[PPkiller]

udial at 44k, sampled at 96k :
udial upsampled at 96k, sampled at 96k:
 
will like to ask, how do u go about setting upsampled, and the sampled frequency? will like to find out the method u used, jus to rule out software resampling which took place in the OS audio path.
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

 

[j2kei]

Ok so I was one of the earlier people that complained about the NFB-12 sounding bad, however it has been 2 weeks and i've come to report my findings as promised. lemme say that the sound on this baby HAS CHANGED! the amount of change is significant. the sound is so much cleaner, bass is not bloated and messy as it was before. the mid and highs are starting to come out. in fact, now i sometimes wonder if there is enough bass. the bass has definitely become more textured instead of some bloated blob that irritates the hell out of me. i feel the highs are pretty extended and sound pretty good on my hd600. i also felt that before, there was a sort of veil. the sound was there, however it felt mask and unclear. now it is nice, clear, and squeeky clean. also, before when i made the report, i felt that the nfb-12 made my hd600 sound like an hd650. now i can gladly say that my hd600 out of my nfb-12 sounds like how an hd600 should sound, neutral with some light sparkly smooth highs . i urge you guys to continue using your nfb-12. this thing required A LOT of burn-in for it to sound good. i had it on for 2 weekends w/o playing sounds, around 6 nights where i would let it connect to a cheapo headphone and run music through. 

 

[Audiophil]

Nobody of us is able to listen locally to this instrument's 'sound' in advance. We have to trust on the information we get from audio-gd (and on this nice discussion in forum of coarse).
I don't like 'sounds', because this means to distort the original signal. Several companies are offering products with a 'sound', this is usually a company 'sound', some may like it and know it in advance. With audio-gd it's a bit more complicated, they offer different sounds and describe it with nice words, all a matter of interpretation, no facts. And I cannot judge, because I have not heard it...
Therefore it seems to be essential to get detailed spec's of proper equipment
 
For me it makes no sense to implement the abilty to work on high resolution in compromising the basic resolution 44,1/48 kHz. At least for me this represents 95% of my music.
 
P.S.: Maybe they implemented the filter because of the artefacts??

 

[Poetik]

Quote:

Nobody of us is able to listen locally to this instrument's 'sound' in advance. We have to trust on the information we get from audio-gd (and on this nice discussion in forum of coarse).
I don't like 'sounds', because this means to distort the original signal. Several companies are offering products with a 'sound', this is usually a company 'sound', some may like it and know it in advance. With audio-gd it's a bit more complicated, they offer different sounds and describe it with nice words, all a matter of interpretation, no facts. And I cannot judge, because I have not heard it...
Therefore it seems to be essential to get detailed spec's of proper equipment
 
For me it makes no sense to implement the abilty to work on high resolution in compromising the basic resolution 44,1/48 kHz. At least for me this represents 95% of my music.
 
P.S.: Maybe they implemented the filter because of the artefacts??

 
Yea you figure on a budget amp/dac it would be optimized to use the basic resolution as opposed to the higher ones.  It sounds pretty good so far but knowing that it could have been better is another thing.  It is a promo item though so I wouldn't expect it to be fully optimized yet.  Luckily we have people on here that will do the tests to help us out in the long run.

 

[Slaughter]

It could have sounded better? If the "issue" was resolved, it would sound like an NFB-11. That's like saying that if the HD650's "issue" was fixed, it would sound better. Well some people like the sound of the 650 over something like the 880. The roll off is what gives it the smooth sound as described on their website. What is the difference between headphones that sound smooth and headphones that sound detailed and analytical?

 

[killkli]

Well, I see that the resampling is indeed an issue since non-existing information did produced by the setting. This is not "right."
 
