MHDT Lab Pagoda: A review ...
Feb 21, 2016 at 1:54 PM Thread Starter Post #1 of 42


Headphoneus Supremus
Sep 27, 2007

I've used MHDT Laboratories DACs for quite some time.
Having spent a great amount of time exploring the DAC landscape so many years ago, it became quite clear to me after agonizing hours and long protracted moments of soul searching, that I by far preferred the old school, NOS (non over sampling)/R2R DAC circuit when compared to any oversampling, filtering delta-sigma type DAC (this latter circuit being by far the most common among the DACs offered today).
And over those initial months of exploration, I found that among all those NOS/R2R DACs that I could actually afford, the MHDT Labs DACs were the sexiest, most musical, most engaging and addictive listens that I had found and by far worked the best with my similarly loved and near traditionally used Grado headphones (though I should be careful here not to pigeon hole these DACs as strictly “Grado DACs” .. because they are phenomenal with all headphones but they just happen to be stupidly good with Grado's notorious idiosyncrasies).
From the Constantine to the Paradisea, then to the Havana and ultimately the Stockholm 2 DAC, over the past decade I have found little reason to deviate from my use of MHDT Lab's music rendering machines as my main listening vehicles.
And so, after having a few months with The Pagoda and especially after realizing the somewhat marked differences in performance that were immediately evident between it and it's predecessors, I knew I would be wanting to write a bit about how it differed and what I personally have found to be it's more distinct and appreciated hallmarks as a digital to analogue converter.
But before I get in to such issues as performance and sonic impressions, I wanted to write a bit (a good bit) about some of the technical differences between the Pagoda and previous MHDT Labs DACs (mostly the Stockholm 2). 
PCM1704 versus PCM56P / 16 bit vs 24 bit
As with all of the previous DACs mentioned above, the Pagoda is a wonderful realization of the classic NOS/R2R design and yet it differs from the majority of MHDT Lab's previous offerings in a couple of significant and quite frankly, astonishing ways.
First off and most significantly, the Pagoda uses dual PCM1704 DAC chips in it's circuit whereas all previous MHDT DACs have employed the venerable PCM56P DAC chips in a very similar “dual monolithic” architecture (especially the Stockholm 2 DAC, which according to MHDT Labs' lead designer, Jiun-Hsien Wu, employs the identical “discreet, no OPAMP, no feedback” circuit as employed with the Pagoda DAC).

Secondly, the Pagoda DAC is a true 192khz/24 bit design, both receiving a signal in 192khz/24bit while also outputting in 24bit format. All previous PCM56P DACs released by MHDT Labs have been limited to 16bit output, with some (like the Havana 2 and Stockholm 2) being capable of at least receiving 192khz/24bit signals. But even though some might be able to accept this high resolution signal, all have been limited to 16bit output; this is simply an inherent limitation of the PCM56P DAC. By the way, if you would like a nice breakdown of every MHDT Labs DAC currently in production, be sure and check out this handy specs table at the MHDT home page: MHDT Labs DAC Families

