Guitar Sings as DSD DACs Shoot it Out
If ever an audio event were in living color, it was the Bay Area Audiophile Society’s (BAAS) February 23 live / DSD event at Blue Coast Studios in Belmont, CA. Jointly hosted by Blue Coast Records founder Cookie Marenco, a long-time proponent of DSD who has engineered or produced five Grammy nominated records, and the dedicated president of BAAS, Bob Walters, the two-session event prefaced comparisons of four DSD-capable DACs at three different price points with intimate, live-to-DSD recording sessions with Flamenco guitarist Jason McGuire.
I happily sat in the sweet spot, just a few yards in front of McGuire, in the first session. He in turn was seated between the Sony AR1 loudspeakers used for playback. From this delicious vantage point, I could evaluate the four DACs with the sound of the “live reference” in my head.
McGuire is anything but a one-song artist. With over 30 years experience performing and teaching, he is a major exponent of Flamenco, and can easily move between different styles and improvise while remaining within the classical form.
At close range, McGuire’s playing was astounding. I often think of the acoustic guitar as a small-sounding instrument, but on its own, in a room that could not have held far more than 20 people, it was a mighty force. If I was struck by the depth of McGuire’s sound, and the length of its natural decay within the guitar’s resonant body, I was literally stunned by the contrast between the softness of his touch in more ruminative passages and the brutality of his rapid strums on multiple strings and alarming raps on the guitar’s body. I could feel the fire and passion in his playing, and imagine how vividly his renowned Flamenco dancer wife converts that fire into movement.
At the same time, I wondered if any recording and sound system combo could together convey the in-your-face realism of this event without resorting to brittleness or squashing dynamics. The next time someone talks to me about system speed and slam, I’ll tell them to listen to a flamenco guitarist up close.
Marenco, who recorded McGuire’s performances in 128x DSD for imminent release on her site, had to adjust her mikes to compensate for the 20 extra bodies in the room. Comprising her recording chain were a Neumann U87 microphone (top) to a Millennia preamp, and two Neumann KM184s in stereo configuration (below) to Bybee Bullets to Neve preamps. These led to Ed Meitner AD/DAs and then to a Sonoma Recording System. Everything was wired with Marenco’s hand-made cabling, sourced from Jean Marie Reynaud silver/copper alloy cable.
After McGuire came the DAC shootout. Playing files on a PC Laptop using JRiver Media v18 feeding the Sony speakers via a Pass Labs X-600 amplifier, HiDiamond Reference USB cable supplied by Worldwide Wholesales, Marenco’s custom cabling, generic power cables, and a Meitner speaker switcher, we compared the sound of four units:
• Teac UD-501 DSD 5.6MHz / PCM 384 dual mono DAC ($850)All cabling and connections were identical, and equipment supports, vibration isolation devices, and other essential components equally non-existent. Room treatment was courtesy of Marenco’s garden. What was not equal were the playback levels. This was an all-volunteer effort, and there simply wasn’t enough time during set-up to calibrate everything. As if calibration were easy, given the dynamic nature of some of the tracks, and the varying abilities of the DACs to convey those dynamics.
• Mytek Stereo192-DSD USB/FW DAC ($1,595)
• Benchmark DAC2 HGC ($1,995)
• Playback Designs MPD-5 DSD 6.1 / PCM 384 DAC ($13,000)
Note as well that no attempt was made to set filters, upsampling, or other such variables on these units. Many of these DACs might have performed quite differently had time been available to adjust settings and tune them to the playback chain and situation.
Walters came prepared with files of six tracks, in resolutions ranging from redbook CD up to 24/384 and 64x/128x DSD. Our announced goals were to determine if the DSD tracks sounded different, let alone superior, and what equipment was necessary to extract the best from the format. In actuality, the session revealed which DACs were most capable of extracting the most information from any format.
