Monthly Archives: May 2014

When seeking to convey what the DirectStream achieves with high resolution source material you can never afford to ignore the garbage-in garbage-out principal.  High resolution implies that we are talking about source material which rises above and beyond the performance delivered by lower-resolution source material, by which we must assume in turn we are referring to Red Book CD program material.  Extending this argument further, it means that we are obliged to confine ourselves to source material of impeccable pedigree.  It also follows that, unless we spend most of our time listening to DACs of the highest possible pedigree (which I have at times had the opportunity to do), when we hear exceptional playback of an exceptional recording we generally don’t have a context in which to place a qualitative assessment of what we hear.  So in my description of what follows, bear that in mind.

First up, “Quiet Winter Night”, a folk-jazz album by the Hoff Ensemble, from Morton Lindberg at 2L in Norway.  This is a 24-bit 352.8kHz (so-called DXD) recording, which is the format in which the original recording was made.  2L makes all of their recordings available in this format if that is what you want.  I know of nowhere else where you can purchase the entire catalog of a no-compromise audiophile label in what is truly the Original Studio Master format.  They also make available a huge selection of free downloads in a bewildering array of formats, including multi-channel, DSD, and original studio master DXD.  Since the DirectStream does not support 352.8kHz PCM, BitPerfect downsamples it to 176.4kHz.  The sound is effortless and spacious, having been made in the reverberant soundstage of a small church in Oslo.  The instruments are all acoustic, apart from a Strat, and are grouped tightly together.  The recording has a “live” feel, although there is no audience.  It has a slightly bass-heavy ambience, which to my ears is typical of 2L.  Perhaps bass-heavy is wrong.  Maybe I mean bass-weighty.  Because the microphones are very close to the instruments, and there is little in the way of studio adulteration going on, I think they naturally pick up more of the “weight” of the bass drum and double bass than you would normally hear from a typical listening position.  But in any case the DirectStream does a terrific job of presenting the individual instruments and voices, locating them precisely in space.  The percussion in particular is presented with a lightness of touch and clarity of texture, but with appropriate weight in the bass.

Résonance” is a recording of solo Viola Da Gamba music by Nima Ben David, from Todd Garfinkel’s MA Recordings.  I believe the original master recording was made in DSD, and what I have is a 24-bit 176.4kHz conversion.  This recording is quite honestly like being in an intimate highly reverberant studio space with the performer.  Ms Ben David is right in front of you, so much so that getting the volume absolutely correct within 0.5dB is crucial to making it sound right.  The scrape of the bow across the strings is captured cleanly.  The tonality of the instrument is wonderfully present, and, try as I might to listen analytically to the sound, I find myself being drawn inexorably into the music.  The Viola Da Gamba is a predecessor of the ‘cello but with a sound which is at once less refined, with some of the weighty tone of a double-bass, but in the hands of Ms Ben David ethereally expressive.  And in the hands of the DirectStream, there is little to want for in the sound.  The DirectStream’s magical bass was made for music like this.

Meet Me In London” by Antonio Forcione & Sabina Sciubba, produced back in 1998 by Naim Records, is probably the one “Audiophile” record that truly stands up on its own two feet as a musical experience.  It is a collection of covers and standards that showcase Sciubba’s poised and controlled vocal, with Forcione’s dexterous acoustic guitar work, backed up by delicious bass, and occasional other musicians as required.  The recording was magnificently captured by Naim, and remastered specially for the 24-bit 192kHz digital download which I have.  The DirectStream reproduces this recording beautifully, with the texture and dynamics of Forcione’s guitar work a particular joy to listen to.  The bass is wonderfully tasty, with real presence and weight.  Sciubba’s vocal is easy on the ear and shows remarkable restraint in terms of dynamic compression, but some compression is inevitably there, and the DirectStream makes it plain if you want to go to the trouble of listening for it.  I feel I want to use the word “presence” here again.  This is emerging as one of the core characteristics of the DirectStream – it has consistently great “presence”.

I have an odd track – I have no idea where it came from, or who the performers are – but it is a beauty.  It is an excerpt from Copland’s “Fanfare For The Common Man”, and was recorded by someone who had a wonderful idea:  “Why don’t we do a proper job of capturing those drums?”.  Both bass drum and tympani are prominently featured on this recording, and the DirectStream gives free rein to both, limited only by the bass response of my speakers which doesn’t go down much below 28Hz (and neither does my room, for that matter).  But when the bass drum strikes, it has an enthrallingly accurate and detectable pitch, something that I have previously only heard before on seriously good headphones, such as Tim’s Stax SR009s.  The bloom and decay are wonderfully accurate.

