Review and Measurements of Audio-gd NFB28.28 DAC and Headphone Amp

[Killingbeans]

Looking at their website, you might think they're more obsessed about making pretty cirtcuitboards over well-measuring ones.
As NwAvGuy has said, electrical performance has nothing to do with components being arranged in neat, symmetrical rows and columns.

All of those through hole film capacitors (Wimas?) for supply decoupling on SMD chips also puzzles me. Sure, the red caps on a black PCB looks pretty, but high quality ceramic SMD caps would be a far better choice.

 

[Veri]

All of those through hole film capacitors (Wimas?) for supply decoupling on SMD chips also puzzles me. Sure, the red caps on a black PCB looks pretty, but high quality ceramic SMD caps would be a far better choice.

Yes, at this price, "WIMAs" you're looking at what, almost $100 worth of caps in there? I'm betting counterfeit. Nothing against Chinese-made audio gear really, but for audio-gd's asking price it's somewhat suspiciously "impressive" to stuff so much in there.

 

[rajapruk]

I get impressed the other way around, if very few components acheive god measurements.

 

[bunkbail]

@amirm, I know that you're already fed up measuring these badly engineered gears, but can you please make one last test on this unit? Audio-gd higher end stuffs tend to have a stupid amount of components especially relays for volume control, so I've been wondering how good their volume matching actually is.

 

[Andrex]

Maybe an explanation...
.... A lot of Burson AOP

https://jipihorn.wordpress.com/2013/04/02/le-mythe-des-aop-burson-et-audio-gd/

In French, but with a lot of measures.
Take the time to watch the video

 

[Andrex]

A lot of details and schemas to explain why Burson Op amp are really crap

 

[amirm]

Are you sure whatever was rattling around inside wasn't a vital shorting jumper for the filter settings and that has compromised the test results?

Yes, I am pretty sure. Filter jumpers don't cause these kinds of problems and results are consistent with equipment design (lack of feedback) and prior testing of other units from Audio-gd.

 

[amirm]

I do believe that Kingwa and Audio-gd could make good measuring products if they really want to, at least from their internal build quality, but they just strictly adhere to an ancient and basic principle (no negative feedback) that makes them measure terrible independent of their build quality.

There is a problem there where they show measurements saying despite lack of negative feedback, their products measure great. There are pictures of audio precision analyzer in use and such. This is not proper. They need to fix this as a minimum they need put proper measurements there.

How much lack of feedback hurts you also depends on design, verification and component selection. The more linear the basic design, the less need for feedback.

 

[amirm]

omg... been seriously considering the NFB28.38 to pair with the HD6XX. Some even say the NFBs sound "tube-ish". After reading lots of good reviews on NFB , with these measurements I now feel so lost and confused.

I think most people are confused about "tube sound." It is not like this is something that you can test for with a toggle switch. They buy a tube product with completely different characteristics, like it for whatever reason (due to improper testing), and then assign advantage to a product having "tube sound."

I have yet to find a high-distortion product -- tube or solid state -- do anything for me. I just hear the distortions that take away from the fidelity of music source. Yet all of these products are praised for having a positive tube sound. The difference for me is that I do controlled side-by-side testing and there, it becomes clear with careful listening, you just hear the impact of distortion. Or it is not audible.

Any "warmth" would have to come from rolling off the high-frequencies which a tube amp with high output impedance can impact.

Anyway, the jury is out on whether there is any merit here on making designs with such high levels of distortion.

 

[Veri]

Here's another beautiful piece of bullshit from their R2R product:

True NOS design. NOS is less sensitive to jitter and offers neutral sound quality .To avoid the digital artifacts, a uniquely designed noise rejection processing technology is applied.

NOS is less sensitive to jitter?!!

 

[JJB70]

I still think if people want distortion its more sensible to introduce it via DSP then end up lumbred with it. If you tire of DSP induced euphonic distortion you switch the DSP off, bit pricier if you've gone and spent $$$$$$$$$s on an amp just to have a distorted sound.

 

[Jorj]

I'm happy I sold my NFB28, v2015 version. It was not awful sounding, but when my D10 arrived, I realized that the D10 sounded just as good, if not better (the differences are hard to hear, if it's possible at all), and cost a small fraction of the NFB28. I should have sent it to Amir first, but this one is a newer version with similar design, and my bet is it would have tested just as badly.

Kingwa is a nice guy, always responsive, but he's making claims his gear does not seem to back up. He could do better if he chose to, I'm sure.

 

[VintageFlanker]

Here's another beautiful piece of bullshit from their R2R product:

True NOS design. NOS is less sensitive to jitter and offers neutral sound quality .To avoid the digital artifacts, a uniquely designed noise rejection processing technology is applied.

NOS is less sensitive to jitter?!!

Yep. As I owned two of their R2R units, I strongly disapprove this claim. Both (R2R-11 and R2R-2) sounded way more linear and "neutral" set to 8X oversampling. With jumpers set to NOS, you get nothing but sluggish bass that bleed entirely to the low mids and backward, inaccurate and distorted treble.

 

[graz_lag]

Shittier than Schiit.

Difficult ... but possible ...

 

[bunkbail]

Yep. As I owned two of their R2R units, I strongly disapprove this claim. Both (R2R-11 and R2R-2) sounded way more linear and "neutral" set to 8X oversampling. With jumpers set to NOS, you get nothing but sluggish bass that bleed entirely to the low mids and backward, inaccurate and distorted treble.

But the R2R-11 is NOS only though.

 

[Veri]

But the R2R-11 is NOS only though.

Software oversampling maybe?

 

[bunkbail]

Software oversampling maybe?

How do you do that? HQPlayer?

 

[VintageFlanker]

But the R2R-11 is NOS only though.

I believed it's not. You can switch jumpers as well as NFB-11 with ESS filters. I made Experiments using jumpers to set what I was thinking to be a "true" NOS mode ("sound warmer" says the "manual").

Edit: I may be wrong, maybe it's only an analog tuning

 

[bunkbail]

I believed it's not. You can switch jumpers as well as NFB-11 with ESS filters and at the exact same location. I made Experiments using jumpers to set what I was thinking to be a "true" NOS mode ("sound warmer" says the "manual").

Nope. I own an R2R-11 myself, the only thing the "warm" jumper do is to roll-off the high frequency. The DAC is still operating in NOS mode, regardless of the jumper settings (neutral/warm). Neutral jumper sounds great to my ears.

 

[Veri]

How do you do that? HQPlayer?

That's one option. There's many others though, I have used SoX on foobar before which is also available on Pi-hats. SoX is great and free