As I would find over the months and hours of review listening and hyper vigilant A/B switching between the Pagoda and Stockholm 2 DAC, this inclusion of the PCM1704 DAC chips in the same circuit would make a remarkably apparent difference between the two machines. Further, speaking strictly of high resolution listening, when rendering files encoded at 192khz and 24 bit, the differences would truly become jaw dropping. Really, as I hope to make clear with the following essay, the differences are quite startling and will resound more or less with any given user depending on his/her listening preferences, the types of digital files owned and other gear in the sonic chain.
Because the Pagoda's closest competitor and design sibling is the Stockholm 2, my comments here will focus primarily on a detailed discussion of the differences between these two DACs. They are both insanely good listens, fantastic values and overall remarkably well designed machines; I adore both. But they really will, upon careful listen, render significantly different sonic landscapes. These differences should be closely considered by any person trying to decide whether to buy the Pagoda or the Stockholm 2.
So as I approach the daunting task of reviewing the Pagoda, I have found the greatest challenge to be how to make clear that the Pagoda cleaves entirely to the same musical, engaging and utterly seductive design ethos of all of the MHDT DACs previous, while also representing a sure and undeniable departure sonically, from all of those PCM56P DAC based designs that came before it. Really, I've agonized over this review more than any other I've written, nearly having wrecks on the freeway as imaginary metaphors and analogies unfolded in my mind and/or busting fingers at work while doing the same. So bear with me as I stumble through these difficult things and forgive the sure-to-come run on sentences and gross attempts at conveying emotions in easy to understand, intellectual and literary terms. Cause it really might get “gross” at times. :wink:
The PCM1704 DAC
As I constantly repeat in any bit of writing I do that involves circuits and capacitors and resistors and PCB's and the like, **** man... electronics in general: I am no expert in regards to these things. I've never soldered a damn thing in my 48 years on Earth and I have no idea how circuits and or circuit design works. About the most technical thing I've ever done involving electronics is to use conductive ink pens to do volt mod hacks on computer video cards and motherboards in attempts to overclock them further (overclocking and computers in general being an even more insidious addiction of mine than is audio equipment and music listening).
So some months ago as I began considering the writing of this bit of tripe, my first task was to learn whatever I could about the PCM1704 DAC and what made it special in regards to the building of non oversampling DACs.
I spent a lot of time reading online and discovered quite a bit of interesting information regarding the PCM1704 DAC. Following, you will find some of the highlights of these findings.
Of initial interest was that I discovered the PCM1704 is currently out of production and that it is still considered among a vast grouping of today's top “boutique” builders to be the very best DAC chip currently available among those DACs now out of production or even those currently in production for that matter. Have a brief read around the intrawebz and you will find overwhelming love for this somewhat antiquated and expensive DAC chip.
Further, I learned that all PCM1704 chips were produced with varying grades of quality, being marked as certain “grades” … PCM1704U, PCM1704U-J and PCM1704U-K and the like, with the “K” variants being the highest quality. Because I could see no markings on my Pagoda's PCM1704 chips, other than a plainly indicated PCM1704 screen printing, I decide to contact MHDT Labs lead designer Jiun-Hsien Wu to inquire about this.
Jiun replied:
PCM1704 chips have many grades, usually in three grades, 1) no letter, 2) J grade, 3) K grade. In fact, in early years of manufacture the chip had one white dot printed for J grade and 2 white dots for K grade. Also, there were some specifically classified for car audio where you will find L grade and G grade. The grade we used in our Pagoda is 1) no letter. We may consider to release J grade and K grade in the future.”
Jiun goes on to say:
The PCM1704 chip is very expensive, costing upwards of $70.00 each in orders of 1000pcs or more and is no longer in production. In this way, it is just like the best of the ancient vacuum tubes. The chips just sold out. As for our being R2R designers and builders, we needed to have a product at this level of DAC quality. If you know how hard it was to manufacture a PCM1704 chip, even by today's new IC machine standards, then you will know that the PCM1704 will NOT be in production again in the future.”
Telling words, indeed...
Otherwise, I spent a lot of time reading comments from other builders such as Lessloss Audio and from “aficionados” such as Berry Willis from Stereophile. Across the board, I found nothing but mad respect for the PCM1704 DAC and in fact, a near reverential attitude from many such experts in the high end audio design field.
According to my reading online, the Burr Brown PCM1704 is the only option available for the construction of non-oversampling, 24bit digital to analogue converters. As already mentioned, though the PCM1704 is still recognized among many builders as the best DAC chip available today.
Berry Willis, from Stereophile had the following to say:
The 1704---priced at $12.95 (currently $70.00 as of 02/2016) each in OEM quantities of 1000 or more (SEE EDIT BELOW) ---is a small 20-pin SOIC requiring a ±5V supply. It will accept input data words of 20- and 24-bit lengths at sampling frequencies up to 96kHz, and will support 8x oversampling at the highest sampling rate. Its SNR is near the theoretical maximum for all electronic devices, and enables a 112dB dynamic range---a 6dB improvement over the highly regarded PCM-1728 (see previous story). Unlike delta-sigma converters, Burr-Brown's BiCMOS device is "not sensitive to clock jitter," according to audio product marketing manager Mike Centorino. "Sign-magnitude designs are inherently better than delta-sigma types at handling jitter," he said. "It's a different architecture, specifically designed for high-end performance." Read more at HERE.
EDIT: the above text from Berry Willis is somewhat dated, according to Jiun at MHDT Labs, the current price of the "no label" PCM1704 DAC chips used in the Pagoda is $70.00 per chip in orders of 1000pcs or more.  An INSANELY expensive bit of silicon!
Further, the fine builders at Lessloss audio had this to say:
The PCM 1704 is out of production and represents the end of the era when quality was first on the minds of DAC engineering laboratories. In today's production, all remaining DAC's are of the Sigma/Delta type. These incorporate more technologies into one chip, including two channels for stereo, volume control, up sampling, and often even clock oscillators of their own. Production costs have been saved, but the issue of quality of sound reproduction is no longer the primary issue … “
And further, they say:
There is a fundamental difference between the way parallel multibit converters (like the PCM1704) and the sigma/deltatype work. The parallel type use a separate cascade of resistors and switches for each dynamic modulation of the audio signal, whereas the sigma/delta type (or one-bit, as they are also called), rely on a constant comparator to define changes in the audio signal's dynamic magnitude. Each method has its own advantages and disadvantages.
The main advantage of multibit conversion is that they are theoretically less susceptible to the influences of clock jitter. Their main disadvantages lie in their sensitivity to the influences of heat fluctuation.