Moving on to DSD.  In terms of its specifications, what the DirectStream offers that its predecessors did not, is DSD support.  I have left this aspect until last, because I don’t want to over-emphasize its importance.  While DSD is at the core of how the DirectStream functions, I would not wish to suggest that it should only be of interest to those with a DSD collection.  Far from it – it is a high-performance DAC that happens to do DSD.  But its DSD performance still needs to be examined.

Let me review what I think of DSD at this point in time, recognizing that this a moving, evolving target.  For me, the bottom line is that the best sounds I have ever heard have been produced by DSD.  At BitPerfect we are quite involved in format conversions between DSD and PCM.  We have found that the very best DSD-to-PCM conversions we know of – those produced by our DSD Master product – can be very, very close to their DSD originals.  The differences are really quite subtle.  We can also do it the other way round – make conversions from PCM originals to DSD – but the best quality PCM-to-DSD conversions are still those made by professional third party software such as Weiss Saracon.

An interesting thing happens when we make PCM conversions from two types of DSD recordings.  The first type are original DSD recordings that have never seen PCM, while the second type are DSD conversions from a PCM master.  What we have found in the past is that PCM conversions made from “pure” DSD masters (i.e. the first type) are not quite as good as the DSD originals, whereas those made from “impure” DSD masters (i.e. the second type) can be much harder to distinguish from the DSD master.  This suggests that a certain “PCM-ness” may be imprinted upon a recording whenever it enters a PCM format, and that once having acquired this “PCM-ness”, it can never again regain what you might call its “DSD-ness”.  All very wishy-washy, I know, and not at all scientific (although I do have a rational basis for making that argument).  And given that all this was determined exclusively using the Light Harmonic Da Vinci Dual DAC, it is far from conclusive either.  But I wanted to mention it, because one of the things I want to accomplish with the DirectStream is to see if the same sort of behaviour can be evinced using it.  And just to be clear, I am only going to report preliminary findings in this post.  We’ll be doing a lot more work with the DirectStream in this area down the road, and may report on that later if I come up with anything interesting.

In comparing PCM and DSD versions of the same track, how can you avoid comparing Apples and Oranges at some point?  Unless you have access to detailed information concerning how the different versions were mastered, how do you know that what you are hearing is not ultimately a difference in the mastering that you end up ascribing to a component in the review system?  To attempt to minimize these issues, I set about using DSD Master to make a number of PCM conversions from a selection of DSD tracks that I have, and comparing what I heard from them.  That way, the only differences inherent in the source material are those arising from DSD Master’s ministrations, which, if nothing else, are at least known.  In each case the PCM versions were 16-bit 44.1kHz, 24-bit 88.2kHz, and 24-bit 176.4kHz, and all were in Apple Lossless format.  The DSD tracks were a mixture of “first-type” and “second-type” DSD64 recordings; mostly the latter since the true provenance of a DSD recording is actually desperately hard to establish conclusively.

I don’t have the time to go into details but the results were all remarkably consistent.  First of all, the differences – from CD all the way up to DSD – were much smaller than I expected them to be, and the most noticeable of these differences was moving up from CD to 24/88.2, which is probably due to the step up in bit depth.  With each upward step in format resolution the same two general things were observed to happen.  The bass got cleaner and firmer, and the detail resolution of instruments across the spectrum improved.  The soundstage imaging got crisper, and everything just seemed to acquire that extra smidgeon of presence.  However, the difference between 24/176.4 PCM and DSD was most often too difficult to reliably detect.  I suppose I should be making a big deal of that – how DSD Master produces PCM conversions which are close to indistinguishable from DSD – and yes these conversions are good, but I don’t think that’s the whole story here.

The bottom line is that the DirectStream produces exceptionally good sounds playing DSD material, but it also produces virtually indistinguishable sounds playing DSD Master’s PCM conversions of the same material.  If pushed to get off the fence here, I would suggest that the DirectStream’s DSD processing may be at the root of this.  Using the terminology I introduced earlier, it sounds to my ears as though the DSD processing in the DirectStream may be introducing “PCM-ness” into the DSD data stream.  We should not lose sight of the fact that there is a lot of signal processing going on in the DirectStream, just as there is in other DACs built using common commercial DAC chipsets.  Each of those chipsets uses its own proprietary implementations of those signal processing algorithms, and DirectStream merely has its own.  This signal processing is wickedly complex, and does not yield to a few pithy sentences.  It is also an area in which great progress will continue to be made, aided by the inexorable advancements which will continue to be made in signal processing technology.

