Jump to content
  • The Computer Audiophile

    Review | Berkeley Audio Design Alpha DAC Reference Series 3

    R1-06312-0002.jpgBerkeley Audio Design is a unique company to say the least. When USB audio was the darling of the HiFi industry, Berkeley said no way to putting a USB interface inside its DACs. When another three letter audio format named DSD started coming back to life from the ashes of SACD, the company also kept the format out of its DACs. Over the last several years Ethernet has become a major part of many audiophile systems, yet Berkeley's founders again say the technology isn't for them (to include inside a DAC). This decision wasn't made due to lack of understanding, especially considering Berkeley co-founder Michael "Pflash" Pflaumer was the inventor of the highly successful local area network TOPS. To say he has directly related engineering chops is an understatement. The real reason for excluding USB and Ethernet comes down to a fundamental goal of the company and that's reducing noise as much as possible. The more noise Berkeley can keep out of its products, by excluding noisy physical interfaces and ancillary processing, the better the performance. This aforementioned mantra goes back to my earliest conversations with Berkeley Audio Design. The company marches to its own drum, and an incredibly high resolution drum at that.

     

    The original Berkeley Audio Design Reference Series DAC was released in 2014. At that time I concluded, "The Berkeley Audio Design Alpha DAC Reference Series is a DAC for the ages. Delicacy, detail, and unparalleled transparency are hallmarks of the RS... If I could afford it and my job allowed it, the Alpha DAC RS is the only DAC I'd use for the foreseeable future." Near the end of 2016 I wrote about the second generation of the RS DAC by saying, "The difference in sonic quality between Series 1 and Series 2 is substantial. This difference makes the decision to upgrade from S1 to S2 a no brainer... The Alpha DAC Reference Series 2 is easily Berkeley Audio Design's finest work to date and a reference by which other DACs will be measured.

     

    I've had the Berkeley Audio Design Alpha DAC Reference Series 3 in my system for a few months and I must say the trajectory from RS1 to RS2 to RS3 is linear. Each new release brings us audiophiles a new level of performance, based almost entirely on the intellectual know-how of Berkeley engineer Pflash Pflaumer. Like all good engineers, Pflash is constantly perfecting his creations and attempting to squeeze more out of each design. Most of us know how engineers work, suggesting one more change or one more week of development will be all that's needed. However, if I had to guess, I'd say Berkeley co-founder Michael Ritter has played a large role in getting these version updates to market.  If left to their own devices, engineers like Pflash who have all kinds of ideas, will perfect things until someone says enough and a product must go out the door. 

     

    These design ideas from Pflash take months and even years to work through and implement in actual products. Thus, the reasons for the RS1, RS2, and now RS3. We aren't talking about a resistor model number being changed or an internal part being out of production. Yes, some manufacturers release a MK2 or MK3 version if a supplied part is no longer available. Not Berkeley Audio Design. The company works through major code and design changes, and concludes all of this by ensuring sonic improvements. None of these changes would mean anything without the end result being a major sonic improvement. Based on my time with the RS3, I unequivocally conclude the RS3 is the best sounding Berkeley Audio Design product released to date. 

     

     

    RS3+Front.jpg

    RS3+Rear.jpg

     

     

     

    Listening, Comparing, Concluding

     

    I recently found artist Ted Poor's latest album, named You Already Know, via Qobuz new release recommendations (Qobuz, Tidal). Like all books I read, I judge albums by their cover. Just kidding, but scrolling through streaming services like Qobuz I tend to look at album covers when making decisions about which albums to give a few minutes of my time. Ted Poor's new album caught my eye and I'm very happy it did. 

     

    R1-06312-0009.jpgListening to the opening track titled Emilia, through the Alpha DAC RS3, I was immediately touched by the sound of Ted's drums. The delicacy, subtle sonic textures, and overall sense of realism was striking. Perhaps this realization was related to me recently attending my daughter's first drum lessons and hearing very similar sounds in the flesh. Certainly not similar artistry, but the drums on this recording when played through the Alpha DAC RS3 just sounded so dang real. With each hit of the drum head, I could hear the sonic equivalent of a color palette full of tones, overtones, fundamentals, and harmonics. These sounds were placed on a soundstage easily recognizable as that from a Berkeley DAC. The stage isn't overly expansive and nothing near closed-in. This realistic sound emanates from just to the right and left of the loudspeakers, with a good sense of height, and a three dimensional floating image in front of the listener. 

     

    Moving along to track four, titled To Rome, the Alpha DAC RS3 presented me with something incredible that I haven't heard previously. When Ted Poor hit his drums, I could hear not only the drum head but I heard deep down into the shell of each drum. This explanation may sound strange, as it's quite different from anything I've previously experienced. Yet, it's related to hearing so much of the drum's sound as described in the previous track. It's as if things are taken to a whole new level on this track through the Alpha DAC RS3. 

     

    I encourage readers to listen to this track and really focus on the drums about thirteen seconds into the song. I'm not suggesting everyone will be able to hear this, but it's as if I can tell how deep these specific drums are in inches and as if I can hear the sound going from the top drum head to the bottom, and resonating from within the entire shell. This is really something, to say the least. Heck, maybe I'll be able to tell what color the drum set is painted through the RS4 :~)

     

    Given my recent score of the entire Three Blind Mice Supreme Collection (1500), it should come as no surprise that I listened to quite a bit of this music through the Berkeley Audio Design Alpha DAC RS3. In this review I'll focus on the 1973 album titled Live! From the Terumasa Hino Quintet (TBM-17). Good luck finding it anywhere for a reasonable price. The album is an absolute gem, with superb sonics, musicianship, and music. 