But as far as RMAA is concerned, I don't think it's such a good reference. See how my Firestone Spitfire wins in every score in RMAA testing:

RightMark Audio Analyzer test report

Testing device	Spitfire
Sampling mode	24-bit, 48 kHz
Interface
Testing chain	External loopback (line-out - line-in)
RMAA Version	6.2.3
20 Hz - 20 kHz filter	ON
Normalize amplitude	ON
Level change	-1.0 dB / -1.0 dB
Mono mode	OFF
Calibration singal, Hz	1000
Polarity	correct/correct

Summary

Frequency response (from 40 Hz to 15 kHz), dB	+0.00, -0.05	Excellent
Noise level, dB (A)	-106.9	Excellent
Dynamic range, dB (A)	106.9	Excellent
THD, %	0.0010	Excellent
THD + Noise, dB (A)	-93.4	Very good
IMD + Noise, %	0.0021	Excellent
Stereo crosstalk, dB	-106.4	Excellent
IMD at 10 kHz, %	0.0024	Excellent
General performance		Excellent

Frequency response

From 20 Hz to 20 kHz, dB	-0.17, +0.00
From 40 Hz to 15 kHz, dB	-0.05, +0.00

Noise level

	Left	Right
RMS power, dB	-106.0	-105.4
RMS power (A-weighted), dB	-107.2	-106.5
Peak level, dB FS	-84.3	-83.6
DC offset, %	+0.0	+0.0

Dynamic range

	Left	Right
Dynamic range, dB	+106.0	+105.4
Dynamic range (A-weighted), dB	+107.2	+106.5
DC offset, %	+0.00	-0.00

THD + Noise (at -3 dB FS)

	Left	Right
THD, %	+0.0010	+0.0010
THD + Noise, %	+0.0019	+0.0021
THD + Noise (A-weighted), %	+0.0020	+0.0023

Intermodulation distortion

	Left	Right
IMD + Noise, %	+0.0018	+0.0023
IMD + Noise (A-weighted), %	+0.0015	+0.0018

Stereo crosstalk

	Left	Right
Crosstalk at 100 Hz, dB	-102	-104
Crosstalk at 1000 Hz, dB	-105	-106
Crosstalk at 10000 Hz, dB	-99	-98

IMD (swept tones)

	Left	Right
IMD + Noise at 5000 Hz,	0.0024	0.0024
IMD + Noise at 10000 Hz,	0.0022	0.0024
IMD + Noise at 15000 Hz,	0.0025	0.0025

 
It's almost perfect score as far as RMAA testing is concerned.
But in reality, I enjoyed TDA1543 based non-feedback & non-oversampling DAC more than Spitfire. Remember, TDA1543 has the same roll-off in high frequency as shown in NFB-12.
 
One more point: in reality, your headphone's FR response difference is "MUCH" greater than what little roll-off the DAC could present.
For example, my HD600 has about almost 15dB differences in high frequency domain.
 
Really...... it's not such a big deal. Though the resampling setting does seemd to be problematic.
There should not be any information higher than 22kHz when the input digital data is 44kHz only.

 

[Audiophil]

sorry Slaughter, we are not talking regarding 'better'. This is a matter of individual taste.
In case of HD650 I don't know the data, but if people report it's smoothness or softness or less agressiveness I would assume, it has a slight decrease at high frequencies.
- If you wold combine it with NFB-11, maybe this can be perfect compensated.
- If you wold combine it with a more or less linear DAC, it would be very similar than 'original'.
- If you wold combine it with NFB-12, you will combine 2 times decreased high frequencies, which is maybe too much lost.
 
Please, take it just as examples, I don't know the data in detail.
 

 

[Audiophil]

Hi Killkli,
 
perfect data indeed, I agree. This box should sound more or less neutral, not necessarily perfect, because not everything influencing the sound is easy measurable..
 
If you enjoy the 1543 in combination with HD600 is simple your personal taste, but means also, that you may have now 18 dB change of high frequencies.
You are totally right, the deviation of speakers, headphones and ears are much larger and more dificult to design
 
 
Quote:

  There should not be any information higher than 22kHz when the input digital data is 44kHz only.

 
Yes, and if you at the same time the 20kHz want to be transmitted, this is also a complicated task and not possible with a 6dB RC filter...

 

[JaMo]

Hi everybody.
I am a person who likes the fact that Kingwa runs a company under upgoing based on a very sund idea by bringing the market a very competent product line of gear and this for generally, very attractive prices. I respect him alot for this. More people should.
 
It hurts to see the critisism from some users though. I fully agree with Kingwa when he sais that when You design an item You must have a very precise idea/goal with the item.
 
The NFB-12 which I also have ordered is a very competent piece of equipment on paper. I haven't heard it yet but I am sure it will be great, based on what users experienced and reported here. For USD200 it is a bargain... and everybody here knows this for sure. Of course these lower priced products in this line have a potential for further improvements.... But please don't expect Kingwa to make excuses for his work. If You want to squeese out more.... please do.. And if You do it, the smarter way is to wait a while for the warranty to expire and a proper burn in.
 