In summary, the DirectStream is very nearly the best DAC that has ever passed through my system, yielding only to the Light Harmonic Da Vinci Dual DAC.  At $6,000 it is far from cheap.  But if you are willing to consider spending that kind of moolah, the DirectStream handily outperforms anything I have heard for anything short of seriously silly money.  And since PS Audio has a pretty comprehensive worldwide dealer network, the prospects of being able to audition one locally are better than for a lot of its serious competition.

As a kind of coda to this review, I have implied that the Light Harmonic Da Vinci Dual DAC is better than the DirectStream, and indeed it is.  I lived with one for the best part of a year.  But it comes in at $31,000 last time I looked.  What do you get for the extra $25,000?  Simply put, the Da Vinci places the performers right there in the room with you.  Sonic textures are so convincingly real.  If you can describe the DirectStream’s sense of presence as exceptional, then that of the Da Vinci is quite uncanny.  It can make you stop and turn your head when you are elsewhere in the house, which is something you have to experience to believe, and is a party trick the DirectStream cannot quite pull off.  It is possible, though, that the DirectStream’s bass might be even better than that of the Da Vinci.  Listening to DSD on the Da Vinci can be heartbreakingly good, just as the Linn Sondek LP12 was, in its way, when it burst on the scene nearly 40 years ago.  The Da Vinci Dual DAC illuminates differences between PCM and DSD which the DirectStream apparently cannot.  Granted, those differences probe deeply into the realm of diminishing returns.  So if a Da Vinci Dual DAC costs $31,000 how much is a DirectStream worth?  I would suggest that $6,000 sounds more and more like an absolute steal.

Read my follow-up review on the important 1.2.1 firmware update.

This component changed more dramatically, and more rapidly over the break-in period than any other component that has passed my way.  Admittedly, I was also breaking in its power cord and its balanced interconnects at the same time, but I have broken in many of those cables before and not noticed such a huge and rapid improvement.  Not that this means anything in and of itself, but knowing that the DirectStream has a transformer in its analog output stage, my left brain can’t help but speculate whether this might have been responsible in some large way.

When you listen to enough high end audio equipment, my first concern with a new product is usually to figure out where it fits in as regards the detail/fatigue dichotomy.  It has been my own experience that the more detailed a system the more likely it is to be fatiguing.  This applies to all aspects of the audio reproduction chain, from electronics to loudspeakers.  DACs are not excluded.  Only the very best of the very best seem able to beak this mould and allow a fatigue-free, detailed and dynamic listening experience.  By far the most dramatic example of this in my experience are the astonishing sounds produced by German manufacturer MBL.  Dan d’Agostino’s new Momentum amplifiers also fall into this category.  As does Light Harmonic’s Da Vinci Dual DAC.  But all of these items come at eye-watering prices.

The DirectStream’s sound is non-fatiguing in an almost tube-like way.  It is smooth and relaxing.  Surely, you tell yourself, it must be sacrificing detail to deliver this.  But no.  With recording after recording, the DirectStream presents layers and layers of detail in such a way as to produce a sound which is very easy to listen to, and with a low fatigue factor, even with CD sources.  As someone with very little tolerance for the fatiguing nature associated with CD, I find this to be a most appealing characteristic.

Another noticeable attribute of the DirectStream is its command of bass.  The DirectStream’s bass is deep and tight, tight, tight!  I can follow bass lines, and the texture and precise intonation of bass drum strokes all the way through complex pieces and performances.  Bass drum thwacks have a physical impact which you can feel in your chest.  Which is not to say that the bass is in any way prominent.  [Bass done right should sound bass light most of the time.  If the first thing you think of when you hear a system for the first time is “wow, listen to that bass!” then the chances are that its bass is bloated and overblown.]

