Musical Fidelity V-DAC II D/A processor Measurements

Sidebar 2: Measurements

To measure the V-DAC II (serial no. SDY0367), I used Stereophile's loan sample of the top-of-the-line Audio Precision SYS2722 system (see www.ap.com and the January 2008 "As We See It"; for some tests, I also used my vintage Audio Precision System One Dual Domain analyzer.

Unlike the original V-DAC's S/PDIF input, which was limited to 96kHz, the V-DAC II's input operates at sample rates from 32kHz up to a maximum of 192kHz. The Macintosh USB Prober utility identified the product as the "Musical Fidelity V-DAC 24/96," but no serial number. USB Prober confirmed that the V-DAC II did operate in isochronous asynchronous mode, as claimed, and that the USB input would handle 16- and 24-bit data with sample rates of 32, 44.1, 48, 88.2, and 96kHz.

The maximum output level from both of the V-DAC II's single-ended outputs was 2.24V, 1dB higher than the CD standard's 2V and the original V-DAC's 2.08V. If not compensated for, this difference will work in favor of the new products in side-by-side comparisons with the Mk.1 V-DAC. The Musical Fidelity preserved absolute polarity (ie, was non-inverting). Its output impedance was higher than the original version's, but was still low at 99 ohms at high and mid frequencies, and at 124 ohms in the low bass.

Fig.1 shows the V-DAC II's frequency response with sample rates ranging from 32kHz (yellow and gray traces) to 192kHz (blue and red). The response at each sample rate follows the same basic pattern: a gentle rolloff followed by a sharp drop-off just below the Nyquist frequency (ie, half the sample rate). Unlike the original version of Musical Fidelity's more expensive M1DAC, the response with 192kHz sampling does extend an octave higher in frequency than at 96kHz. Channel separation (fig.2) was superb, at >110dB in the treble and midrange, but did decrease in the bass, reaching 100dB at 20Hz. This is presumably due to the increasing source impedance in this region of the MF's tiny wall-wart power supply.

Fig.1 Musical Fidelity V-DAC II, frequency response at –12dBFS into 100k ohms with data at 32kHz (left channel yellow, right gray), 44.1kHz (left green, right blue), 96kHz (left cyan, right magenta), and 192kHz (left blue, right red) (1dB/vertical div.).

Fig.2 Musical Fidelity V-DAC II, channel separation (5dB/vertical div.)

Whether measured using a swept 1/3-octave bandpass filter (fig.3) or with a modern FFT technique (fig.4), the increase in bit depth from 16 to 24 resulted in a 20dB reduction in the level of the V-DAC II's noise floor, suggesting that this inexpensive device has a resolution of around 19 bits. However, the lowering of the noise with 24-bit data (blue and red traces) does unmask some static tones just under 8kHz. Admittedly, these are at a very low level, but their presence does suggest some mathematical limitations in the Musical Fidelity's digital signal processing. Fig.4 was taken with S/PDIF data; repeating the test with USB data gave an identical result, confirming that the V-DAC II correctly handled 24-bit data via its USB input.

Fig.3 Musical Fidelity V-DAC II, 1/3-octave spectrum with noise and spuriae of dithered 1kHz tone at –90dBFS, with 16-bit data (top) and 24-bit data (bottom) (right channel dashed).

Fig.4 Musical Fidelity V-DAC II, FFT-derived spectrum with noise and spuriae of dithered 1kHz tone at –90dBFS, with 16-bit data (left channel cyan, right magenta) and 24-bit data (left channel blue, right red).

Unlike some processors that use a sample-rate–converter chip as a jitter filter, the V-DAC II suffered very little from noise modulation. Fig.5 shows the spectra of its output while it reproduced a 1kHz tone at 0, –60, and –90dBFS. While there is a slight rise in the noise floor as the signal level increases, this is mild, though AC-supply components at or below a still-very-low –130dBFS can be seen in the left channel. With its low noise and very low linearity error (fig.6), the V-DAC II's reproduction of an undithered 16-bit tone at exactly –90dBFS was essentially perfect (fig.7). Not only are the three DC voltages representing this signal clearly resolved, so is the Gibbs Phenomenon "ringing" on the tops and bottoms of the waveform. This waveform also indicates that the V-DAC uses a conventional time-symmetrical FIR reconstruction filter rather than, for example, a minimum-phase or apodizing filter that, all things being equal, tend to give a sound that listeners prefer. Extending the bit depth to 24 gave a well-defined sinewave (fig.8), despite the very low signal level.

Fig.5 Musical Fidelity V-DAC II, FFT-derived spectrum with noise and spuriae of dithered 1kHz tone at 0dBFS (left channel blue, right red), –60dBFS (left green, right gray), and –90dBFS (left cyan, right magenta).