If You want a more complete setup, choose the higher gear: REF- or NFB-7 , -8
 
@Kingwa and his staff. Thank You and keep up Your work and Your high ambition as before.
 
/Jan

 

[Audiophil]

Hi Jan,
I underline most of your post.
Remember, I had also my reasns, when I ordered the NFB-12.
But why not improve, when it is easy going? Personally I would also go for 50% more, if the result is as expected.
On the other hand not everybody is able or willing to solder SMD parts...
 
My only concern to audio-gd is: As they are located far away and not present at local dealers, where you could test the staff, it would be nice to get more exact information regarding the characteristic of the boxes.The use of nice words instead of tech data causes misunderstandings.
 
BR
Michael

 

[Poetik]

Quote:

It could have sounded better? If the "issue" was resolved, it would sound like an NFB-11. That's like saying that if the HD650's "issue" was fixed, it would sound better. Well some people like the sound of the 650 over something like the 880. The roll off is what gives it the smooth sound as described on their website. What is the difference between headphones that sound smooth and headphones that sound detailed and analytical?

 
You're right with the different flavor and all but something tells me even with a rolloff similar to an NFB-11 they're 2 totally different chips so they'd have a different sound regardless of the filter. 
 
As for wanting it to sound like an NFB-11 we'll see.  I get mine in about a week so I'll be able to truly "hear" the difference. 
 
 
 
 
 
 

 

[Slaughter]

Exactly! People are wanting the NFB-12 to sound like something its not and wasn't designed to sound like. The NFB-12 is smooth, period. That's why I have said all along that it would be a good match for Beyers and Grados. I find it excellent with the M50 which can have stident highs. If you don't want to smooth out the highs on your phones, then it's not the amp/dac for you. I just don't understand why people want Kingwa to change the sound when it was clear what they were buying.

 
Quote:

sorry Slaughter, we are not talking regarding 'better'. This is a matter of individual taste.
In case of HD650 I don't know the data, but if people report it's smoothness or softness or less agressiveness I would assume, it has a slight decrease at high frequencies.
- If you wold combine it with NFB-11, maybe this can be perfect compensated.
- If you wold combine it with a more or less linear DAC, it would be very similar than 'original'.
- If you wold combine it with NFB-12, you will combine 2 times decreased high frequencies, which is maybe too much lost.
 
Please, take it just as examples, I don't know the data in detail.
 

 

[haloxt]

Quote:

How would the NFB-12 compare to the original Compass.  Just wondering if I should look into something newer.

The compass has a "natural" or "organic" sound to it. If NFB-12 is similar to wm8741 sparrow, then NFB-12 should be a good amount more technically detailed with more precise soundstage, less "musical", but neutral. I like both, but definitely my sparrow is more technically capable. The main difference between sparrow and higher end audio-gd gear is timbre, sparrow presents such microdetail as bolder macrodetail, while higher end gear can express microdetail more precisely, subtly and nuanced. If you don't need such high resolution microdetail, the wm8741 sparrow is really great, and so should the NFB-12 be.

 

[Kingwa]

In our NFB11 page we always post the different sonice between NFB11 and NFB12.
http://www.audio-gd.com/Pro/Headphoneamp/NFB11/NFB11EN.htm
 
English is not our mother tongue maybe we can't exact explain the sonic flavors.
 I think this link even though reviews the NFB2, but  the NFB12 sound flavor is quite near the NFB2 although sound quality is different level. Maybe can easier understand than our describing . They both are put the targets at smooth sounding and had a different to our ES9018 and PCM1704 based products.
http://www.digitalaudioreview.net.au/index.php/audio-reviews/digital-source-reviews/item/177-audio-gd-nfb-2-dac-19-series

 

[haloxt]

Kingwa, your sound signature descriptions are perfectly fine. Maybe consider increasing the font size so people can notice it better ... something like this.
 
The NFB-11 has quite high fidelity, its sound signature is absolutely revealing, neutral and faithful and very detailed. It can show how good or how bad a recording is. If you are afraid of flaws being revealed in the music files, other gears like those applying WM8741 will be better.