DirectStream is also an imaging champ.  This is another performance parameter which is very high on my personal desirability list.  There are two thresholds which a system can cross when it comes to imaging.  The first is the basic one of throwing a deep and wide soundstage.  Many components can do this, and these days it is almost inexcusable to fail in this category.  The second one is harder to cross, and involves being able to locate the performer in a three-dimensional space not only with great precision, but with great stability.  In systems that fail to cross this second boundary, when you close your eyes and concentrate on individual performers, it can be maddeningly difficult to visualize that performer consistently in the audio space.  This level of performance places great burdens on all elements of an audio system.  In particular, this is the area in which cables – still the greatest bugaboo of the high-end – can make an absolutely critical contribution.  DirectStream is able to cross that second threshold, and, given appropriately recorded source material, and with the right ancillary supporting cast, throws a seriously stable, clearly delineated sonic image.

I said in yesterday’s post that I would place some special emphasis on CD quality playback as a result of Paul McGowan’s claims that the DirectStream especially enhances CD playback.  And so I shall, with the aid of some specific examples.  But as a general observation it is clear that the DirectStream presents CD-based program material with a smoothness, clarity, richness of tone, and dynamic excitement, that will surprise you, and with a bass performance that will stun you (if you are fortunate enough to have equipment capable of exploiting it, and have taken sufficient care in the set up of your listening room).  The best-recorded CDs can sound remarkably close to high-resolution downloads.  This is what Paul McGowan was talking about.

I’ll start with an “audiophile approved” recording which many of you will know:  Acoustic Live by Nils Lofgren.  It is a deeply satisfying album on both sonic and musical levels, and is only available digitally (to my knowledge) in Red Book format.  On the DirectStream Lofgren’s vocals emanate with admirable clarity, presence, and purity of tone from a Nils-Lofgren-sized space bang in the middle, and about two feet behind the line between my speakers.  The vocal affects an unusually emphasized lisp which, being familiar with his voice through a couple of his LPs that I own, I believe is being unnaturally emphasized by the microphones during the recording.  The DirectStream presents the vocal cleanly enough that I feel comfortable making those kind of assessments.  Lofgren’s aggressively picked acoustic guitar solos remain stably imaged even as the dynamics of the recording push towards the limits.  On “Black Books”, a pair of tom-toms play gently about 4 feet behind the voice and a foot or so to the right.  As Lofgren’s guitar solo picks up in complexity, volume and dynamic attack, and the atmospheric keyboard ambient gradually ramps up, it is easy to remain focused on the tom-toms.  This is a first class performance by the DirectStream.

Reference Recordings continue to make some of the finest quality CD recordings available.  Although most of them are now available as high-resolution PCM downloads, I still have a number of CDs that I bought over the years and have since ripped to disk.  One of those is Bruckner’s 9th Symphony by the Minnesota Orchestra conducted by Stanislaw Skrowaczewsky.  Bruckner’s symphonies are scored in a Wagnerian style, calling for plenty of deep, sonorous, highly resonant bass and baritone brass instruments, together with lush strings.  These recordings also capture a wonderful sound stage, which on the best systems can be the size of a concert hall.  Unfortunately, I have not yet got my listening room to the point where I can achieve that (and I’m not sure I ever will), but I do get something that approaches 20ft wide and 30ft deep if you can place yourself in the sweet spot.  The DirectStream reproduces the sonorities of the orchestra quite wonderfully.  It allows me to visualize the complex sound stage convincingly, and handles the most complex passages with aplomb.  Only during the very densest orchestral climaxes is this clarity compromised.  By comparison, on the Reference Recordings 24-bit 88.2kHz download of “Stravinsky” by the Minnesota Orchestra conducted by Eiji Oue, even with the volume raised to scary levels there is no onset of congestion during the loudest climaxes, and the DirectStream’s poise, balance, and control are impeccable.  When the powerful bass drums are struck during the Dance of King Kashchei you can easily discern the pitch of the drum and follow its bloom, decay and reverberation.  Not to mention feel it.

