Bricasti Design M1 D/A converter John Marks July 2013
If Stereophile has covered another product that has received as many upgrades in as brief a time as Bricasti's M1 DAC ($8595), I missed it. I first wrote about the M1 in my August 2011 "Fifth Element" column; John Atkinson's full review was the cover story of the February 2012 issue; in a Follow-Up (March 2012), I wrote about Bricasti's expansion of the M1's selectable filter options; and in another Follow-Up (September 2012), I reported that Bricasti had now included, via a firmware upgrade, so-called "minimum-phase" versions of M1's existing filters, digital phase inversion, and digital volume control. I also noted the forthcoming availability of options for USB input and remote control. With the exception of those last two options, all upgrades of the M1 have so far been free for registered owners.
Recently, Bricasti's Brian Zolner called to tell me that he urgently wanted to visit me so that he could replace the power supply that feeds the digital "housekeeping," or digital signal-routing, section of my M1 review sample. Bricasti has changed that subassembly from a digital switch-mode supply to an analog, linear power supply. In the photo, the old, switch-mode power supply is on the left; the new, linear supply, based on a custom-wound transformer, is on the right.
I'm not talking about the power supply that feeds the M1's DAC chipsthe circuits that convert digital signals to analog. The M1's DAC chips and analog output stages have always had linear power supplies. The housekeeping functions involve such things as selecting which digital input provides the signal, and which of the various firmware-resident digital filters is applied to it.
Seeing as the critical digital-clock functions and the digital-to-analog stages were already powered by linear supplies (as are the analog output stages that provide the voltage that drives the preamp or power amp), one might be tempted to think that swapping in a different power supply for digital housekeeping circuitry would make little audible difference. However, I found the difference arresting.
Zolner arrived with the new printed-circuit board and some hand tools. (Installing the new board requires no soldering, and is within the competence of a reasonably careful, averagely coordinated audiophile.) I invited my friend Bob Saglio over as a sanity check; Bob is usually skeptical about upgrades and tweaks.
I first played "Easy to Love," from Ella Fitzgerald's The Cole Porter Songbook, Volume Two (CD, Verve 821 990-2). All fine and good. Zolner installed the new PCB, pausing halfway so I could photograph the two boards together. He then buttoned up the M1 and hooked it up to my system.
I hit Play on the Parasound CD 1 that serves me as transport. At about the second phrase of the piano introduction, it became very apparent that the harmonies were, without question, cleaner. Even the piano's individual notes sounded cleaner.
The clarity of a single piano note is a multidimensional phenomenon. Except for those in the low bass (one string per note) and midbass (two strings per note), most piano notes are sounded by three strings together, more or less in unison but not absolutely. No matter how carefully tuned and regulated the piano, two of those three strings are going to be more congenial to each other in pitch, timing, or both. Those congenial two gang up on the helpless third string via destructive interference. That microscopic amount of "beating" is what causes the attack of a piano note comprising the sounds of three unison strings to rapidly decay to the point where the odd-man-out string drops out for lack of mutual reinforcement.
After that tipping point, the note's decay is slower, and its sustain is therefore longer. One would not want things to be any other way. This is a feature of the piano, not a bug. The complex behavior of the coupled-oscillator system formed by three unison strings is responsible for the fact that a loud attack at first decays quickly, which makes audible room for the note(s) that follow. Were this not the case, the piano would sound as if the sustain pedal were down all the time.
Compared to the new analog power supply, the original switch-mode power supply had been making something of a hash of the M1's presentation of the piano in "Easy to Love," both chords and single notes. With the new power supply, the complex interactions of all those piano strings sounded a lot more natural, even (dare I say it?) analog-like. Bob Saglio agreed.
I'd expected to hear some improvement with the new supply, but only with attentive listening. I'd expected nothing like the unquestionably increased clarity I did hear. What I was truly unprepared for was the tighter bass. In this regard, I think "Easy to Love," a mono recording from 1956, is a great evaluation track, precisely because its bass is on the verge of being a bit out of control, unfocused and woolly. Sometimes a recording that's not all that great sonically can be more helpful in hearing differences than an audiophile reference recording.
Other CDs followed, with similar results. Aaron Diehl's The Bespoke Man's Narrative (CD, Mack Avenue MCD 1066), which I recommended in my June column, was more coherent. The biggest jolt came from the opening bars of the first track of Joel Frederiksen and Ensemble Phoenix Munich's Requiem for a Pink Moon: An Elizabethan Tribute to Nick Drake (CD, Harmonia Mundi HMC 902111)the lute sounded so much more like a real, live lute that it was shocking. Even an old standby, Time for Love: The Best of Julie London (CD, Rhino R2 70737), got the Windex treatment.
I don't have space here to cover in depth what this meant for the M1's filters, but the new power supply made me reconsider my preferences among them. Minimum 0 has supplanted Minimum 2 as my default filter, and the range of the filters themselves seems to have been expanded. The Pavel Haas Quartet's recording of Dvorák's "American" quartet (CD, Supraphon SU 4038-2), which previously had sounded best with Minimum 2, now sounded better with Minimum 4. Brian Zolner: "Getting rid of that noise lets us hear what the filters really are."
Zolner believes that, with the new board, we were hearing the absence of the digital power supply's switching noise, which previously had been leaking into other sections: "You wouldn't think that it would sneak its way [into the signal path], but it does." He said that the M1 measures the same with either power supply.
I have a bit of a problem with the idea that removing some noise can result in a perception of more bass, as well as the increased overall clarity one would expect. Zolner replied that the bass was always there, it was just veiled. I feel a bit chagrined that I heard the veiling only after it had been removed.
Bricasti must be trying to win some kind of award for exceptional customer service. The V2 power-supply hardware upgrade is offered to all registered owners at Bricasti's internal costs of parts and labor, which Zolner says come to under $200; and all M1s made since the beginning of March 2013 include the upgrade with no increase in price.
Bricasti is now working on a firmware upgrade to the M1 that will implement a DSD-via-USB protocol, as well as a 200Wpc monoblock power amplifier. They expect to display both in October, at the Rocky Mountain Audio Fest, in Denver.John Marks