[Elanzer]

What I actually meant by this, is that the other devices hear the whiring as well as the NFB-12 when no resampling is being done, but when I resample it's different as the whiring is pretty much completely gone and replaced with an analog feedback-like sound in the background. I prefer to listen to it with the resampling for obvious reasons (no annoying whiring and thus dial more pronounced), but if it's "wrong" to have the whiring removed, then resampling is doing things worse from a reference point of view.
 
Quote:

Quote:

If the whiring noise is intended, then resampling actually seems to make things worse. All my other devices put forward the whiring noise just like the NFB-12 is doing at 44KHz without any resampling, so I don't see how 44KHz playback is broken.

LOL.  I actually think it improves a bit, the whirring becomes more focused.  However, not tested on NFB-12 as I don't have it yet (may go for NFB-11), but tested on all my other DACs.

 

[holland]

Quote:

What I actually meant by this, is that the other devices hear the whiring as well as the NFB-12 when no resampling is being done, but when I resample it's different as the whiring is pretty much completely gone and replaced with an analog feedback-like sound in the background. I prefer to listen to it with the resampling for obvious reasons (no annoying whiring and thus dial more pronounced), but if it's "wrong" to have the whiring removed, then resampling is doing things worse from a reference point of view.

 
Ah, yes.  Agreed.  It doesn't disappear for me though, still audible at higher sampling rates.  It does become less pronounced when going to 96kHz and above, again identical on all DACs I have.

 

[Cankin]

Do you guys with NFB-12 hear that whiring noise with anything other than that udial.wav file?

 

[baylaef]

I notice supercurio's result.
There are artifacts(noise) between 22k and 44k range.
This means the over sampling of WM8741 is not good.
If oversampling works well, artifacts should go to higher range and removed by digitalflter easily.
 
WM8741 may not do interpolation at middle/high rate mode. So artifacts stay between 22k and 44k range. 
It's not pin configuration issue but WM8741's own performance.
WM8741's oversampling does not work practically.
 
We can't hear the noise itself over 20k, but the beat noise made by sound and higher noise may be heard.
So high artifacts shall be removed.
 
Quote:

Hi again!
 
I ran new analysis with improved settings and setup (more adequate gains, only 1 computer to avoid noise sources, better cables)
 
This time, benchmark results are better, but you'll see easily the artifacts I was talking about with the additionnal spectrograms at 44k.
Roll-off is confirmed due to a digital filter applied.
 
Spoiler alert: Upsampling is not optional.
 
 

44k/24bit performance
http://supercurio.project-voodoo.org/audio/audio-gd/RMAA/audio-gd-NFB-12-44k-24bit.htm
 
96k/24bit performance
http://supercurio.project-voodoo.org/audio/audio-gd/RMAA/audio-gd-NFB-12-96k-24bit.htm
 
192k/24bit performance
http://supercurio.project-voodoo.org/audio/audio-gd/RMAA/audio-gd-NFB-12-192k-24bit.htm
 
Samplerates comparison
http://supercurio.project-voodoo.org/audio/audio-gd/RMAA/audio-gd-NFB-12-samplerates.htm
 
44k/24bit without or with upsampling
http://supercurio.project-voodoo.org/audio/audio-gd/RMAA/audio-gd-NFB-12-44k-upsampling.htm
 
udial at 44k, sampled at 96k :

 
udial upsampled at 96k, sampled at 96k:

 
RMAA test wave at 44k, sampled at 96k :

 

RMAA test wave upsampled at 96k, sampled at 96k:

 
 
PS: sox graphs are generated with: 

  for x in *.wav ;do sox $x -n spectrogram -o $x.png -x 1600 -s -t "$x" -w Kaiser -z 120; done

 

[holland]

Quote:

I notice supercurio's result.
There are artifacts(noise) between 22k and 44k range.
This means the over sampling of WM8741 is not good.
If oversampling works well, artifacts should go to higher range and removed by digitalflter easily.
 
WM8741 may not do interpolation at middle/high rate mode. So artifacts stay between 22k and 44k range. 
It's not pin configuration issue but WM8741's own performance.
WM8741's oversampling does not work practically.
 
We can't hear the noise itself over 20k, but the beat noise made by sound and higher noise may be heard.
So high artifacts shall be removed.

The NFB-12 sets the WM8741 into 2x oversampling mode (192kHz sampling rate for the oversampling rate control).  It would be best to analyze with a FFT instead of udial.

 

[Greg121986]

Would the NFB-11 be a better choice than the NFB-12?

 

[Slaughter]

I am probably the first, but just sold my NFB-12. It's been good conversation. Hopefully everyone continues to enjoy this great amp/dac.

 

[Poetik]

Quote:

I am probably the first, but just sold my NFB-12. It's been good conversation. Hopefully everyone continues to enjoy this great amp/dac.

Probably not the first but I'm sure you're one of them.  I just got my NFB-11 today and I'm thinking of selling off my NFB-12 as well.   The differences are pretty much as Audio-GD states so it really boils down to how you like your music presented. 

 

[Moving Air]

I believe wolfson has good DAC chips....won't say that the NFB-12 is not good.  Diff. people have different expectations and requirements....as I'm not using the module for headphone purpose...instead using it for my music system....Thus, my expectation is clean/pure signal path quality, as flat/wide a freq. response as possible,...no noise and minimal intermodulation distortion...hopefully the NFB-11 can deliver it.
 