Another “Audiophile Classic” is Rebecca Pigeon’s “Spanish Harlem”.  I have a 16/44.1 version on Chesky’s “The Ultimate Demonstration Disk”, plus a 24/96 version on the downloaded album “Retrospective”.  The song begins with an upright bass line.  On the 24/96 version there is both air and weight to the bass, and an element of phrasing to the simple bass line.  All of these things are audibly diminished on the 16/44.1 version.  When the piano enters, it sounds like it has been placed at one end of a long cardboard tube with the microphone at the other end.  Strangely, this tonal defect is less immediately evident on the 16/44.1 version than on the 24/96 version.  Pigeon’s vocal has what sounds like an electronic reverb applied on the 24/96 version, and this sounds less somehow artificial on the 16/44.1 version.  The track progresses with the introduction of a pair of maracas about 30ft away towards the right, together with violins and guitar.  These are altogether more distinct on the 24/96 version, but start to coalesce slightly on the 16/44.1 version.  These differences are not nearly so stark using my Classé CP800.  What are we hearing here?  Is it the deficiencies in the playback of the DirectStream on 16/44.1 source material?  Or are we more clearly resolving the limitations of this CD’s performance?  It is surely the latter.  If the DirectStream was having trouble resolving 16/44.1 then how could it possibly do a better job with 24/96, unless the two signal formats were being digitally processed in an incompatible manner (which, from all I have been told, is not the case).

U2’s album “How To Dismantle An Atomic Bomb”, released in 2004 was roundly (and justifiably) panned by audiophiles for its massively compressed “loudness wars” sound.  According to the web site dr.loudness-wars.info, it has a DR (Dynamic Range) rating of 5 which is pretty bad (The Rebecca Pigeon 24/96 album, by contrast has a DR rating of 12, which is still only modest).  What will the DirectStream make of this?  The answer is that it sounds like a very bad MP3.  It is truly terrible, and quite frankly I could not listen to more than three tracks before I had to find something else.  The DirectStream cannot make a silk purse out of this particular sow’s ear.

How about an album that very few of you will have heard of?  Released in 1996, “Decksanddrumsandrockandroll” is an album of electronica by the British band Propellerheads.  I bought it exclusively for one track “On Her Majesty’s Secret Service” because I am a sucker for James Bond music.  OHMS is a terrific blast.  It can only be listened to at one volume setting – way too loud.  The deep, powerful, agile, and impactful synthesized bass line is tailor made to showcase the tightness of the DirectStream’s Stygian end.  I don’t know about foot tapping – it had me leaping about the room!  With the curtains closed, of course.  With electronic music, there is no such thing as any kind of sonic reference, so you can’t really say anything about how “natural” or “accurate” the sound might be.  All you can go on is how you respond to it.  Going back to the Classé CP800 this track becomes surprisingly pedestrian.

In terms of pure unadulterated macro dynamic detail, the ultimate test is the famous Telarc recording of Tchaikovsky’s 1812 Overture featuring real brass cannons captured with as little compression as technology permitted.  I have both the CD layer plus the DSD layer which was (with apologies to Shakespeare) “from its mother’s SACD untimely ripp’d”.  The DSD version plays back the cannon blasts with stunning impact.  Each of the 16 brass cannons has its own distinct sound, which come across very clearly.  One of them (the second-to-last one, if you’re following this) even sounds as though it has blown itself apart when it fired.  Listen for yourself!  How does the CD layer compare on the DirectStream?  There is no doubt that the cannons appear to be wrapped in cotton wool by comparison.  There is a loss of immediacy.  Also, it is slightly harder to listen through the cannon blasts and follow the underlying music, which includes a recording of the actual cathedral bells of St. Basil’s in Moscow (which is what Tchaikovsky specifically called for in his score!).  But if you had never heard the DSD version, you would undoubtedly have been seriously impressed by the CD version.

So, summarizing this part of the review, the Red Book (16/44.1) performance of the DirectStream is quite impeccable.  It is a massive step up in quality from my Classé CP800, and from every other DAC which has passed through my system, with the possible exception of the Light Harmonic Da Vinci, to which I will return before I finish.  The bottom line is that CD played on the DirectStream sounds better than hi-res played on some lesser (and cheaper) DACs which have passed through my system in recent months.  And not just in sound, but in character.  As I type this I am listening to the third movement of Shostakovich’s 5th symphony, played by the London Philharmonic Orchestra, conducted by Bernard Haitink (part of an awesome symphony cycle).  It is hard to believe this is plain old Red Book.

Click here, and I will conclude by focusing on the performance of the DirectStream on hi-res source material, including DSD.

At BitPerfect we needed a reference-quality DSD-compatible DAC, and, without mentioning names, the ones we had to hand were proving not be up to the standards we were hoping for.  To a certain extent, this is a self-fulfilling prophesy.  At the end of the day the differences between DSD and PCM, while real enough, are actually quite small.  On the amazing Da Vinci Dual DAC (which costs over $31,000) it certainly is subtle, although real enough that once you’ve heard it you want more of it.  So at the sane end of the price spectrum, where should one be setting realistic expectations?