Fig.6 Musical Fidelity V-DAC II, linearity error, 16-bit data (2dB/vertical div.)

Fig.7 Musical Fidelity V-DAC II, waveform of undithered 1kHz sinewave at –90.31dBFS, 16-bit data (left channel blue, right red).

Fig.8 Musical Fidelity V-DAC II, waveform of undithered 1kHz sinewave at –90.31dBFS, 24-bit data (left channel blue, right red).

Although the V-DAC II did produce some high-order harmonics with a full-scale signal, these were all very low in level, even with the extreme 600 ohm load (fig.9). The subjectively innocuous second harmonic was the highest in level—more so in the left channel (blue) than the right (red)—but was still inconsequential in absolute terms. The picture was similarly excellent on the demanding high-frequency intermodulation test (fig.10), with the second-order difference component lying at –96dBFS (0.0015%).

Fig.9 Musical Fidelity V-DAC II, spectrum of 50Hz sinewave, DC–1kHz, at 0dBFS into 600 ohms (left channel blue, right red; linear frequency scale).

Fig.10 Musical Fidelity V-DAC II, HF intermodulation spectrum, DC–24kHz, 19+20kHz at 0dBFS into 100k ohms (left channel blue, right red; linear frequency scale).

The V-DAC II effectively rejected word-clock jitter on all its inputs. With 16-bit data (not shown) there was no accentuation of the odd-order harmonics of the J-Test's low-frequency squarewave. No jitter-related sidebands can be seen with 24-bit data via the S/PDIF or USB input (fig.11), though the idle tones just below 8kHz in fig.2 that I mentioned above are present.

Fig.11 Musical Fidelity V-DAC II, high-resolution jitter spectrum of analog output signal, 11.025kHz at –6dBFS, sampled at 44.1kHz with LSB toggled at 229Hz: 24-bit USB data (left channel blue, right red). Center frequency of trace, 11.025kHz; frequency range, ±3.5kHz.

The Musical Fidelity V-DAC II may be affordably priced, but its measured performance is almost beyond reproach. Extraordinary!—John Atkinson

COMPANY INFO
Musical Fidelity Limited
US distributor: Tempo Distribution LLC
PO Box 541443
Waltham, MA 02454-1443
(617) 314-9227
ARTICLE CONTENTS

COMMENTS
FSonicSmith's picture

JA-your review of this MF DAC could not come at a better time. DACs are a hot product category with more offerings in all price ranges than ever before. Being a headphone enthusiast, I can tell you that among young burgeoning audiophiles worldwide (particularly space-craved European and Asian enthusiasts who frequently use computers and headphones in bedroom or tiny apartment rigs), DACs are a constant topic of discussion on headphone oriented audio sites (I would name the most trafficed one but am not sure that is permitted in this forum).

I LOVE this review because here we have a no-nonsense product from a very well respected manufacturer that has "measurements beyond reproach" and yet the sound is respectable, but not state of the art. It is "Exhibit A" to that old but absolutely undeniable chestnut that "If the sound quaility and the measurements don't coincide, you must not be measuring the things that matter the most". From time to time Stereophile has published long essays on a particular topic of facination i.e. cartridge alignment theory. I look forward to the day when someone as well versed in independant engineering as you, dear JA, undertakes a similar essay and takes a very hard look at what sets apart the state of the art DACs from the merely respectable in terms of design and implementation. Perhaps some input from someone like Gordon Rankin, whom I gather is one of this world's leading software engineers and experts on digital audio would assist in the essay. Is the size and sophisticatioin of the power supply and it's isolation from the digital components a fundamental aspect? Are operating temperature control (the so-called oven) for the DAC and clocking circuitry critical? Is the recreation or simulation of "real" (as in naturally rounded" soundwaves rather than the notched one we see via scope a critical goal? Where are the current engineering short-comings that need to be overtaken before digital recording and playback can have an absolute break-through in SQ? Just my thoughts, likely to be worth 2 sense and not more.

Archimago's picture

Nice review.

I found the opening comments curious however.  Having bought hi-rez downloads from various sources, I think it's obvious which hi-res format *should* win:

FLAC format - open source, easy to decode, capable of multichannel, good compression (in my tests better than Apple's ALAC). Apple's refusal to include native support in iTunes after all these years is an example of why I am reluctant to buy Apple products.

No DRM. Since you'd likely want to convert to lower-res versions for iPod/phone/tablet without hassle.

I prefer PCM over DSD. After all these years, DSD has not gained much of a foothold - I think there is a reason despite claims for the press and manufacturers like Sony. The standard DSD64 just doesn't sound all that much better and most albums have to be mixed in PCM anyways in the studio.

I really hope the whole affair around hi-res gets standardized so more people can enjoy the benefits of higher quality digital.