The sigma/delta microchips have the advantage of being less sensitive to heat fluctuations, however, they react very readily to any amount of clock jitter.”   There's lots more good technical data at LessLoss' original article.

Now again … such technical banter above reads like so much delicious geek voodoo and in my wildest imaginings I would totally understand every word and symbol used in such verbose mysticism … but truth is I don't understand much of it and I include the above for those of you more technically minded who might further comment on such technical minutia. I do know this, though. The PCM1704 is undeniably held in a near mystical regard by a great many out there in Hi-fi builder wonderland and it really does represent quite a departure from MHDT Labs typical PCM56P DACs.

But why call it "The Pagoda?"

In conversation with lead designer, Jiun, I asked why they chose to name this new DAC offering “The Pagoda;” Jiun replies:

We are all Buddhist. For Buddhists, the term “Pagoda” has two distinct meanings: First, Pagoda is the highest achievement of memorial in all things, especially architectural forms. And second, every eminent monk has a sacred, protected space where he keeps all his treasures and those things he holds most sacred; this is also known as a Pagoda. Since we consider the PCM1704 chip to represent the highest standards in regards to the R2R DAC design, we in turn decided to call this new DAC The Pagoda.”


Not to side track this discussion of audio towards that of philosophy or religion, but I personally found these comments fascinating and I was very glad I thought to ask. Because though this discipline of circuit design and construction and sonics is certainly one of a mechanical, electrical nature … it's also one that arises out of personal passions, beliefs and principles as related to those who create such works. I know this myself as a decades long visual/interface designer in the software field and so Jiun's words uniquely resounded with me in that regard. It's nice to find that the creators of the Pagoda are so concerned about “sewing” their more “ephemeral” and/or personal beliefs in to the fabric of the machines they design and build and it increases my enjoyment of having a chance to experience their works.

Build quality and technical specs

As with the Stockholm 2, the Pagoda is housed in the very same brushed aluminum casing and utilizes the very same “smoked” acrylic face plate that allows a user to just barely glimpse the internal circuitry while also giving a nice window on the subtle glow of the Pagoda's tube buffer stage. Further, as with previous DACs from MHDT Labs, the Pagoda is available in an all black chassis or a silver chassis as seen in the included photographs here.  The look is understated elegance; all the seams and fittings are perfectly aligned and overall, the entire kit, as with previous releases, is very easy on the eyes.


Here once more I am reminded of myself in a design regard as MHDT Labs has seen very little need to change the same basic exterior/industrial design of the chassis of their DACs over the years. Go all the way back to the first Constantine DACs and you will find a variant of this same approach (albeit in all acrylic). It's indicative of good design I think that a designer might stick to a proven approach/solution for things such as mechanical casings (and in my case, framing for page layouts or certain element arrangements which are repetitive through any given interface project I might be working on).



It's nit picky of me and is certainly not meant as criticism but my inner typographer likes the “script” typeface used on the Stockholm 2 face plate more than the more modest, “san serif” type face used on the Pagoda. The script typeface yields a bit of understated elegance to the package which works very well with the otherwise “spartan” and wholly efficient use of materials in the visual presentation of the Stockholm 2 case. The Pagoda's more conservative type face doesn't really convey that visual elegance as neatly; but then again … I'm a type guy/ad guy/design guy from way back and these things jump out at me and so I thought I'd comment.

Here's the technical specs of the Pagoda:

DAC chips: dual PCM1704
Input Capacity USB (Max) : 24bits/192kHz
Input Capacity SPDIF (Max) 24bits/192kHz
Output Format: 24 bit
I/V Stage: Current Out Discrete Transistors I/V, No OPAMP, No feedback

Output level: 3.0v
Available inputs: 4 Inputs -- USB/RCA/BNC/Optic
Dimensions (WxDXH): 276 x 150 x 60 mm
Weight: 2kg
Tube Buffer: utilizes a single 5670/2C51 tube



As with every other MHDT Labs DAC, the Pagoda employs a single 5670/2C51 tube in a non-amplified, "buffer stage".  The affect of this tube on the overall sound is extremely subtle. This buffer stage does, in no way, create what one might call a "tubey" sound or an overtly warm sound.  More so, as one changes tubes, one will here very subtle changes in the rendering of high, mid and low frequencies; one tube will yield slightly smoother highs with a deeper bass (WE396A) while another tube might yield more detailed, forward highs and snappier, tighter bass (Bendix 2C51). These 5670 tubes are generally quite affordable (other than the Bendix variants, which can get very expensive) and so rolling different tubes can be quite fun. 

