Technical Question

Search4Tone

Inspired
This question is for someone who is certain of the answer. If you aren't certain of the answer please leave the response to someone who is. I don't mean to be inconsiderate, I just want to get the correct answer.

It's my understanding that the front input uses an analog high frequency emphasis circuit.

My question is: At exactly what point in the signal chain is the high frequency emphasis removed and the signal re-equalized to a flat response?

Thanks.
 
..And this was where I was headed with my inputs question. Thank you for asking.. and thank you Java for the answer.

Java: Not sure if this is answerable then, but I must ask: that which was asked and answered as being proprietary, is it present only on the front input? Or is this present on the other inputs? If no, is there a similar or other process going on with them? Or is the front input the only "effected" input?
 
The front's the only 'effected' input from what I've read, but I could be totally wrong - such is the nature of proprietary matter.

Question for those in the know as well - is the front input's circuit physically different or is it just the special sauce algorithm that makes it different from the rear?





edit - found this clip

Cliff: "You have to set the input selection to match the input you're using. If you're using the front input then you must set the input selection to front and vice-versa. If you plug something into the front and set the input selection to rear it will get MUCH brighter. The front input is optimized for guitar-level inputs and has spectral shaping and more gain than the rear input. The front input is optimized for guitar pickups. This is a combination of hardware and software processing. If you set the input source to Analog Rear this turns off the software processing part. If you are plugged into the front it will change the tone since you're still going through the hardware processing. This is why I say you must match the input selection to the input you are using. The rear inputs are standard line-level inputs and can be used with any program material. The front input, as stated above, is optimized for guitar pickups. As such it has more gain and less headroom and may clip if used for non-guitar program material. If you plug a guitar directly into the rear you may find you don't have enough signal level."

from this thread: http://forum.fractalaudio.com/axe-fx-ii-discussion/48798-inputs-question-soft-clip-normalize-3.html
 
Last edited:
The front's the only 'effected' input from what I've read, but I could be totally wrong - such is the nature of proprietary matter.

Question for those in the know as well - is the front input's circuit physically different or is it just the special sauce algorithm that makes it different from the rear?





edit - found this clip

Cliff: "You have to set the input selection to match the input you're using. If you're using the front input then you must set the input selection to front and vice-versa. If you plug something into the front and set the input selection to rear it will get MUCH brighter. The front input is optimized for guitar-level inputs and has spectral shaping and more gain than the rear input. The front input is optimized for guitar pickups. This is a combination of hardware and software processing. If you set the input source to Analog Rear this turns off the software processing part. If you are plugged into the front it will change the tone since you're still going through the hardware processing. This is why I say you must match the input selection to the input you are using. The rear inputs are standard line-level inputs and can be used with any program material. The front input, as stated above, is optimized for guitar pickups. As such it has more gain and less headroom and may clip if used for non-guitar program material. If you plug a guitar directly into the rear you may find you don't have enough signal level."

from this thread: http://forum.fractalaudio.com/axe-fx-ii-discussion/48798-inputs-question-soft-clip-normalize-3.html

That information was for the legacy units, not the II. The II cuts off the front when the rear is selected and vice versa. So it's not possible to screw up your EQ anymore.

However, assuming the II has the same optimized front input which I believe I remember reading it does, the only thing I can infer from the above is that the hi-freq de-emphasis occurs immediately after the A/D when using the front input. If it didn't, a signal with high frequency emphasis would be fed to the amp blocks when using the front input, and a signal with flat EQ would be fed to the amp blocks when using the rear inputs. This would radically change the character of the distortion. This conclusion assumes the II has similar architecture as the Ultra and Standard.

Setting the controls to front input must engage the hi-freq de-emphasis circuit that follows the A/D, and setting to rear input must disengage the de-emphasis circuit since the rear input has no "soft limiter" hi-freq emphasis/clipping circuit.

Perhaps I've missed something, but that's the only inference I can draw from the behavior of the inputs and knowing that the front input has the soft limiter which as I recall is simply a hi-freq emphasis circuit that precedes the A/D and clips the EQed signal if the front input is driven too hard and also prevents the input of the A/D from ever being overloaded, and the rear input does not have the "soft limiter" (which is actually a hard clip of an EQed signal) but is simply a normal input.

It's possible there is more to it than that and that the rear input has some form of emphasis/de-emphasis, but I found out what I wanted to know and this jogs my memory of how the legacy units operated. The simultaneously active front and rear inputs of the legacy units was revealing in terms of how the front input was EQed.

Thanks for the quote!
 
as they say BUSINESS IN THE FRONT, PARTY IN THE REAR!!!!!!!!!!!!!!!!!!!!!!!!!!! ps you said rear input, huh huh huh. oops did I digress again, time for bed.
 
The II cuts off the front when the rear is selected and vice versa. So it's not possible to screw up your EQ anymore.

IIRC, the front Instrument Input and rear Input 1 L&R are grouped together in an either / or. But the rear Input 2 L&R can be active at the same time as the front Instrument input (I omitted USB, SPDIF, AES/BU on purpose for simplification).

Richard
 
I really don't get why this would help you in any way. It seems more like you're trying to reverse engineer the unit that just playing and tweaking it.
 
It doesn't appear that any emphasis is used. Here is a frequency curve of just the AD/DA - analog I/O (left rear for both) with everything is bypassed for the Axe II.

I had this curve from some testing I did in posting the update to the frequency response curves of the factory speaker cabinets (in the User Cabs & IR section) of the Axe II.

This doesn't prove emphasis/de-emphasis isn't happening (if so it's implemented very well) but that it doesn't appear to be an "effected" input (as some have worded it).

Hope this helps.

Keith
 

Attachments

  • Analog I:O Left Rear.jpg
    Analog I:O Left Rear.jpg
    69.2 KB · Views: 38
It doesn't appear that any emphasis is used. Here is a frequency curve of just the AD/DA - analog I/O (left rear for both) with everything is bypassed for the Axe II.

I had this curve from some testing I did in posting the update to the frequency response curves of the factory speaker cabinets (in the User Cabs & IR section) of the Axe II.

This doesn't prove emphasis/de-emphasis isn't happening (if so it's implemented very well) but that it doesn't appear to be an "effected" input (as some have worded it).

Hope this helps.

Keith

The effect is apparently reversed in software (somewhere before the grid). The net effect is a flat response. However, it is on the front input. So your test would not capture that.
 
Last edited:
Here is a contentious post from the good 'ole days where some of this was discussed. It was referring to the ultra/standard, but imagine it is still mostly valid.
Axe-FX :: View topic - TIPS SECTION

H
ere is the most telling post:
"

AlbertA wrote:
Aahhh, so the hardware pre-emphasises the high frequency content before sampling and then de-emphasizing it in software once sampled I suppose....

http://en.wikipedia.org/wiki/Dolby_noise_reduction_system



Exactly, but it's not dynamic like Dolby so you don't get the funky pumping artifacts.

The spectrum of a guitar is pink(ish). Above 800 Hz or so the energy rolls off dramatically. As luck would have it, humans perceive noise above 800 Hz or so to be most objectionable as it manifests itself as hiss.

So the front input pre-emphasizes the high frequencies and then does the inverse in software. This has the net effect of a flat frequency response but pushes the noise floor down by the amount of the pre-emphasis.

It's an old trick, used in FM radio and vinyl records. The basic premise is to optimize the data conversion to the information content of the source.

CC"




I don't think Cliff will reveal anymore detail about it than this.

 
Last edited:
If the info about the processing is proprietary *and* the reason for needing the info is proprietary...

does that cancel out? or is proprietary ** 2? (squared).

Richard
 
Back
Top Bottom