[Tilpo]

I finally got my hand on one of Audio-GD's Digital interface.
Since they both used the same USB receiver chip, the result should be similar.
But it's not! At least not on my NFB-12:

NFB12's USB receiver, with the same driver, recorded by the same recorder. Both my RME's optical & Audio-GD's Digital Interface got 96kHz properly.
However, my NFB12's USB receiver got serious cut-off after 24kHz.
If it's driver or my foobar's setting error, it should be the same with Audio-GD's Digital Interface, but it's not.
So the problem would be something wrong from NFB12's USB receiver to DAC. 
 

That does seem like a bad thing, but do you honestly expect to hear anything at all at those frequencies? As long as it's flat until about 18kHz (and it is) it doesn't matter what happens after that. In fact, playing back the high/low frequencies that you can't hear in addition to those you can hear will only be a source of distortion.

 

[Kingwa]

The PCM  audio signal usually have limit at 20Hz to 20KHz.
 
The fast roll off must by the digital filter built in WM8741, its duty is cut the non-audio signal (noise )  which is higher than 20KHz.
Please read the article of the digital filter can make you understand.
 
The NFB12 if change the digital setting can keep the 20-20KHz no roll off.
We don't want give users a subjective point while listen, so we have not  post which one is best for the frequency band test. This is not contact to the sound quality. Some gears have large roll off still consider sound good .

 

[killkli]

Quote:

That does seem like a bad thing, but do you honestly expect to hear anything at all at those frequencies? As long as it's flat until about 18kHz (and it is) it doesn't matter what happens after that. In fact, playing back the high/low frequencies that you can't hear in addition to those you can hear will only be a source of distortion.

You didn't get to my point.
ALL THREE LINE ARE THE TESTING RESULT FROM NFB12
Just the digital sources are different.
 
1. My computer -> Audio-GD Digital Interface Coax-out -> NFB-12(Coax)
2. My computer -> RME babyface Optical-out  -> NFB-12(Optical)
3. My computer -> NFB-12(USB)
 
Using a 24bits/96kHz test file.
 
If the signal was correctly sent from the source to the DAC, all three line should be the same with minor jitter noise, but NOTHING should be different in frequency response.
It's the same with both Audio-GD Digital Interface/RME babyface Optical-out. BUT my NFB12's USB ALTERED the frequency response. THe high is not simply roll off, It's kind of "CUT" off.
Which means from the USB receiver chip to DAC, there is something LOST.
It's not about sound signature, it's about design flaw or unit flaw.
 
KINGWA, your response is still not even touched my point.

 

[Digital-Pride]

Quote:

You didn't get to my point.
ALL THREE LINE ARE THE TESTING RESULT FROM NFB12
Just the digital sources are different.
 
1. My computer -> Audio-GD Digital Interface Coax-out -> NFB-12(Coax)
2. My computer -> RME babyface Optical-out  -> NFB-12(Optical)
3. My computer -> NFB-12(USB)
 
Using a 24bits/96kHz test file.
 
If the signal was correctly sent from the source to the DAC, all three line should be the same with minor jitter noise, but NOTHING should be different in frequency response.
It's the same with both Audio-GD Digital Interface/RME babyface Optical-out. BUT my NFB12's USB ALTERED the frequency response. THe high is not simply roll off, It's kind of "CUT" off.
Which means from the USB receiver chip to DAC, there is something LOST.
It's not about sound signature, it's about design flaw or unit flaw.
 
KINGWA, your response is still not even touched my point.

I see your point.  At which point in the upper frequencies is the cut off you mentioned? 
 

 

[killkli]

Around 20~22kHz, which means that perhaps the DAC received 44 or 48kHz data.
 

 

[Digital-Pride]

Quote:

Around 20~22kHz, which means that perhaps the DAC received 44 or 48kHz data.
 

Hmm, how's the frequency response to that 20 to 22 kHz point?  

 

[scannon18]

I think I'll go for a different dac/amp. Since hearing the sansa clip + ive decided that i dont really care for "warm" gear.