For the sake of this review, I am listening with a AEG 5670 tube, one of my personal favorites with any of the MHDT Labs tube DACs. The stock GE tube sounds pretty good too, but the AEG (and also the Bendix 6385) is a definite upgrade in regards to overall sonics (more clean, better detail, bass … less grain, etc.).

Just regarding tubes? Here's a pic of a JW WE396A and Bendix 6385 tube:

Also, and this applies to any of the MHDT Labs DACs that use tubes (Havana, Stockholm, Constantine, Paradisea, etc.) ... a little known tube that is honestly right up there with the Bendix 6385 in regards to cleanliness, detail and gorgeous high frequency retrieval is the humble, AEG 5670 tube. This is a top notch tube for use in MHDT Labs DACs and is cheaper than dirt.  It's also almost impossible to find.  But if you can find one, go ahead and spend the $6 bucks or so for it and save yourself about $150 to $300 on that Bendix 6385 purchase because they are almost indistinguishable from each other when listening (spatious, airy, detailed, tight bass, etc.) ... 


Sound and listening impressions

Even before I get in to the meat of this, I know this particular section of the review will not only be the most difficult to elaborate on, but also probably the shortest. Reason being, the Pagoda DAC is basically the exact same very well realized NOS/R2R circuit found in the Stockholm 2 DAC but with PCM1704 DAC chips in tandem as opposed to the Stockholm's PCM56P DAC chips. Therefore, in many ways and certainly upon a casual listen, the two DACs are very similar sounding and obviously “cut from the same cloth.”

Like the Stockholm, the Pagoda is imminently natural sounding, highly detailed while never sounding artificial or digital or mechanical in any way (poorly designed Delta Sigma DACs? I'm talking at you!). The Pagoda, as with any well designed NOS/R2R DAC is just a stones throw away from what some of the best vinyl rigs can deliver in regards to musicality, ease of listening and pure enjoyment of music; it does not beg that you analyze and it does not convey any affectations whatsoever. One will not find oneself squinting at strident highs or woolen bass and though plenty detailed, one will never find oneself wanting to hyper focus on some anomalous “tick” or “shuffle” or “page turn” that is a given with well recorded music … rather, these subtle but evident sonic nuances are more so rendered as part of the overall musical whole. As repeated so many times in various other reviews and posts in general, I've just never found anything so musical and engaging as well designed NOS/R2R DACs. And being a lover of music first and the machines that deliver it, second … I've never found cause to venture off the path of this antiquated but still highly regarded circuit design.

I've said so much along these lines already in a previous review I did of the Stockholm 2 DAC, that I'm going to forego any more winded dissertations on exactly how fine and musical and engaging of a listen the Pagoda DAC is. Rather, the remainder of this bit of writing will focus entirely on how and where I've heard the Pagoda to differ sonically from the Stockholm 2.

Pagoda vs Stockholm 2 - It's all in the details...

That said, the inclusion of the new PCM1704 DAC chips in the Pagoda does indeed create some rather striking and very notable differences in the sound of the Pagoda which are undeniable and frankly, quite astonishing when compared to the Stockholm 2 and it's PCM56P DAC chips.

Upon mentioning these differences in conversation with MHDT Labs lead designer, Jiun, he was careful to mention that because both the Stockholm 2 and Pagoda DACs use a circuit completely free of any operational amplification, being a simple “current out/no feedback/discreet transistor” design, that this allows a listener to clearly hear the differences between only the DAC chips that are used in the circuit.

Jiun comments: “The main circuits of the Stockholm 2 and Pagoda are almost identical. Using such a discrete I/V circuit such as this, one can easily here the differences between the two simply based on the DAC chips themselves.”

Personally, I love this idea of cleaning up all the various pathways, interconnects and various exchanges that go with circuit design so that one can focus simply on the main digital to analogue conversion components. That being said and because I'm always debating myself in my own mind, one could argue that the inclusion of a tube buffer stage in these DACs is in no way “clean” and is in fact a bit counter to the comment I just made...and technically, that argument would be correct. But I see the tube buffer stage simply as a goal of the designer to better sculpt the final rendering of music to a tonality of his/her desired project aesthetic, an intended bit of v e r y subtle coloration if you will and also a feature that can make the machine more enjoyable as one can roll different tubes to achieve a slightly different sound; I'm quite okay with that and even enjoy it. Take it from me, as designers go, myself included … it's all about “gestalt,” … and if the inclusion of a tube buffer is what does it for the MHDT Labs gang and if the results happen to please my ear as well? Then so be it. But see the final paragraph on possibilities of future DAC releases having the ability to bypass this tube buffer stage because such things are indeed in the works... :wink:

Otherwise, regarding my critical listening...



