Benchmark DAC1 USB D/A processor & headphone amplifier Page 2
Starting off with the DAC1 USB used as the primary digital source in my system, the sound fed via AES/EBU or TosLink was very appealing. The low frequencies had both weight and definition, while the highs were more silky-smooth than I remembered from my original DAC1. I have been playing Alison Krauss and Robert Plant's Raising Sand collaboration (CD, Rounder 11661 9075-2) almost nonstop since I bought it at our local Starbucks. The Benchmark opened a clear window on T Bone Burnett's rather dark-sounding mix, and while the two voices blend much better than I would ever have expected, the threads could still be easily distinguished through the DAC1, due to its impressive resolving power. Similarly, the subtle but real-sounding room reverb on this issue's "Recording of the Month," Neil Young's Chrome Dreams II (CD, Reprise), was reproduced with a lack of ambiguity.
Fed high-resolution data, such as the DVD version of the Beatles' Love remixes (Capitol 9463-79810-2), with which the Ayre player will output a 96kHz stream, such low-level details as the birds in the background of "Because" were clearly outlined without the high frequencies being thrust forward at the listener. The 24-bit/96kHz FLAC download of Linn Records' recording of the Mozart Requiem, with Sir Charles Mackerras conducting the Scottish Symphony Orchestra (also released on SACD as Linn CKD211)—sent to my Slim Devices Transporter from the Mac mini that serves music to our household via WiFi, and sent in turn to the Benchmark via the AudioQuest optical S/PDIF connection—was reproduced with a rich, broad sweep of sound, yet without any smearing or obscuring of fine detail.
As well as standalone auditioning in my main system, I compared the Benchmark DAC1 USB both with the original version of the Benchmark DAC1 that I'd bought following my 2004 review, and with the Bel Canto e.One DAC3 ($2495), which I wrote about last November. To permit instantaneous comparisons, I fed the Ayre C-5xe's AES/EBU output first to the Mark Levinson No.30.6 DAC, then from that processor's two AES/EBU data outputs to the two DACs under test using identical lengths of DH Labs AES/EBU cable. Levels were matched to within 0.05dB at 1kHz by using both processors in fixed-output mode and using the Input Offset function on the Mark Levinson No.380S preamplifier.
My original Benchmark DAC1 has well-extended, well-defined lows, a natural midrange, and clean highs. But the newer Benchmark had smoother highs, with less grain. Considering the fact that the earlier version has identical DAC and output circuits—other than the use of Philips 5532 op-amps in the 2003 sample compared with Texas Instruments 5532s in the 2007 sample—it was hard to see why this should be the case. Still, it was a consistent factor in my auditioning of the two Benchmarks.
Against the Bel Canto DAC3, it was extremely difficult to hear any differences with rapid switching. But listening to longer passages of music through each DAC, two distinct characters emerged. With Raising Sand, for example, the Bel Canto had a very slightly smoother low treble and a touch more lower-midrange bloom, the Benchmark slightly more "bite" and less soft-sounding lows, which benefited the double bass on this CD. I noted in my review of the DAC3 that the string players on the Mozart Flute Quartet movement on Editor's Choice (CD, Stereophile STPH016-2) were set farther back in the stage than I was expecting. Via the Benchmark, the musicians moved forward a little in the soundstage, although, paradoxically, the reverberation tails were just as clearly delineated.
In the cans
Turning to the Benchmark's sound via its USB connection, I auditioned it exclusively using a pair of Sennheiser HD650 headphones. Playing the 24-bit master files for my Stereophile recordings, everything seemed fine. But with iTunes running on my PowerBook as the source of 16-bit music files, the sound seemed a little grainier than I was expecting from my experience using S/PDIF and AES/EBU sources. I stuck with it, but the more I listened, the less satisfactory I found the sound.
I normally leave the measurements until after I have finished my auditioning, but it looked as if something was wrong. And something was, as you can read in the "Measurements" sidebar. My sample of the DAC1 USB appears to go deaf below –67dBFS when driven by my laptop. I listened to the 500Hz fade-to-noise-with-dither track from the CBS Test CD, ripped to my PowerBook, and the tone first became very dirty-sounding, then vanished after about 5 seconds from the start of the fade. As I noted in the measurements, the low-level deafness didn't happen when there was also high-level data present. I mixed the fade-to-noise signal with a 19kHz tone at –10dBFS. As I can't hear this high, it wouldn't mask the low-level tone. And I could indeed now hear the fade cleanly dropping in level, as it should.
I switched to my Mac mini. Now everything was fine: 16-bit music sounded identical to how it had done with AES/EBU and, switching in the 10dB extra gain on the DAC1's headphone output, I could follow the tone on the fade-to-noise track down to inaudibility even without the high-frequency, high-level tone present.
Given the problems I had using the Benchmark to play back 16-bit files from my PowerBook and from my PC via the USB connection, this inevitably will have to be a two-part review. But used via USB from my Mac mini and from its conventional data inputs, the DAC1 USB is a superb-sounding device. Its balance is slightly more forward than the twice-the-price Bel Canto DAC3, but that will make it a better choice for systems in which the DAC3 sounds too laid-back. And the fact that it has an excellent-sounding headphone output with two choices of maximum gain makes it a bargain. Very highly recommended.