Thanks a lot for telling me that NFB-11 is available....I have just changed the order with Kingwa.   ))  Keep the fire burning for headfi.....really cool to meet people like you ....the pursuit for good sound quality....
 
 
ps: someone ask about freq.roll off, this link might help with your understanding...
 
http://forum.ecoustics.com/bbs/messages/34579/131062.html

 

[BrianMendoza]

Hey guys, quick question... I bought an NFB12 to use as a DAC for my Concerto until I can afford a StageDAC. Anyways, I've read a couple post and saw a couple graphs showing the NFB12 having some treble roll off.
 
This may sound dumb, but do the results of those charts also represent the DAC out? If it only matters to those using the NFB-12 as both a DAC & an amp, than it won't be a worry for me, because I only plan on using my NFB-12 as a DAC.

 

[Elanzer]

Quote:

Hey guys, quick question... I bought an NFB12 to use as a DAC for my Concerto until I can afford a StageDAC. Anyways, I've read a couple post and saw a couple graphs showing the NFB12 having some treble roll off.
 
This may sound dumb, but do the results of those charts also represent the DAC out? If it only matters to those using the NFB-12 as both a DAC & an amp, than it won't be a worry for me, because I only plan on using my NFB-12 as a DAC.

 
It will effect the DAC out as well. It's very minor (less than -2dB for the audible range), but it does give the DAC a slightly warm bias because of it.
 
Audio-GD reintroduced the NFB-11 (No USB), so if you haven't been shipped your order yet you could always have them change it, it might have better synergy with the LCD-2 as well, based off how many LCD-2 users have enjoyed the NFB-10ES with it.

 

[BrianMendoza]

Quote:

It will effect the DAC out as well. It's very minor (less than -2dB for the audible range), but it does give the DAC a slightly warm bias because of it.
 
Audio-GD reintroduced the NFB-11 (No USB), so if you haven't been shipped your order yet you could always have them change it, it might have better synergy with the LCD-2 as well, based off how many LCD-2 users have enjoyed the NFB-10ES with it.

Oh ok. Thanks for the quick reply btw! I actually just bought mine from another member, and it'll be in the mail tomorrow.
 
I guess as long as it's satisfactory as a DAC, my real intention for it is just to keep me satisfied until I can save up some more and go for a StageDAC or something along those lines.
 
And maybe the warmth will be a good thing...

 I just recently went from a Mapletree Ear HD+ to the Concerto, and immediately felt the loss of tube warmth. I can't say whether I miss it or not yet, but maybe the NFB-12's warmth will be a good medium.

 

[dw1narso]

@Elanzer: thanks a lot for very thorough tests...
 
Things actually seems more complicated but also become clearer at certain extend...
 
In my case, I go ahead and convert and upsample Supercurio's file, using R8Brain, to produce fixed 16-bit 44KHz WAV file and 16-bit-48KHz WAV file.
 
Here are the files: http://www.sendspace.com/file/pjsofp (48KHz) and http://www.sendspace.com/file/i1o7zp (44KHz)
 
What I hear on my PC is still the same, whether I play the file on foobar or load it on Audacity... 44KHz introduced "noise" , while 48KHz sounded clean.
 
Now, here is the interresting part. I upload those files to my BlackBerry (BB) phone... and played both file. Guest what... I got the reverse result!!! the 44KHz sounded clean while the noise now appear on the 48KHz file.
 
I try to play music with lots of high frequency content (like cymbals) but I never experience the same problem as the Supercurio's udial.wav file.
 
So, the possibilities now are:

• There's a problem on my PC driver/codec/Conexant chip that make it noisy when handling 44.1KHz sample rate.
• My BB actually cannot handle sample rate higher than 44.1KHz, but can handle 44.1KHz rate properly
• Would it be possible that there's a problem with Supercurio's PC soundcard's chip that outputing SPDI/F signal? (notice that he experience the same problem with Windows and Linux)
• Would it be possible that there's a problem on the WM8805 (or WM8741) inherited issue or implementation issue on NFB-12 when receive the 44.1KHz stream
• There's something wrong on Supercurio's udial.wav file.... maybe he can update us, how did he get/create this file? downsample from a file with higher sample rate? (now, there's possibility that there is anti-aliasing error/noise during the process...)

 
@Elanzer: if you don't mind, could you download both files on my link and tried them on your Cowon D2 and Nokia? Just for curiosity, whether these two devices can accept the 48KHz rate properly...

 

[dw1narso]

 
Quote:

Quote:

It will effect the DAC out as well. It's very minor (less than -2dB for the audible range), but it does give the DAC a slightly warm bias because of it.
 
Audio-GD reintroduced the NFB-11 (No USB), so if you haven't been shipped your order yet you could always have them change it, it might have better synergy with the LCD-2 as well, based off how many LCD-2 users have enjoyed the NFB-10ES with it.

Oh ok. Thanks for the quick reply btw! I actually just bought mine from another member, and it'll be in the mail tomorrow.
 
I guess as long as it's satisfactory as a DAC, my real intention for it is just to keep me satisfied until I can save up some more and go for a StageDAC or something along those lines.
 
And maybe the warmth will be a good thing...

 I just recently went from a Mapletree Ear HD+ to the Concerto, and immediately felt the loss of tube warmth. I can't say whether I miss it or not yet, but maybe the NFB-12's warmth will be a good medium.

Signature wise, IMO, LCD-2 would be benefited from ESS chip character...
 

 

[Audiophil]

checked both signals with:
Dell D630 - foobar 1.1 - asio4all -usb - iBasso Viper - Beyer DT880studio
 
Both are sounding identical, no obvious distortion.
But when i had a look at foobar's spec-analyzer view I recognized in the second half of both signal a loud peak @20kHz (-20 dB). The normal signal is only about -40 dB. But again, on both. Maybe some decoders have problems with this signals...