When first setting out on the review listening process, I was fortunate enough to be able to borrow a really swell Mapletree Audio Line Router from Head-fi user Shellylh (thanks!) and this simple device has made the process of direct and quick “A/B” listening very effortless and very effective. Using an optical splitter, I was easily able to run from the optical output of my computer and into the splitter and from the splitter I could then run an optical cable to the Pagoda and another cable to the Stockholm 2. From both DACs I then ran into the line router and from the router I then ran a single RCA cable to my Mapletree Purist Ear+ Purist HD amplifier. Once those connections were made, and after starting playback of any given track, switching between the Pagoda and the Stockholm 2 was as neat and easy as turning a knob from position 1 to position 2. The switch is/was instantaneous with no gaps or silence between the transfer. This really assisted me in nailing down the following listening impressions.

Otherwise, all listening was done via the above visible audio chain and with a pair of Grado PS1000 headphones.  That said ... on to the listening impressions...


General detail and micro detail retrieval

One of the key differences that became apparent upon first comparing the two DACs is that the PCM1704 based Pagoda is a significantly better performer in regards to the retrieval of musical detail than is the PCM56P based Stockholm 2. Neither DAC is “clinically” granular in it's detail retrieval which is to say that neither DAC sounds clinical or unnatural or as though they are artificially enhancing details in the music. But the Pagoda is certainly the more capable in regards to rendering such details and especially, the “micro” details of any given track.

And by details and “micro” details I mean the oh-so-important, subtle nuances inherent in any well recorded bit of music that are so often key to inducing chills in the listener or that might make one look over one's shoulder for the thought that there is someone moving behind you. Things like the high pitched slide of a bass players fingers over a fret board, or on the other hand, the lack of the obvious “bump” sounds that frets on a fret board make as those fingers slide, which might clue the listener in to the fact that the bass players fret board doesn't have frets on it at all. Or the wooden “tap” of a traditional drumstick on a high hat versus the soft “scratch” on the same high hat which would indicate the use of brushes instead of sticks. Little things like the misplaced tap of a foot against a mic stand during a live recording that invariable will emit as a “thump” through the house PA or the perennially mentioned “page turn” so often heard in well recorded symphonic works. Shuffling feet, coughs from the audience, the way reverb in a live recording can reveal the size of the room being played in, strings buzzing on frets, strings breaking and even the turning of tuning pegs can all be heard in well recorded music and both the Pagoda and the Stockholm 2 DAC render them wonderfully and naturally.



But the Pagoda does it better.
It's not exaggerated and it's certainly not artificial sounding but the presence of these details is noticed more so with the Pagoda than with the Stockholm 2. There is a certain “haze” that is lifted when moving from the Stockholm 2 to the Pagoda and suddenly the musical details become more apparent, more defined overall and what you might call having a “more visible shape” or being “more solid around the edges.” This “haze” I mention isn't necessarily a bad thing and in fact, I think some users will prefer the more laid back, easy going nature of the PCM56P based Stockholm 2 over the more technically minded Pagoda … but more on that in my closing thoughts (because I really think either DAC will appeal more or less depending on one's listening preferences).
As with so much of this review pontificating, I find myself reaching for metaphors and descriptions which always seem inadequate after the writing but I know of no other way to communicate it.

But on to another key difference that I have noticed and one that is no less apparent than the comments made regarding detail retrieval above...


Sound stage and imaging

As with the above comments, I also heard a considerable difference between the Pagoda and Stockholm 2 in regards to sound stage and general imaging. By “sound stage” I simply mean how “outside of the head” the music sounds and also how much perceived “space” one might hear around or between the players in any given track, how “three dimensional” the recording sounded and even if one might imagine a stage in front of themselves while being able to place individual players from left to right in that imagined stage space.

Headphones, admittedly, are really not good at this sort of critical listening and imaging. Most every headphone I've heard has simply either created the impression of the musicians being “inside the head” or even, oddly enough and frequently, emanating from behind me. Call it an idiosyncrasy of headphones in general if you will but this has been my experience (especially among all closed back headphones I've owned). The few exceptions to this rule have been several mid range to high end Stax headphones I've owned, the Sennheiser HD800 and especially, more than any headphone I've ever heard, my current go-to headphone, the Grado PS1000. Really, I've never heard any headphone from any manufacturer that can or has equaled the spooky spatial qualities I get from the Grado PS1000's; they are, hands down, the most “speaker like” headphone I've ever experienced and with them, the differences between the Pagoda and Stockholm, specifically and especially in regards to sound stage and imaging, are quite easy to hear.