We found out, eventually, that we weren’t going to find what we needed at the truly sane end of the price spectrum, so we set our sights a little higher.  Talking to Paul McGowan of PS Audio, we decided that their new DirectStream DAC, which was still in development at the time, was in all likelihood what we needed, so we committed to purchase one of the first production models.  I have been listening to it for the last week.  I’m listening to it as I type.  I’ve been doing little else for the last week.  I don’t normally do product reviews – I am not a professional reviewer – but I felt that in this instance the effort was probably justified.  Please read on.

The PS Audio DirectStream DAC is not cheap.  At $6,000 it is priced at a point which all of my friends would unhesitatingly label “insane”.  But in the world of high-end audio, the market to which it is targeted, it has the potential to be classified as a bargain.  Provided it delivers on the hype, that is.  C’mon, you already know it does, because otherwise I wouldn’t be writing this!  But let’s play along anyway, and pretend you haven’t already figured out that the butler did it.

The DirecstStream’s design eschews the conventional wisdom regarding the design of a high-end DAC, which says you go out and buy a chip-set from one of the established vendors such as ESS, TI, and Wolfson, and build yourself a DAC around it.  Rather like you would build a computer, having chosen an appropriate CPU from Intel’s catalog.  PS Audio has taken a different tack.  They have avoided using a DAC chip entirely, and have built their entire core converter functionality around a FPGA which switches the output voltage between two seriously stable voltage levels, representing the ‘1’ and the ‘0’ of a DSD bitstream.  This is quite a smart approach, but it requires every single input data stream to be converted to a 1-bit format, and the one they have chosen is 1-bit / 5.6448MHz, otherwise known as DSD128.

As it happens, the chip-based solutions don’t quite do it this way.  Instead of a 1-bit format, they use a 3-bit, 4-bit, or even 5-bit format, and there is a good reason for doing that.  The devilishly complex Sigma-Delta Modulators (SDMs) used to perform these conversions are fundamentally unstable when configured with a 1-bit output.  This instability can be avoided entirely by using a multi-bit output format.  But the actual D-to-A conversion using those multi-bit bitstreams is more complex.  In effect, PS Audio has traded the electronic complexity of a multi-bit high-speed D-to-A converter for a design approach they can get their considerable electronic design chops around.  All that needs to be done is to navigate a way around the 1-bit SDM instability problem.  As it happens, there are various techniques to mitigate this problem, although it cannot be conclusively eliminated.  But by taking the right measures it can be beaten into submission, otherwise SACD/DSD couldn’t be made to work.

Inside the DirectStream, all input formats – DSD128 included – are converted to a single 30-bit 10MHz format, and from there converted to DSD128 which is fed to the core converter.  There are a couple reasons for this.  The first is that digital volume control can be easily implemented in this intermediate stage.  The second – behind which hide an advanced degree’s worth of technical matters – is that the analog output filters in the output stage assume a certain implementation of the DSD128 data stream to which an arbitrary incoming data stream might not adhere.  [My thanks to Ted Smith of PS Audio for clearing that up for me.]

Paul McGowan has created a lot of hype around the DirectStream through his intriguing claim that it manages to extract more musical information from ordinary CDs (read 16/44.1-formatted music) than any other DAC.  By and large, this claim seems to be backed up by other people who have heard the DirectStream for themselves.  So a significant part of this review will be devoted to assessing its Red Book (16-bit, 44.1kHz PCM) performance, although don’t expect me to pronounce one way or the other on McGowan’s claims.

My test setup was as follows.  The DirectStream was fed by a 2013 base-spec Mac Mini, running OS/X 10.9.3 with iTunes 11.2.1 and BitPerfect 2.0.2 (a pre-release beta).  The DirectStream was fed into a Classé CP800 preamplifier which in turn fed a Classé CA2300 power amplifier.  The loudspeakers were B&W 802 Diamonds.  The Mac Mini used a PS Audio Jewel Power Cord, and the USB cable was a 1m Nordost Blue Heaven.  All other Power Cords, plus all the (balanced) interconnects were BitPerfect Digital Precision products.  The speaker cables were dual (bi-wired) runs of Cardas Golden Cross.  Not only the DirectStream, but also its Power Cord and its interconnects, were all new and unused and so a period of break-in was required.