 

wozwoz's picture

i) The article seems a little johnny-come-lately to me. I mostly purchase hi-rez stereo on SACD, which, when done with a proper hi-res DSD recording (without PCM messing about in the editing), yields the most natural, warm analog sound I have heard ... rather like vinyl without the pops and scratches. There do exist D/A converters that can play DSD files, but the above reviewed item does not appear to be one of them.

 

ii) In any event, there are vastly more hi-rez files available on SACD (about 8000) than available as downloads, ... and most of the download files are ripped from SACD anyway, even if they started life as 24bit/96kHz files. 

 

iii) Finally, I am confused at the articles pointed assertions NOT to buy downloads. I personally do not ... I buy discs because I like the physical touch, I like the hi-rez stereo DSD format, and I think downloads are a waste of money because they have zero re-sale value. But, if this product is designed for downloads, and the article advocates that people NOT buy them ... then we are EITHER in the bootleg piracy world (where who knows what you are actually downloading anyway) OR the source is the hi-rez physical world (i.e. SACD) ... and if so, the SACD will yield better quality than an imperfect analog rip of same. 

Jon Iverson's picture

I strongly disagree with Sam's anti-download stance here and encourage readers with digital music servers to explore the dozens of high resolution download options now available.

It is a trivial matter at this point to convert file formats as things change.

Check out AudioStream for a list:

http://www.audiostream.com/content/hd-music-download-sites

Stephen Scharf's picture

Jon, I agree completely and have been buying more and more true high-resolution content from various providers e.g. HDTracks and Linn. I find them to be markedly better than their Redbook equivalents  (try the 24/192 version from HDTracks of Ella Fitzgerald's Clap Hands, Here Comes Charlie! for a revelatory experience). I find Sam's perspective on this rather curmudgeonly, personally, 

Also, I'm personally a bit tired of seeing repeated Musical Fidelity DAC reviews....and no review of game changers like the Schiit Bifrost. For the same price as a the V-DAC you can get a DAC with a built-in power supply and superior industrial design that you can use a with a real power cord. And for $100 more, you can get that DAC fully upgradeable with asynchronous USB up to 24/192 KHz performance. I owned a Bifrost for the better part of 4 months (before getting my Wadia 121), and IMHO, it blows everything in the $500-$1000 price class completely into the weeds. I've listened to the Musical Fidelity V-DACs and the M1(?) DACs, and agree with JA about being too dry for my taste as well. Once fully burned in, the Schiit Bifrost is musical, sweet and mellifluous, and anything but dry. It does require a good 250-300 hours of burn in, though (approx 3 mos. of use). I think Jason Stoddard and Mike Moffat at Schiit are genuine movers in this computer audio/DAC arena, though, and I think two years from now their products will be very highly regarded. Wish I was seeing their products reviewed in Stereophile...

Jon Iverson's picture

I've got a Bifrost review in the works right now.

Stephen Scharf's picture

Excellent! 

Thanks, Jon, I'll look forward to it.

 

SpinMark3313's picture

The DAC on my Marantz SA-8004 is quite good, but a bit more "hi-fi" sounding than the V-DAC II.  The II helps tame the digital nasties on some older CD's in particular -- worthwhile music but often poor recording / mastering.

Progressive rock, large scale classical = Marantz onboard DAC.  Small combo jazz / chamber music = MF V-DAC II.  My preference anyway...

AA's Pangea power supply for the V-DAC II is well worth the extra shekels by the by...

attilahun's picture

Did Sam say the V Dac II handles 192 and John say 96?

I would prefer 192 and feel 96 will be obsolete sooner as I'm a fan of hi-res audio. 

John Atkinson's picture

Quote:
Did Sam say the V Dac II handles 192 and John say 96?

The VDAC II handles sample rates up 192kHz via S/PDIF (coaxial or Toslink) but only up to 96kHz via USB.

John Atkinson

Editor, Stereophile

attilahun's picture

Thanks for the info John. 

There seems to be a consistent resolution drop in many USB inputs. 

I'm aggravated by the 48 limitation on the USB input on my bryston bda1. 

Hopefully these will slowly fade away. 

I've heard rumors that bryston is exploring a USB upgrade but no solid news.

earwaxxer's picture

I dont know - it seems to me with computer audio we have come to the point with digital music playback where we can do with it what we wish, and we dont have to settle for the set in stone hardware manipulations of the past. From my experience, SRC algorithms each have a different sound, with many methods to tweak phase, dither etc. That is the future and present of getting the most from redbook. Forget about letting the chip do the work.

jim777's picture

You have the choice of sending your 44.1k redbook to the dac and living with its upsampling (with the advantage that it might retain more bitlength than 24bits internally, or be optimized for the type of sigma-delta conversion that is done), or you can choose your SRC algo and send hi-res (already upsampled) data to the DAC. So I'm not sure that your concern is valid.

X