Said plainly and via the grossest analogy I can think of: direct A/B switching between the Stockholm and Pagoda is like hitting an Expand Sound Stage Button(tm). I had never thought of the Stockholm 2 as sounding “closed in” … or “narrow” … or “constricted” or anything even remotely critical in regards to sound stage. It was only when I began the A/B listening and when directly and immediately switching over to the Pagoda that I began to realize the differences.
The Pagoda is a sound stage champion in comparison to the Stockholm 2. Again, not that the latter is bad in this regard, it's just not nearly as good as the Pagoda. The difference is really that dramatic. For almost one entire evening, I spent hours just listening to live jazz tracks and chamber music and other sorts of music where sound stage becomes important (IE > not Agent Orange or Fugazi unless well recorded live tracks) … switching back and forth, over and over I was repeatedly floored. The PCM1704 DAC is simply a far better imaging DAC than the PCM56P DAC.  And this last comment I think is where the importance lies; since both DACs are identical circuits, it's obvious that this difference regarding sound stage (and all differences, really) boils down to the inherent limitations and/or strong points of either the PCM1704 or the PCM56P chips.
What's more, the room itself becomes more “visible” as one listens to the Pagoda. The players are more defined and individual in their placement within the “soundscape;” though using headphones, with my Grado PS1000's I could easily place individual instruments and players within an imaginary space out in front of me and there is a spooky, three dimensional quality to the rendering that isn't as pronounced via the Stockholm 2. Switch back to the Stockholm and it's like hitting that same button but this time it turns the sound stage down a bit, it narrows the image a tad … things are still “right” in a dimensional regard, the imaging is still there but it's not as focused as the Pagoda's take on things.
Those are rather heady attempts at describing the differences between these two star players and I want to be careful to reiterate that both of these DACs are exceptional performers. But, being honest? The differences are really that noticeable...

This sound stage bit really comes in to play in regards to the delivery of orchestral works wherein one's ears might be assaulted by 40 or more individual players. The Pagoda handles these complex arrangements better than the Stockholm 2, allowing one to more precisely hear the myriad of players and instruments in the hall; again, the ability to image the entire scene is just uncanny with the PCM1704 based Pagoda (the word “spooky” keeps coming to mind). The Stockholm 2 does a fine job here as well, but the PCM56P DAC just doesn't handle things as meticulously, presenting a more “smoothed” out image where instruments ever so slightly seem to blend together and players are not as easily placed individually in the overall sonic image. Here again, some listeners might prefer this...
To re state things, as with the “details/micro details” comments above, the Pagoda is the more precise machine in regards to the rendering of music. The well dressed Pagoda might be slightly less “sexy” sounding than the more casually dressed Stockholm 2, it might not be quite so forgiving of bad recordings but it has the same NOS/R2R soul and the same delicious musicality as the Stockholm 2 but with considerably more refinement. If I were to continue with a female metaphor of sorts? (It would be disingenuous of me to discuss things in a male context here): Where the Stockholm 2 arrives all tan and sweaty in it's sun dress and flip flops, hair a bit unkempt but alluring as all get out, sipping a Belgian Ale? The Pagoda in turn arrives dressed to the nines in patent leather black pumps, hair meticulously sculpted in the finest of enticing shapes and layers, nary a strand out of place and wearing a red, split back evening gown, casually and quickly tipping back neat shots of Mortlach 70 Year Scotch. And as each of these turn through the room, their respective hemlines occasionally part or rise to reveal identical crescent moon shaped tattoos on their ankles.
That was bad but man, it's the best I could do. The point is that they are from the same club. They are sisters. But they are certainly not twins...
Other subtle differences...
High Frequencies
This is perhaps related to the former section on details but the high frequency performance of the Pagoda is markedly different from that of the Stockholm 2. The Pagoda presents highs with more acute, sharper edges than the Stockholm 2. The highs are more defined and again, rendered in a more detailed, precise way. On the other hand, the Stockholm 2 renders highs in a slightly more “smooth” manner … not necessarily what I would call “recessed” or “rolled off” but certainly a bit more forgiving, a bit more laid back. As I will discuss towards the end of this horrible essay, many users will prefer this aspect of the Stockholm 2, depending on their ear, their brain and especially their gear. But it should be said, there is more “shimmer” with the Pagoda … there is more “sizzle” to splash cymbals, more "snap" to snare hits and generally, sharper edges on all things high frequency.  Neither DAC is ever, in any way fatiguing in regards to this critical frequency range, however ... so those of you (like myself) acutely sensitive to strident/peaky highs ... either DAC will be completely acceptable.
Bass Frequencies
Further and just in general, I would like to imagine that I heard ever so slightly tighter, more well behaved bass performance with the Pagoda. But this was/is not nearly so matter of fact and certain of a claim as is everything I've said heretofore (details/sound stage/imaging/highs). But another thing about my main headphones, the Grado PS1000's, is that they are a little “bass forward” past a certain point in the volume level, especially in the mid bass region. Therefore, I'm extremely sensitive to bass performance. Be it a difference between the PCM1704 and the PCM56P DACs or be it due to the use of slightly smaller output caps in the Pagoda, real or imagined, I think the bass a tad more neutral and a tad more well behaved with the Pagoda. By the way, for what it's worth … the Pagoda uses 2.0uf output caps where the Stockholm 2 uses 2.2uf … a quibbling difference at best.
I would write something here about mid range frequency performance but I could hear no discernible difference between the Pagoda and Stockholm 2 DACs.
High resolution listening
Another area where I found the Pagoda to excel was when listening to high resolution tracks (192khz/24bit, etc.). Here again, as with the significant difference in regards to sounds stage, I also found a marked difference between the Pagoda and Stockholm 2 when rendering high resolution music. 192Khz/24bit tracks were simply more resolved with the Pagoda, sounding markedly more dynamic, more detailed and more expertly rendered than when listening via the Stockholm 2. When first listening to some of the Chesky Records high resolution tracks, I finally heard what makes this sort of music stand out from typical Red Book 44khz/16bit music, whereas for years, with the Havana DAC and also the Stockholm 2 DAC, I could really detect know significant difference. There really is something to that whole high res thing, after all and it took a proper 24bit DAC for me to hear it. Both the Havana and Stockholm 2 can receive these files just fine but ultimately, due to the use of the PCM56P chip, they still output in 16bits which, in essence, truncates the data present in 192khz/24bit music. For those interested in high resolution listening, the Pagoda is without a doubt the obvious best choice.
In Closing...
The Pagoda is a more “airy” more “spacious” sounding DAC than the Stockholm 2; it is more detailed and has noticeably more defined imaging capabilities and a wider, more three dimensional sound stage. The Pagoda is far less forgiving of badly recorded music and is a much more “precise” vehicle for the rendering of digital music than is the Stockholm 2, especially in regards to the rendering of high resolution music. The Stockholm 2 on the other hand is the more laid back, more casual listen … one could even argue, the more “hypnotic and/or engrossing” of the two as it doesn't so much beg that you study things so precisely as with the Pagoda. The Stockholm 2 is all about being sexy, smooth, playful and inviting. The Pagoda on the other hand comes dressed in a similarly sexy outfit, but with a microscope. The Pagoda will ask more of you as a listener while the Stockholm 2 begs more so that you sit back and enjoy the music. The Stockholm is a more “liquid” sounding DAC, a more “saturated” rendering, if that makes any sense. On the other hand, the Pagoda may seem a bit “dry” in comparison, though still highly natural sounding and at the same time significantly more detailed.
Whether the Pagoda will appeal to you as a listener or if you might prefer the more easy going Stockholm 2 will depend entirely upon your musical preferences and the gear with which you listen to music.

