at the other end. That includes the attachment leads to the resistor, and the resistor itself. Both are unshielded, and they will be the biggest source of noise pickup. If you were holding those connections by hand, the interference source will be much bigger: it'll be you. :)
 
Rex said:
Location, location, location. :)

The interference being picked up in that location will be different from what's picked up at the other end. That includes the attachment leads to the resistor, and the resistor itself. Both are unshielded, and they will be the biggest source of noise pickup. If you were holding those connections by hand, the interference source will be much bigger: it'll be you. :)
Click to expand...
Right. The cable was not being held or moved between the front and back tests. The attachment leads were very short, two alligator clips soldered to one another. While I understand the 1 inch of unsheilded hot lead that connected to one end of the resistor could have formed a small antenna, it is very likely much less susceptible than a guitar pickup. Either way, I measured slightly less noise with my test lead than I did with an actual guitar, but the difference between the front and rear input noise floor was the same in both test cases. ~10dB

Here is an FFT plot of both, Yellow is the front input
https://www.dropbox.com/s/to2cn25z39lp2s3/axefx front vs rear noise floor freqs.png?dl=0

What was more interesting to me was that the noise at the rear input was periodic and the noise at the front input was aperiodic.

I contacted support this morning asking about the difference in noise floor between the front input and rear. Fractal support stated "There is a difference in the two inputs (front and rear) in that the front input is an instrument input whereas the rear is a line input level. Also, factory settings for the front are left only and for the rear it is summed left +right. I do not think there is anything abnormal with the unit. You have been sent a parts link if you choose to replace the front input board if you still believe there is an issue with the unit. "

Please keep in mind, I never used the noise gate prior to FWQ8. Never needed to. Nothing has changed in my setup, only the firmware. My question to fractal support was in order to hopefully verify that my unit had an issue, but according to their response they dont believe there is anything abnormal about my unit.

I know something is wrong with my unit, what I dont know is if it is hardware or software. Very few people seem to be having the issue, so I have to guess hardware. I will order the replacement input board, since it affordable and hope that it solves the issue. Either way I will report back my findings. For now I am using the rear input and with no noise gate. It sounds like it did before the update (noise wise) only better tone and feel (FWQ9).
 
Last edited:
Here's another weird thing. I can get a 6dB improvement in noise on the front input if I change the I/O >Audio> Input 1 Mode from Left to L+R Sum regardless of the input setting on the amp block
 
Last edited:
That's because it sums the channels and reduces the level. Your guitar signal level will drop by a slightly greater amount than the noise dropped, so ultimately it just gives a slightly worse SNR.
 
You must log in or register to reply here.
Back
Top Bottom