 

[Digital-Pride]

Quote:

I think I'll go for a different dac/amp. Since hearing the sansa clip + ive decided that i dont really care for "warm" gear.

But, how are going to keep yourself warm during those cold winter nights?

 

[scannon18]

But, how are going to keep yourself warm during those cold winter nights?:wink_face:

Audio equipment is for music. Females for warmth

 

[renyiandhan]

Quote:

Around 20~22kHz, which means that perhaps the DAC received 44 or 48kHz data.
 

 I got you, but it looks like you just confused different concepts
44.1, 48, 96, 192,384k ... are sample rates, they have nothing to do with frequency response range(20~20khz...)

 

[killkli]

Quote:

 I got you, but it looks like you just confused different concepts
44.1, 48, 96, 192,384k ... are sample rates, they have nothing to do with frequency response range(20~20khz...)

I have not confused anything.
All test was done with 96kHz sample rate data, which means they should all be able to reproduce up to 48kHz sound.
But when the data goes from NFB-12's USB receiver, the reproduction is not correct. Other digital interface such as RME's babyface or even Audio-GD's own Digital Interface got 96kHz data to the DAC correctly.
It clear from the FR plot. Both babyface and digital interface to NFB12's DAC reproduced quite flat FR to about 30kHz, then roll off to 48kHz, it's okay since there is LPF(low pass filter) working.
At 48kHz sound, it's about -9dB roll-off, actually, NFB-12's LPF's roll-off is about -6dB, but my recorder card has it's own roll-off too, so I got about -9dB here.
 
BUT!!!!! From NFB-12's USB receiver, the roll-off (it's actually CUTOFF, but RMAA smoothed it, can't disable it) start around 22kHz, the sound is virtally none from 24kHz onward.
 
Why all you guys keep saying I mistaken something or not? I have complete idea about what I'm doing, and I have correct knowledge about what I'm doing.
I do recording work sometimes. It's part of my job okay?
 

 

[renyiandhan]

take it easy, buddy
but FR is NOT sample rate, right? and your test is about FR roll-off, so it is unrelated to the sample rate
PS: from some earlier tests, the NFB12's FR rolled off from about 10khz,  worse?

 
Quote:

I have not confused anything.
All test was done with 96kHz sample rate data, which means they should all be able to reproduce up to 48kHz sound.
But when the data goes from NFB-12's USB receiver, the reproduction is not correct. Other digital interface such as RME's babyface or even Audio-GD's own Digital Interface got 96kHz data to the DAC correctly.
It clear from the FR plot. Both babyface and digital interface to NFB12's DAC reproduced quite flat FR to about 30kHz, then roll off to 48kHz, it's okay since there is LPF(low pass filter) working.
At 48kHz sound, it's about -9dB roll-off, actually, NFB-12's LPF's roll-off is about -6dB, but my recorder card has it's own roll-off too, so I got about -9dB here.
 
BUT!!!!! From NFB-12's USB receiver, the roll-off (it's actually CUTOFF, but RMAA smoothed it, can't disable it) start around 22kHz, the sound is virtally none from 24kHz onward.
 
Why all you guys keep saying I mistaken something or not? I have complete idea about what I'm doing, and I have correct knowledge about what I'm doing.
I do recording work sometimes. It's part of my job okay?
 

 

 

[renyiandhan]

killkli's test told us that how much a better digital interface could enhance(or extend) the NFB-12's frequency response range

 

[renyiandhan]

whether a FR roll-off is good or bad , it has nothing to do with the ability of  taking music of high sample rates, they belong to different categories.
though i'm also sad to own an older version without filter

  
a tasteless cake compared to new version or 12.1

 

[Tilpo]

"The ACSS is a non-feedback technology made with fully discrete amplifiers. Most people know the global feedback design can offer better specs in test measurements, and non-feedback can't do well in test measurements but can offer better sound for the human's ears"
This lack of negative feedback stated here would mean a reduction in bandwidth. Maybe this is what causes the high frequency roll-off, and not the USB receiver chip or anything like that.

It only states that it doesn't have global feedback, on the other hand. Maybe it does have local feedback.


Kingwa, if you read this, would you be kind enough as to elaborate?