EDIT: Reduced rating... After spending some additional weeks with the product I came to feel less satisfied with the audio quality. When compared to the iBasso D12, it performed well enough. However, in absolute terms, and for a desktop unit, there's just too much distortion to give a 4 star review. Running some basic test tones clearly allowed me to hear low-frequency THD that should not have been audible. So I reduced my ratings somewhat. The bottom line is that this seems like a rather under-realized discrete system. From a discrete product, I would have expected better, even at this price. I'd have to say it's good value, but that there are probably other alternatives in the $300+ range that might warrant consideration as well.
EDIT #2: I hard a little time to spare so I isolated the problem I was experiencing with THD to the SPDIF input. I seem to have a defective receiver, which is adding major THD to low-frequency signals. When I redid my tests recently via USB, the signal was very clean. At this point, it seems I have a defective product and a warranty issue. I am bumping my rating back to 4 stars as the USB input works well. The sound is definitely better on USB than SPDIF, which is not typical with devices. With the right filter selection, the unit is quite great for the price. It's not a giant killer, but for under $300 I have heard nothing else that is close. I will post another update to discuss the warranty support, and if this addressed the SPDIF issue.
Decisions…
I wanted a desktop DAC/AMP for my office. My goals were cheap and bulletproof operation, and multiple inputs to support my various devices. I have used a D12 for this purpose, but moving it from home to work every day and constantly plugging stuff in and out seemed like a lot of effort.
I considered the NuForce HDP seriously. It’s a very elegant unit, and they appear to have done a smashing job on the DAC, in particular. However, for my budget, it was more than I really wanted to spend at work, because honestly I do critical listening in my man-cave, on the Mac/Burson 160D rig.
I thought of another D12, but when I saw the Audio GD NFB-12, I liked the idea of the discrete topology, as well as the fact that it had the same DAC I have enjoyed in the D12.
Time to order. Well, it was weird. Audio-GD is very Chinese. The website is about the worst I’ve ever seen, and while I got same-day response to every question, I had to ask my questions over several times in different ways, and on some issues, like shipping fees, never actually understood the answer. I used Paypal as a way to ensure that if I didn’t get what I expected I would have some financial remedies available.
Arrival and Unboxing
Via DHL, it took four working days from placing my order to delivery, not quite as quick as the iBasso D12, but pretty snappy. When the Audio-GD NFB-12 arrived, it was in a generic, heavily taped cardboard box.
Inside, the unit was armored in tape and foam, very beefy packing for a small unit. On unwrapping the unit, I found a few spare LEDs and some plastic bits with no docs to explain them. Given the MTBF of LEDs, I don’t get why there are extras at all. There was also a four wire connector with no explanation and generic/cheap USB and TOSLINK cables. The shorting jumpers, I figured out going back to the website, are for changing the upsampling rates and the type of digital filter used. (I didn’t have this info when I bought the product… Whoopee, an unexpected, very cool extra feature!). Very geeky, and probably much easier to have done using a DIP on the back of the enclosure.
The fit and finish is OK but not stellar. The metalwork is not particularly high tolerance, and the design is, well, functional. The switches are old-school, and even partially cover the text on the front panel. Also, for some reason the font size varies, being larger for Gain and Volume than for Output and Input.
The metal is rounded, which is nice, and is pleasingly hefty. The unit does feel more like industrial gear than a HP amp, it’s quite “built” in the way that Arnold Schwarzenegger used to be… I note that in the back corner, the metal is not well aligned, there’s about a 0.3 mm height variance in the metalwork, so it’s not flush. Also, the metal had a lot of oil on it and took a few passes to get clean. The volume pot had a very nice feel to it, and tracked extremely well at low levels.
On the back panel, the connectors are very high quality, way beyond what one would expect for $240 (DHL shipping included). In fact, this unit was roughly $40 less than the D12.
I think the best way to describe it is that the FEEL and look of the unit is high-end US Audio gear from the 1980s, functional and tough, with a discrete high-quality components vs. the ICs and op-amps of the consumer market.
Setup for Test
I am using 100% lossless audio for testing, 16/44.1, 24/88.2 and 24/96, stored on a firewire drive, feeding a MacBook Pro, via iTunes and PureMusic. Outputs are being tested via USB and via M2Tech HiFace for SPDIF.
Phones include the LCD-2, Rastapants T50rp mod, and JH16 IEMs.
Generally, I prefer to use upsampling to 88.2 for my CDs, but as the NFB-12 uses the Tenor chipset, I left the unit in 44.1 for USB. Unfortunately much of my HD music is in 88.2, so I can only get that via TOSLINK or SPDIF.
Thus, my methodology will be to test via USB at native 44.1 only, so as to compare the D12 and the NFB-12. I will then compare the two devices via TOSLINK and SPDIF at 88.2 and 96KHz.
Once this is done, I will output the DAC from the NFB-12 to the D12 (the NFB does not have an analog input, USB, Coax and TOSLINK only, another limitation of the NFB-12).
The Sound via USB
I very much have enjoyed my time with the NFB-12 so far.
On the excellent Kirov Orchestra performance of Prokofiev’s Romeo and Juliette, the interplay between the strings was beautifully rendered during “The Balcony Scene.” Cellos had real bit to them, and the violins had a very pleasing tone to them, though I felt a degree of air was absent that I’m used to from higher end gear, and the sound had a somewhat darker character than I’d expected.
