Question about the 24bit 48Khz Fractal AX FX III

Les Paul Rider

Les Paul Rider

Member
I recently bought some new monitors (Focal solo6 ST6 "the new 2022 model") I went to my store intending to buy the Dynaudio Lyd-48 but I was comparing them to other monitors Neumann KH 310, Adam, Genelec etc. and for me the Focals were clearly superior in sound quality, I can't say anything in terms of flat sound reproduction, so I couldn't test it, anyway I don't have a studio ready for good sound reproduction. The question that comes to me is that I wanted to test the sound quality and I wanted to use an album in DSD audio format and the player tells me that the quality of the files are equivalent to 32bit 705.600Hz, if I play the file from my laptop - USB to the Fractal AX FX III - XLR to Monitors, the question is: Will the sound quality that comes out of the monitors be a maximum of 24bits 48KHz? Do I need a DAC to handle files over 48KHz? Thank you so much.
 
Try it with headphones plugged into the AFX3. I mention this because it will instantly answer your question (same conversions).

Also, I think you meant 705.600kHz not Hz.
 
Last edited:
Tha
pedalbuilder said:
Yes, although you almost certainly won’t hear any difference unless you have magical ears that can hear frequencies over 24k.
Click to expand...

I read an article that mention this:

“To an Objectivist, 20kHz and above does not matter for one plain and simple reason: humans cannot hear above 20kHz. Ergo, if we cannot hear something then it doesn’t matter. Case closed.

While the base statement is accurate, it ignores the consequences of implementation.

If we instead look towards the practical application of high bandwidth amplifiers we see that we could ask a very different question and get a very different answer. The right question to be asking is this:

Does phase accuracy matter within the passband of human hearing?
Click to expand...
Now we have a proper question that circumvents all the narrow focus arguments that plague us.

The answer to question 2 is an unequivocal yes! Human hearing is very sensitive to phase differences within the audio passband.

Circling back to the first question, we can now answer it with another unequivocal yes! because there are no practical means of bandwidth limiting an analog amplifier to 20kHz without affecting the phase. It’s typically important to have a bandwidth approaching 100kHz to get phase accurate performance in the audio passband.”

So.. since my monitors only reproduce up to 48Khz, I guess the only difference from a DAC would be given by the 32 bits instead of 24 bit from the fractal, or would the difference not be noticed either?
 
Les Paul Rider said:
Tha


I read an article that mention this:

“To an Objectivist, 20kHz and above does not matter for one plain and simple reason: humans cannot hear above 20kHz. Ergo, if we cannot hear something then it doesn’t matter. Case closed.

While the base statement is accurate, it ignores the consequences of implementation.

If we instead look towards the practical application of high bandwidth amplifiers we see that we could ask a very different question and get a very different answer. The right question to be asking is this:


Now we have a proper question that circumvents all the narrow focus arguments that plague us.

The answer to question 2 is an unequivocal yes! Human hearing is very sensitive to phase differences within the audio passband.

Circling back to the first question, we can now answer it with another unequivocal yes! because there are no practical means of bandwidth limiting an analog amplifier to 20kHz without affecting the phase. It’s typically important to have a bandwidth approaching 100kHz to get phase accurate performance in the audio passband.”

So.. since my monitors only reproduce up to 48Khz, I guess the only difference from a DAC would be given by the 32 bits instead of 24 bit from the fractal, or would the difference not be noticed either?
Click to expand...
I found the answer to my question in this link:
https://www.makeuseof.com/16-bit-vs-24-bit-audio/
 
I hope Axe Fx 4 is 32bit 192khz or higher cos why not? At least on the digital i/o side, because I record at 96khz.

Psychoacoustics are real, but so are placebos. Maybe they are mutually inclusive, maybe generic measurement techniques are shit. Regardless there's some very nice recordings in those formats, especially because they take so much effort to produce.

Case in point, guitar amps and modelers.
 
Les Paul Rider said:
Tha


I read an article that mention this:

“To an Objectivist, 20kHz and above does not matter for one plain and simple reason: humans cannot hear above 20kHz. Ergo, if we cannot hear something then it doesn’t matter. Case closed.

While the base statement is accurate, it ignores the consequences of implementation.

If we instead look towards the practical application of high bandwidth amplifiers we see that we could ask a very different question and get a very different answer. The right question to be asking is this:


Now we have a proper question that circumvents all the narrow focus arguments that plague us.

The answer to question 2 is an unequivocal yes! Human hearing is very sensitive to phase differences within the audio passband.

Circling back to the first question, we can now answer it with another unequivocal yes! because there are no practical means of bandwidth limiting an analog amplifier to 20kHz without affecting the phase. It’s typically important to have a bandwidth approaching 100kHz to get phase accurate performance in the audio passband.”