After auditioning virtually all the files on the Teac – this took a long time, given that we were attempting to sink into the music rather than resorting to evaluating short sound-bites disguised as music – we trimmed the playlist. By the time we had finished with the Mytek, we had pared our selections down to Marcia Seebaran singing Billy Strayhorn’s “Lush Life” on a free track from Audio Jazz Prologue III, Jane Selkye singing “Slow Day” from the Blue Coast Records Collection, the first seven minutes of the first movement of Mahler Symphony No. 2 (Budapest/Fischer) on Channel Classics; and brilliant jazz pianist Vijay Iyer’s solo, “Human Nature,” from Solo on ACT, which Marenco recorded.
Cut to the chase. Without regurgitating all my copious notes on each track in each format, it is fair to tell you this. To my ears, the Teac UD-501, which seems to be optimized for DSD, sounded absolutely flat and lifeless on 16/44, and first began to transmit the body, substance, and presence of music on 24/384. DSD was another story, producing tighter bass, a more clearly defined leading edge, and more natural bite on the sound of horns, guitar, and piano. I would not be surprised if this DAC performs far better on all formats when its filters and settings are optimized.
Listening to the Mytek Stereo192-DSD DAC with DSD files of the Mahler 2, it was immediately apparent that the unit yielded richer bass, greater detail, and more hall sound. I know this track very well – I’ve been carrying the hybrid SACD room-to-room at shows for years as I’ve blogged for Stereophile – and still failed to hear much of the bottom line produced by basses and the lowest pitched percussion. But the DAC was very good. When listening to Vijay Iyer in 24/96 rather than DSD, I still heard for the first time some of the piano’s bell-like tone that was absent from the Teac’s presentation. This was also the first time that I liked jazz vocalist Marcia Seebaran, who had sounded very second rate on the Teac. An outlay of $750 more, IMHO, yields far more musical information, at least when the settings on all four DACs are in default mode.
The Benchmark DAC2 HGC was a mind-blower. Although questions remain as to whether some of its perceived superiority was due to a treble boost that some found worrisome, it produced, on all tracks we auditioned, richer bass, a much bigger soundstage, and greater height.
In DSD, Seebaran sounded more intimate and immediate than ever before. I could hear, for the first time, the reverberation surrounding her voice. I could also appreciate, for the first time, her voice’s warmth, sweetness, and expressive power. She emerged, at last, as an artist whose impact, in part, depends upon subtle variations in volume and nuance that lesser DACs cannot convey.
The Mahler 2, as well, was a revelation on the Benchmark. In DSD, it completely left the speakers behind, with an expansive soundstage that dwarfed them in every respect.
After we rebooted to deal with the Playback Design MPD-5’s proprietary interface, I discovered that the Mahler in DSD was set even farther back behind the speakers. It sounded smoother, with noticeably greater dynamics and a firmer, far more audible, and richer-sounding bottom bass line. The timbre of brass also sounded far more natural than with the other DACs. This superiority also extended to the bass accompanying Seebaran. With all selections, the additional natural warmth and greater body of instruments and voice, increased color saturation, organic (as opposed to hyped) leading edge of sounds, and the better integration of the leading edge with the body and undertones of the sonic envelope produced the most musical presentations of the session short of McGuire’s stunning performance.
At the end, we briefly returned to the Teac. I felt as if a piece of cardboard had been placed between the loudspeakers and me. That’s how much smaller, diminished, and damped everything felt with the TEAC set to default mode.
I should note that, in the second session, Walters began with the Playback Designs DAC and went down the chain. He also unintentionally tested people’s sonic memories by playing longer selections. I wasn’t present, but it seems that opinions varied widely as to which of the three top DACs sounded the best.
The second session also had six women in attendance (plus Marenco). Comprising about 30% of the attendees, this was a new record for BAAS. Their reaction to the extended DAC comparisons was, in the paraphrased words of one attendee who generated nods from her fellow females, “Why don’t you guys just listen to the music? It’s great.” My husband would have nodded along with them, if he hadn’t just plain nodded off out of boredom.
At session’s end, someone called for a blind listening test. Half the people present said it would be a waste of time, but Walters proceeded. Of the 10 who stuck with it, the only person to correctly guess the identity of the four DACs was seated in the worst listening position. Which goes to prove that roses are red and violets are blue, but in a blind world, nothing is true.