Powered up for the first time, and prior to any break-in, the sound was noticeably thin, brittle, and edgy.  This could be down to the DirectStream or any of the brand new cables (and I have little interest in finding out which was responsible for what), but nevertheless you could tell immediately that here was something with serious detail in its sonic presentation.  I set it up for a period of continuous break-in which I was prepared to last for at least a good week.  In fact the improvement over the first 24 hours was quite dramatic, and by the end of the fourth day the previous 24-hour period seemed to have wrought no significant improvements, so at that point serious listening commenced.  If significant break-in over a time frame of weeks should occur, I might report on it later.

PS Audio makes a point of recommending that you connect the DirectStream directly to the inputs of your Power Amplifier for a potentially ideal listening experience.  It took me until the sixth day of listening before I got round to that, but the result was so immediately and obviously superior that this instantly became my preferred configuration.  This entire review uses the DirectStream connected directly to the Power Amplifier’s inputs.

Before going on, there are a couple of matters specific to BitPerfect users.  The first is that DirectStream only supports PCM sample rates up to 192kHz.  However, in order to deliver DSD128 support, it must also support a PCM stream format at 352.8kHz which it duly announces to OS/X.  However, if you send it a true 352.8kHz PCM data stream it does not play properly.  Therefore it is important to set BitPerfect’s “Max Sample Rate” to 192kHz.  I would also check “Upsample by Powers of Two” so that any DXD (24/352.8) tracks which you may have – and I have quite a few – will be downsampled to 176.4kHz rather than 192kHz.

The second matter relates to volume control, and requires an extended discussion.  First of all I used BitPerfect’s Volume Control (the slider in the menu bar drop-down menu), which addresses the DAC’s USB-accessible volume control.  While this does indeed control the volume it has some unexpected behaviour.  The first is that it operates independently of the DirectStream’s own volume control, and cannot be addressed within the DirectStream itself.  The second is that when used on a DSD track it causes the volume to be muted.  BitPerfect Users used to using the volume control via its keyboard shortcuts may find this behaviour slightly annoying.  According to Paul McGowan this behaviour arises because the XMOS USB receiver is unexpectedly acting upon the USB-delivered volume control commands.  At this point, I am not sure which direction PS Audio will take when they get around to addressing this issue.  Having a USB-accessible volume control is usually quite desirable in a Computer-based audio system, and I hope they will take that into account.

A personal irritation is that the DirectStream’s internal volume control is calibrated on a 0-100 scale, with 100 representing 0dB and each ‘1’ dialing in 0.5dB of attenuation (and ‘0’ being muted).  I personally prefer to see the dB attenuation scale indicated (which is actually more common these days).  In comparison, 0-100 is a bit like a car’s speedometer calibrated as 0-100%.  “Honestly, officer, I was only doing 55%!!”.  But I appreciate that other users may have a different perspective.

Build quality of the DirectStream is everything you would expect from a $6,000 component.  Mine is finished in silver (black is an option), and has an unusual glossy piano-black top surface which looks very crisp, and as far as I can tell has a primarily cosmetic function.  Sitting on top of my equipment rack it also ties in visually with my piano-black 802 Diamonds.  For a DAC it is surprisingly heavy, not least because its analog output stage unusually uses a transformer (although I have not opened it up to see how big the transformer is).  The heft also suggests a generously-specified linear power supply.  Given PS Audio’s expertise in audio power supply, this would not be in the least surprising.  The front panel is bare except for a small touch-panel display.  This display can be dimmed if preferred (turned off, actually) using a button on the supplied remote control, but having done so I could not un-dim it without powering down the DirectStream.  Am I being unfair by observing that the remote itself is a bit plasticky, in comparison with the hewn-from-a-solid-billet look and feel of the DirectStream itself?  The similarly-priced Classé CP800 has a remote that might break your toe if you were to accidentally drop it.  All things considered, the DirectStream’s physical presence is fully consistent with its price tag.

So how does it sound?  Click here and let’s find out.

For a long time, in order to get gapless playback to work reliably, I have been enabling the setting “Default to Fixed Indexing” in the iTunes section of BitPerfect’s Preferences Window.  However, I have noticed that, since performing the latest 11.2 update to iTunes, gapless playback has been problematic. Today I tried disabling “Default To Fixed Indexing”, and now gapless playback is working properly again.  Users encountering gapless playback problems might like to try doing the same thing.