So.. since my monitors only reproduce up to 48Khz, I guess the only difference from a DAC would be given by the 32 bits instead of 24 bit from the fractal, or would the difference not be noticed either?
Click to expand...
All modern DACs have linear phase response. Poorly written article.
 
FractalAudio said:
All modern DACs have linear phase response. Poorly written article.
Click to expand...
and were Rupert Neve still alive he would tell you the story of Geoff Emerick and the console input module with an analog anomaly at approx 43kHz — that produced clearly-heard sonic side effects inside the range of human hearing. Don’t kid yourself that anything above 20kHz “doesn’t matter”.
 
touch33 said:
and were Rupert Neve still alive he would tell you the story of Geoff Emerick and the console input module with an analog anomaly at approx 43kHz — that produced clearly-heard sonic side effects inside the range of human hearing. Don’t kid yourself that anything above 20kHz “doesn’t matter”.
Click to expand...

Exactly. Maag Air Band, anyone?
 
touch33 said:
and were Rupert Neve still alive he would tell you the story of Geoff Emerick and the console input module with an analog anomaly at approx 43kHz — that produced clearly-heard sonic side effects inside the range of human hearing. Don’t kid yourself that anything above 20kHz “doesn’t matter”.
Click to expand...
I’ve heard that story. Neve would relate an anecdote that implied Emerick could hear something nobody else could hear, but it’s hard to say if what he was hearing was high frequencies or even if the story is true :).

That said, I agree audio above 20 kHz can matter when it undergoes processing, and that includes d/a conversion, which is why we all record at sample rates above 40 khz.
 
touch33 said:
and were Rupert Neve still alive he would tell you the story of Geoff Emerick and the console input module with an analog anomaly at approx 43kHz — that produced clearly-heard sonic side effects inside the range of human hearing. Don’t kid yourself that anything above 20kHz “doesn’t matter”.
Click to expand...
Of course it can but that's unrelated. 43kHz is only about an octave above the range of hearing. All depends on what the "analog anomaly" is. If it's a notch or a highly resonant peaking response it can affect the range of interest. Something outside the range of interest can easily produce measurable effects in the range of interest. Nonlinearities outside the range of interest can also cause measurable distortion in the range of interest. None of this is a surprise but also completely irrelevant to the discussion.

As an example you can have, say, a shelving filter with a center frequency of 40 kHz and a gain of 20 dB. The zero of that filter is at 12kHz so clearly it has an effect on the band of interest despite its center frequency being well outside.

Modern DACs are flat to within a fraction of a dB from 0 to over 20kHz and have extremely linear phase response.

32-bit DACs are a marketing gimmick. The only 32-bit DACs I'm aware of are made by AKM and have a dynamic range of 121 dB. That's only 20 bits. The 12 remaining bits are useless. A 32-bit DAC would theoretically need to have 192 dB of dynamic range to use all the bits. It's extremely difficult to get much over 100 dB out of anything and the dynamic range of virtually every analog transducer is less than that.
 
GlennO said:
I’ve heard that story. Neve would relate an anecdote that implied Emerick could hear something nobody else could hear, but it’s hard to say if what he was hearing was high frequencies or even if the story is true :).
Click to expand...
That sounds like the emperor had new hearing. :)
 
LanceZ said:
Exactly. Maag Air Band, anyone?
Click to expand...
Just because the critical frequency of something is ultrasonic doesn't mean that it doesn't have measurable effects in the audio band. See my example a couple posts up.

People are conflating the ability of ultrasonic filters to effect audible frequencies with the ability to hear beyond 20kHz (or more like 14kHz for a typical adult).
 
FractalAudio said:
32-bit DACs are a marketing gimmick. The only 32-bit DACs I'm aware of are made by AKM and have a dynamic range of 121 dB. That's only 20 bits. The 12 remaining bits are useless. A 32-bit DAC would theoretically need to have 192 dB of dynamic range to use all the bits. It's extremely difficult to get much over 100 dB out of anything and the dynamic range of virtually every analog transducer is less than that.
Click to expand...
Uaudio Apollos use ESS Sabre 32 bit adc/dac but no joke it is a meme. Higher sample rates are handy, especially with lazy plugin devs that don't bother researching oversampling algos (so many!). Disk space is cheap, it sounds good, why not?

FWIW, really excellent 48khz conversion is expensive as hell, another side of the argument.
 
johnnypig said:
Uaudio Apollos use ESS Sabre 32 bit adc/dac but no joke it is a meme. Higher sample rates are handy, especially with lazy plugin devs that don't bother researching oversampling algos (so many!). Disk space is cheap, it sounds good, why not?

FWIW, really excellent 48khz conversion is expensive as hell, another side of the argument.
Click to expand...
The dynamic range for their "Flagship" DAC is only 120dB. Again only about 20 bits. They look like good products but they're far from "32 bit".
 
You must log in or register to reply here.
Back
Top Bottom