Personally, I find either to be just as enjoyable and find myself completely unable to form an opinion as to which I prefer more. As with the Havana and the Stockholm 2 … the Pagoda and Stockholm 2 are still very much from the same family (though with quite more noticeable differences than the former two when making comparisons and due to the use of two entire different DAC chips in their respective circuits).
But what can be said about both the Pagoda and Stockholm 2 as well is that they are music lover's DACs. Though more detailed, more analytic, the Pagoda is still nowhere near the realm of “hyper digital sounding” DACs, those DACs that I'm sure most of you with any experience have heard that just sound mechanical or processed or unnatural. Both DACs present music as you would hear it in the studio or when being played live and both are perfectly representative of what makes the classic NOS/R2R DAC circuit so attractive to those of us who enjoy it.
As final word, and as with previous reviews, I feel compelled to cite a closing paragraph from a certain other of my writings which seems to always apply to any MHDT Labs DAC that I am lucky enough to spend time with:
"The Pagoda DAC, like it’s predecessor (the Stockholm 2) is a music lover’s DAC. Music played live, in The Real World™, is clean and beautiful, sometimes a bit fuzzy around the edges and other times strident and boomy; real music made in real spaces and with real recording instruments could be extended and moody, three dimensional or closed in, it sounds like the brass of a trumpet or the soft gut string of a classical guitar. It twangs and resonates at times. Other times it is punctuated, brief and tight. Music, as played by any given player in any given space and upon any given instrument is as myriad and diverse as the sorts of equipment built to reproduce it...and in my opinion, audio listening gear should present any given recording in a manner that is as true to the recording environment as is possible. To this writer's ear? The Stockholm does it."
(Oh yeah!)

I almost forgot to add this, but it's sort of a neat bit of “insider” info. I asked Jiun if they had ever planned on making a DAC with the added feature of being able to bypass the tube buffer stage, via a switch or a toggle or the like.

