[Elanzer]

Quote:

So, what's the best "oversampling" setting?

 
Most people prefer the 8x Linear Phase Apodising filter, it's widely considered the best filter on the WM8741 DA chip. This will limit you to 48KHz output.
 
If you need 96KHz support, try 4x Linear Phase Soft-Knee filter, since the 4x Linear Phase Apodising filter isn't available on audio-gd's gear.

 

[captouch]

Quote:

 
Most people prefer the 8x Linear Phase Apodising filter, it's widely considered the best filter on the WM8741 DA chip. This will limit you to 48KHz output.
 
If you need 96KHz support, try 4x Linear Phase Soft-Knee filter, since the 4x Linear Phase Apodising filter isn't available on audio-gd's gear.

I liked 8x Minimum Phase Apodising filter the best, but don't recall if I tried every single one.  Did you compare 8x Linear Apodising to 8x Minimum Apodising?
 

 

[Chronos]

 
Quote:

I liked 8x Minimum Phase Apodising filter the best, but don't recall if I tried every single one.  Did you compare 8x Linear Apodising to 8x Minimum Apodising?
 
 

I really didn't like the NFB-12 without any filters, in fact I was so disappointed I wanted to get rid of it right away.  I thought it was way too warm, sucked out all the bass and fun from my Denon D7000's.  I tried the 8x Minimum phase adipose filter setting also and really like it now.  That headfonia review really missed out by trying this thing stock.  

 

[captouch]

 
Quote:

 


I really didn't like the NFB-12 without any filters, in fact I was so disappointed I wanted to get rid of it right away.  I thought it was way too warm, sucked out all the bass and fun from my Denon D7000's.  I tried the 8x Minimum phase adipose filter setting also and really like it now.  That headfonia review really missed out by trying this thing stock.  

Agree, I was disappointed as well until I went with the revised filter setting.  Agree that Headphonia missed out and did a disservice to the unit by not at least trying other settings.  Someone suggested it, but guess they were too busy to try it.
 
 

 

[eclipes]

Overexaggerated review and unwillingly to try other filters when numerous people suggested the 8x filter. Ever since that, disappointed in headfonia.

 
Quote:

 

Agree, I was disappointed as well until I went with the revised filter setting.  Agree that Headphonia missed out and did a disservice to the unit by not at least trying other settings.  Someone suggested it, but guess they were too busy to try it.
 
 

 

 

[ac500]

I wish someone would write a simple guide on how to select among the most popular filters. Because you can't expect every reviewer to know either.

 

[glenda]

Quote:

I wish someone would write a simple guide on how to select among the most popular filters. Because you can't expect every reviewer to know either.

Come on its right on the Audio-gd website plain as day.  
 
The least the reviewer could have down is upsampled so the right filter was being used, I think the reviewer purposely handicapped the NFB12 to prove a point about plug and play or something.   He's forgeting that the average buyer is not so inept as to make a few adjustments for their system. 
 
Again this discussion points to why we need to push Audio-GD to software control which unlocks the 4x apodizing filters for Hirez, even the 10SE doesn't have the WM8471 in software mode.
 
The other thing is the dither features,  I know that is a bad word to some but all oversampling has it and it would be nice to be able to select each of the four available inside the WM8471.   Its a top end chip when it is under software control,  thats the difference between its use in such a wide range of gear (ie everything from ($100 to near 5 figures.)  
 
Thinking about opening my NFB12 and trying the hifidino controller (which is opensource with the code right on his webpage for complete software control) and possibly Ciuffoli's LL1527 output stage but changing the output stage would probably be a mistake:
 
See I've been measuring the NFB12  as a standalone DAC and its distortion is relatively high,  not bad but surprisingly higher than even a decent tube dac.   But the key unique feature/design of this nfb12 is the inegration to the headphone amp and the complete absense of typical color adding negative feedback.   In other words after I was measuring the dac output I also measured the headphone out and the distortion spectra/level was the same,  even as the volume knob (power out) was turned up !  
 
This is the beauty of Kingwas's analog design in that the entire system is scaled as the volume knob is turned,   so that up to a point a change in volume does not equal a change the sound quality.   Thats one of the dirty secrets about most any gear you buy,  the dac output is attenuated then the amp with its fixed gain amplifying an attenuated signal has increasing distortion with input level.  Ever notice how most separate amplifiers sound "different" with a slight turn of the volume knob?   And many times you fight it trying to find that sweet spot where everything sounds good at the right volume?  
 
With the NFB its volume control is handled completely different.   So as standalone DAC performance of the NFB12 is average but the way it is integrated to its amp the end result is less distortion,  and without huge amount of NFB added to your music there is much less color than the typical dac+amp.

 

[ac500]

I'm not sure why this is, but I've noticed if I turn the volume knob down very near absolute 0, the left ear cuts out completely before the right ear. In other words extremely low volumes the volume becomes unbalanced, shifting the sound to the right ear only. Not a big issue because I can just switch to low gain.
 
Also interesting is that turning the knob introduces very audible noise while I'm turning it, but that's not too unexpected I think assuming the knob is a potentiometer. The noise goes away when I stop turning it of course.