I have been using the new updates to OS X 10.9.3 and iTunes 11.2 overnight.  It has run my standard battery of compatibility tests and no new problems have emerged.  At the same time, it is clear that none of the old problems that have been known for some time have been addressed either.  I think BitPerfect users can feel safe upgrading.

Particularly with Digital Audio, you often run into situations where logic appears to dictate one thing, but experience favours another.  As a consumer there is little to be gained by pursing the logical solution in the face of a better sounding practical alternative.  The only reason for listening to music is to enjoy the experience.  Why choose to listen to something you enjoy less because some arcane explanation says that it ought to actually sound better?

However, for those of us involved in this business on a professional level, we cannot afford to take such a simplistic approach.  If something that should sound better actually ends up sounding worse, it is usually in your best professional interests to find out why.  In most circumstances this is far easier said than done.  Sometimes it can seem an impossible thing to get done.   In many instances the cost of doing the necessary work may be the limiting factor, and that is a very valid consideration for most people in the audio business.  In the software business, it is often a more manageable task.

Take the question of DSD playback vs playback of a PCM conversion of the same DSD data.  It is a known fact that you cannot convert losslessly, back and forth, between DSD and PCM.  DSD stores some information which cannot be adequately represented by PCM, and vice versa.  Simple information theory therefore states that if you make a PCM version of a DSD track, no matter how advanced the quality of the conversion, some information is going to be lost along the way (I will come back to this later, but for the time being we will stick with it).  If information is being lost, then the result is, by definition, a poorer quality representation.  This is an inescapable fact.  It may be, of course, that the information that is lost is information which contributes adversely to the enjoyment of the playback experience, but it is lost information regardless.  And if we were able to identify it as such, we could take steps to eliminate it from the original recording in the first place.  No, the fact remains, a PCM conversion made from a DSD original, is – de facto – a retrograde step.

So what happens if the PCM version is reported to sound better?  Do we simply argue that it can’t possibly sound better because it is inherently inferior?  No, of course not.  There may be a perfectly good explanation for it that we have not yet stumbled across.

Don’t forget that we never actually listen to digital files.  We listen to analog signals which a DAC has created using the digital files as its source material.  And there is a rather serious amount of processing that goes on between the digital data being sent to the DAC and the analog signal being transmitted out of it.  There is no doubt that this digital processing has the capability to alter the signal substantially enough to account for audible differences.

I have mentioned before that we have ordered a PS Audio DirectSream DAC as our new reference DAC.  I can’t wait to receive it.  Paul McGowan of PS Audio has been using BitPerfect and DSD Master to play his DSD content through the DirectStream.  The Hybrid-DSD files enable him to switch easily between the original DSD content and the PCM conversions for all his DSD music.  Paul reports the surprising observation that he tends to have a slight preference for the PCM conversions over the DSD originals in his system.  I will be doing the same comparisons myself when my own DirectStream arrives.

This would be an example of a circumstance where I would say that the DSD original ought to be “fundamentally” superior to the PCM copy.  Or at the very least as good as the PCM copy.  Having written DSD Master (which was used to do Paul’s conversions), I would say with some certainty that there is no basis for supposing that our PCM conversions could actually improve upon the DSD original.  Assuming that Paul’s observations are accurate and reproducible – and I have no reason to suppose otherwise – this would be a perfect example of a situation where one has little professional alternative but to try and figure out what is behind this unexpected result.  For consumers, though, it perhaps provides another – and quite unexcited – reason to go out and purchase DSD Master!!

I said earlier that I would elaborate upon whether or not data must perforce be lost in a conversion between DSD and PCM.  The obvious caveat here is that, clearly, the DSD must not have been sourced from a PCM master!  There are some glaring real-world examples here.  Some SACDs, for example, have been publicly exposed as having been mastered from 16-bit 44.1kHz CD masters!  The DSD64 format is perfectly capable of storing all of the information that is held in a red book recording.  Therefore, those SACDs, at best, have no possibility of ever sounding better than a perfectly-played CD.  Many other DSDs are sourced from Hi-Rez PCM masters.  Generally, if you were to convert such a DSD file to Hi-Rez PCM using DSD Master, it is quite likely that the resultant PCM may well sound close to indistinguishable from the DSD version.  Still and all, I don’t see why it should sound better