He replied:

Yes, we have. But it will probably be in an adapter form. With such an adapter you can choose to bypass the tube buffer and listen in a pure solid state condition. As well, using certain adapters in the same design, you can even use other tubes like sub-miniature tubes (6021, 6111, 6112, 6BF7) or even different triode tubes.”

I found this pretty exciting because for as long as I've used MHDT Labs DACs, I've always wondered what sonic change it might make to bypass the tube buffer and simply listen to the pure solid state circuit. So hopefully we'll see such abilities in the future. My thinking is that the adapter solution will most likely be applicable to any existing MHDT Labs DACs and not necessarily tied to some new release. Schitt Audio has a similar option available for their tube amps and I'm thinking MHDT's adapter will function in a similar way.

And finally, I asked if they had any new projects in the works that he could comment on and he replied:

We have no new projects in the works at the moment, but we may after this Lunar New Year.”

So some exciting stuff from Jiun at MHDT Laboratories (thanks, Jiun for being so amicable in our discussion and keep up the great design work. You and your crew have a lot of fans out here...


Also, if you want to read more about the full range of MHDT Labs NOS R2R DAC offerings in general, be sure and check out the thread recently created by Head-fi user "LuckBad":  MHDT Labs: R-2R NOS Tube DACs

Happy Listening!




MHDT Lab Home Page--> MHDT Lab Home

MHDT Lab Ordering information--> MHDT Lab Order

MHDT Lab Product Comparison--> MHDT Lab DACs

Feb 21, 2016 at 2:06 PM Post #2 of 42
Incredible review, @s1rrah
I'll be hearing @Soundsgoodtome's Pagoda later this week (swapping an Atlantis for his Pagoda so we can both hear them) and his Stockholm at the end of the tour.
Really cool that they might have an adapter coming to bypass the tube buffer. I'd love to see what these sound like without a tube to know what sound we're influencing.
At some point, I'd love to get an Atlantis in your hands for a week or two in a swap as well.
Feb 21, 2016 at 2:07 PM Post #3 of 42
one of the best review i have read thanks
Feb 21, 2016 at 2:24 PM Post #4 of 42
thanks Joel, good work... I think I'm gonna get the Pagoda in the nearest future, I'm just wondering how it compares to my other reference DACs, no worries, no upgrades this time :)...
but replacing the chips PCM1704 with K grade would tempt me...
Feb 21, 2016 at 2:51 PM Post #5 of 42
  Really cool that they might have an adapter coming to bypass the tube buffer. I'd love to see what these sound like without a tube to know what sound we're influencing.

The DAC may lose its old school sound character without the tube buffer especially with PCM1704 DAC on board. I think it loses that magical, sweet sound the people love it for...
PCM1704 it detailed and cold and sounding rather like the modern DACs.
Feb 21, 2016 at 2:57 PM Post #6 of 42
The DAC may lose its old school sound character without the tube buffer especially with PCM1704 DAC on board. I think it loses that magical, sweet sound the people love it for...
PCM1704 it detailed and cold and sounding rather like the modern DACs.

I'm mostly interested in getting an idea of what the DAC sounds like without the tube buffer so I can have a full understanding of what I'm doing to the sound. I don't actually plan to use it long term without a tube or anything. :wink:
Feb 21, 2016 at 3:01 PM Post #7 of 42
This is super helpful. I think I'm leaning towards the Stockholm end of things when I look at an MHDT DAC since the Pagoda sounds similar to a lot of the characteristics that my Master 11 has
Feb 21, 2016 at 4:38 PM Post #10 of 42
Great review and comparison Joel! Like the fun tidbits along the way as well.

This won't happen until May for a direct a/b but I'll try to go off memory and give you impressions next week sometime.
I might be able to do that...I do want to see what the Atlantis does against it first though.
Feb 21, 2016 at 4:41 PM Post #11 of 42
Great review and comparison Joel! Like the fun tidbits along the way as well.

This won't happen until May for a direct a/b but I'll try to go off memory and give you impressions next week sometime.

That's great actually. May is when I was targeting to buy anyway. If I bought now it would be more of an impulse buy than anything.
Feb 22, 2016 at 11:28 AM Post #12 of 42
An interesting bit of info on the PCM1704 DAC chips and that is also indicative of just how scare they are becoming is that the current price of the PCM1704 is $70.00 per chip in orders of 1000 pieces or more.  And that's for the "no label" chips as used in the Pagoda DAC.  The "J" and "K" grade marked chips are most likely ridiculously more expensive.
If anybody knows the going rates for typical delta sigma chips, then you'll understand just how expensive $70.00 per chip is.  LOL ... that's some high dollar silicon right there.  
I'm posting this as it required some edits to my review where I had cited very old information.  Recent dialogue with MHDT Labs shed some light on the going rate for those chips and so I've edited the review...

Users who are viewing this thread