 

[BleaK]

Quote:

No there  are two 4x apodizing filters locked on these wolfson DAC's,  the big expensive companies that use these in their high end gear have programmers who can unlock and give acces to these filters with the serial port (see the datasheet.)   These 4x apodizing filters are what made the wolfson famous for 24/96 hirez (no upsampling required.) 
 
Notice that most have found that the hardware acessable 8x apodizing filters the better of the limited selection offered by AudioGD but not compatible with hirez.   The same holds true with 2x for 192k hirez and 4x 96k hirez, for hirez you want the software locked apodizing filters but it looks like Kingwa hasn't figured it out yet.      Waiting till he does out before I upgrade,  I have faith.  In the mean time the NFB-12 is an outstanding value with an excellent analog section.   If you are interested in the subject google the WM8471 datasheet, apodizing oversampling,  and arduino programming.
 
 

 
This would be awsome if Kingwa figured it out and implemented in future products like NFB-12 and NFB-10SE.
 

 

[glenda]

Quote:

I'm not sure why this is, but I've noticed if I turn the volume knob down very near absolute 0, the left ear cuts out completely before the right ear. In other words extremely low volumes the volume becomes unbalanced, shifting the sound to the right ear only. Not a big issue because I can just switch to low gain.
 
Also interesting is that turning the knob introduces very audible noise while I'm turning it, but that's not too unexpected I think assuming the knob is a potentiometer. The noise goes away when I stop turning it of course.

Its typical with volume pots,  but more pronouced in this applicatio,  the main reason the 10se is a big upgrade is that the pot is replaced with ladder relayed resistors,  so near perfect LR matching across the volume range.
 

 

[captouch]

Quote:

Come on its right on the Audio-gd website plain as day.  
 
The least the reviewer could have down is upsampled so the right filter was being used, I think the reviewer purposely handicapped the NFB12 to prove a point about plug and play or something.   He's forgeting that the average buyer is not so inept as to make a few adjustments for their system. 
 
Again this discussion points to why we need to push Audio-GD to software control which unlocks the 4x apodizing filters for Hirez, even the 10SE doesn't have the WM8471 in software mode.
 
The other thing is the dither features,  I know that is a bad word to some but all oversampling has it and it would be nice to be able to select each of the four available inside the WM8471.   Its a top end chip when it is under software control,  thats the difference between its use in such a wide range of gear (ie everything from ($100 to near 5 figures.)  
 
Thinking about opening my NFB12 and trying the hifidino controller (which is opensource with the code right on his webpage for complete software control) and possibly Ciuffoli's LL1527 output stage but changing the output stage would probably be a mistake:
 
See I've been measuring the NFB12  as a standalone DAC and its distortion is relatively high,  not bad but surprisingly higher than even a decent tube dac.   But the key unique feature/design of this nfb12 is the inegration to the headphone amp and the complete absense of typical color adding negative feedback.   In other words after I was measuring the dac output I also measured the headphone out and the distortion spectra/level was the same,  even as the volume knob (power out) was turned up !  
 
This is the beauty of Kingwas's analog design in that the entire system is scaled as the volume knob is turned,   so that up to a point a change in volume does not equal a change the sound quality.   Thats one of the dirty secrets about most any gear you buy,  the dac output is attenuated then the amp with its fixed gain amplifying an attenuated signal has increasing distortion with input level.  Ever notice how most separate amplifiers sound "different" with a slight turn of the volume knob?   And many times you fight it trying to find that sweet spot where everything sounds good at the right volume?  
 
With the NFB its volume control is handled completely different.   So as standalone DAC performance of the NFB12 is average but the way it is integrated to its amp the end result is less distortion,  and without huge amount of NFB added to your music there is much less color than the typical dac+amp.

 
If we're only talking about using the NFB12 as a standalone DAC, does the better signal-to-noise apply as well at higher volume knob levels (on the variable output where the knob actually controls the output level)?  Or are your observations above only applicable to the HP out?
 
 

 

[Jrz]

Sorry for noob question, here it comes:
 
Which one of DT 880 I buy (250 ohm or 600 ohm), I have already ordered NFB-12?

 

[glenda]

Quote:

 
If we're only talking about using the NFB12 as a standalone DAC, does the better signal-to-noise apply as well at higher volume knob levels (on the variable output where the knob actually controls the output level)?  Or are your observations above only applicable to the HP out?
 
 

I dind't open it but I think the headphone and and the variable preamp out are the same, the switch just directs the headphone out to the rca and viola its a preamp, they measured the same.  Its not just snr its the distortion spectra that stays amazingly constent from line level DAC out to higher power headphone out..
 
 

 

[captouch]

Quote:

I dind't open it but I think the headphone and and the variable preamp out are the same, the switch just directs the headphone out to the rca and viola its a preamp, they measured the same.  Its not just snr its the distortion spectra that stays amazingly constent from line level DAC out to higher power headphone out..
 
 

 
If that's the case, then the pre-amp out would best be used in variable mode with volume knob turned near max and volume attenuated at the amp.  Fixed output is equal to volume knob at 12 o'clock, so better distortion characteristics would result near max.  Is my understanding correct?  Thanks.

 

[Digital-Pride]

Quote:

Sorry for noob question, here it comes:
 
Which one of DT 880 I buy (250 ohm or 600 ohm), I have already ordered NFB-12?

The amp inside the NFB-12 is quite robust.  So, go ahead and get that 600 Ohm beast.