Things became much more lively when I popped on Mozart’s Piano Trio. The sound was lively and crisp, the instruments well placed, and the sense of stage rather palpable. Piano was percussive, grain-free and with a lovely sense of decay and black-space between the notes.
Keith Jarrett’s Paris/London Testament, Part VIII in 24/96 simply had me shutting my eyes and having a good listen. This DAC/AMP is just great for Piano. Beautiful overtones, no grain.
Switching to female vocals and rock, on Cowboy Junkies title song “Something More Beside You,” Margo Timmins voice was smoothly articulated in a well defined space, again, without the grain that is so often an overlay on female vocals. The cymbals came through nice and brassy, without sounding like buckshot in a frying pan (again, the lack of grain).
The amp section is clearly way more powerful than I’ll ever need with phones. My low impedance JH16 was as loud as I want on “Low” gain at 9 o’clock, and the T50 and LCD-2 never felt the want for reserves. It seemed this amp would go way louder than sanity or OSHA standards would dictate. Volume pot tracking, btw, was perfect down to dead-silence, even on the IEM.
The unit does show some limitations on classical, which probably explain why the piano trio was simply more fun than the orchestra. In general, the treble feels a bit and lacking in extension, almost soft. At the other end of the spectrum, the bass section sounded a bit muddy, like the amp does not have quite the control I’d expect, given the low output impedance and how power rating. There’s power, but as Yoda says, “control, you must learn control!” For orchestral pieces, these limitations make the amp a bit dark, but not distractingly so. For rock, some tracks came off feeling a bit bass heavy, though again this was not to the point of being a distraction, and at the price point is hardly objectionable.
Comparing the NFB-12 and D12
In a sense, this is an unfair test. The power supply and power output of the NFB-12 almost assures it will sound better. At comparable price points, the D12 has to deliver a lot of functionality, including being self powered, and offering analog inputs, in about 1/12 the volume, and has to be designed to be a power miser, which usually means class A/B op amp operation instead of class A. So this is not meant to detract from the D12 or to be unfair, rather, it’s to take a comparison in performance of a great portable DAC/Amp relative to a budget desktop unit with a lot of high-end concepts in the design.
On Nine Horses’ “Snow Borne Sorrow” I listened repeatedly to “The Banality of Evil” and found the differences to be very apparent in the vocals and strummed guitar. The NFB-12 had a richer sound to it, while the D12 felt then. The NFB-12 simply conveyed a more robust sense of “presence” to the vocals, while the D12 added some grain in the top registers. Male and female vocals had a continuous structure of harmonic integrity, in other words the voice didn’t feel like the voice plus overtones, and it was very grain free. The strumming of the guitar came through a little more crisply on the D12, but it felt comparatively etched and a bit thin, lacking in the sense of body a guitar should have.
Comparing the Kirov Romeo and Juliet, the NFB-12 had a lusher and more spacious sound. Both units did a very respectable job presenting complex passages without congestion, but the NFB-12 had an edge in spatial delivery and in air around the instruments, and on occasion the D12 sounded strident on strings. When listening to cellos, the D12 has an element of grain to it that was extremely low on the NFB-12, though totally absent on the Burson (just to note this amp is excellent, but not a giant-killer).
On Younger Brother’s Vaccine, in Pound a Rhythm, the percussion in the D12 felt a bit thin and dry, and the treble somewhat etched in comparison the the NFB-12, which frankly rocked. The drums were more percussive and you could hear the strike of the sticks on the drumheads very clearly. The NFB-12 also threw a much bigger soundstage, and gave a real wild sonic ride on Safety in Numbers.
Conclusion so Far
This is not a giant killer, but at $240 at my door, it’s rather remarkable to get bulletproof construction and a fully discrete amp with audiophile-grade parts throughout. The overall impression was a highly competent performance via USB, and while a bit less open and slightly more grainy than the Burson, it was a distinct step-up from a D12. Sound was powerful, spatially well presented, with easy dynamics, good sense of air and black-space, and pretty low levels of grain (something that really bugs me).
When compared to the slightly more expensive, but portable iBasso D12, the NFB-12 was clearly more musical, vastly more powerful, and probably going to last years longer. If you need portability, of course, the D12 is a quality choice with excellent performance, but if portability is not a consideration, the less expensive NFB-12 is the clear winner on sound quality, and validates the notion that at a given price, desktop is likely to be higher performance than portable.
The bottom line is that this amp seems to be a ridiculously good value at the price point. It’s an entry into a real audiophile level of construction and performance at a bargain basement price.
Next Installment
In part 2, I’ll play with the digital filter settings, and hopefully compare the TOSLINK and SPDIF inputs. The unit offers a selection of filters, from 2X up to 8x (though only at 48K or 44.1K inputs), with differing filter functions, such as linear phase, soft knee, brickwall, and a “linear phase apodising filter.” Sounds scary, huh? The filter settings are done manually by removing the cover and using the little "easy-to-lose" plastic jumpers.
Installment After That…
In part 3, I’ll try to compare the variable amp outputs by driving the amp section of my Burson.