     

    The opening track Stella by Starlight is a dynamic slow burn showcasing Hino as a master trumpeter. Due to what I previously heard through the Alpha DAC RS3, my ears were wide open to hearing this album again for the first time. The RS3 didn't disappoint. I honestly didn't expect to hear Hino's trumpet produce its own color palette coming from the instrument's bell, but it's all there, on display through the RS3. When I say color palette, this has zero to do with the DAC producing its own colorations and everything to do with the DAC NOT producing such colorations. The colors all stem from Hino's trumpet as he works his magic from Tokyo's Yubin-Chokin Hall on June 2, 1973. 

     

    The amount of subtlety that can be heard, in addition to the fundamental notes played loudly, from Hino's trumpet is truly special. Sure Hino can belt out the notes, but what can be heard as he eases off each note and as the sound fades into the far reaches of the jazz club, is equally as thrilling to me. 

     

    R1-06312-0008.jpgSkipping to track three on this album, titled Be and Now, the Alpha DAC RS3 delivers an experience so reminiscent of being in the club that I expected smoke to shoot from its front panel. This track involves more of the entire quintet than the first track and the sound is spectacular. Less of the trumpet subtleties can be heard but this is more than made up for by the rest of the musicians showing why this quintet was so revered. With percussion in the left channel, piano in the right, and Hino's trumpet dead-center, the soundstage is more realistic than on track one, yet the RS3 delivers each instrument as if in its own world. Each instrument is reproduced in its entirety, supremely delineated from the others, and they all come together to make a whole greater than the sum of the parts. This level of performance reminds me of the fact that DACs are much more than individual chip selections, oscillator specs, or chassis designs. The sum of the parts must be more than a single internal component that's the star of the show. When done right, as is the case with the Alpha DAC RS3, the end result is truly on another level and can transport the listener from the crises of daily life to another world of relaxation and enjoyment. 

     

    Seeking to find weaknesses in the Alpha DAC RS3, I put on some tracks that demand the best transient performance, or else they fall apart and lose their realism. One such track is [untitled] from the album May 24, 1976 by Larry Karush (Qobuz, Tidal). I first played this track with the RS3 connected to the Constellation Audio Inspiration Mono amplifier's DIRECT input. This input bypasses an internal gain stage but also requires much a higher level from the upstream component (DAC or preamp). Connected to this input, the RS3 was a bit dull on some transients. This isn't surprising and isn't indicative of a weakness but rather a design decision limiting the DAC's output voltage rather than building in a very powerful output stage that could reduce audio quality and isn't needed for 99% of the amplifiers on the market. Switching to the balanced XLR input on the same amplifiers resulted in a night and day difference. The transients from Larry Karush's piano were both striking and at the same time full of harmonics. After the initial transient attack, the harmonics that were audible throughout the sustain and decay were magnificent. This is different from other DACs I've heard in that some can handle the transient very well, but through those DACs I never realized there was additional musical information present at the same time. I guess there's something to be said for developing the Berkeley DACs using Keith Johnson's Reference Recordings material, having been present at the recordings, and also developing the analog to digital converter used for the recordings. This provides tremendous insight and enabled the Berkeley team to identify when it gets things right or wrong throughout product development. 

     

    The Berkeley Audio Design Alpha DAC Reference Series 3 has some serious competition from many manufacturers around the world. Two competing products with which I'm intimately familiar are the dCS Rossini and EMM Labs DV2 DACs. The features of these DACs are very different as are the internal topologies and the final product which is the sound heard through the DACs. The dCS Rossini is the most feature rich DAC in that it offers the most types of inputs including both USB and Ethernet, word clock in/output, Roon Ready-ness, and its own iOS app among other things. The EMM Labs DV2 is in the middle, offering a USB input, EMM fiber optic link, and like the Rossini user selectable / variable voltage output. The Berkeley Alpha DAC RS3 may be considered austere to many in that it has no USB or Ethernet input, doesn't accept native DSD or DoP material, and requires the separate Alpha USB ($1,995) digital to digital converter to sound its best. That said, the Alpha USB is really a noise isolation device that converts one interface, USB, to another, AES or S/PDIF, and will make nearly all DACs sound their best. 

     

    Internally these DACs couldn't be more different. Both the dCS and EMM Labs DACs use conversion topologies developed in-house without any off the shelf converter chips. Both DACs are known for converting incoming audio to very high rates of DSD such as DSD1024 for EMM Labs, while the Rossini can also convert to a user selectable DXD rate. The Alpha DAC RS3 accepts incoming audio up through 24 bit / 192 kHz and does its own internal up/oversampling through a SHARC DSP chip and then on to an Analog Devices DAC chip. Readers should be careful when assuming what goes on in and between these chips because in the process of developing its own silicon "chip" Berkeley discovered a way to use only the core of the Analog Devices chip in a way completely unique to the company and nearly identical to what was planned for its own chip design. 

     

    Berkeley Audio Design has always maintained that its custom filtering, developed through knowledge gained in the Pacific Microsonics days, is better than any filters in the business and a large part of what sets it apart from competitors. As an audio writer, it's hard to listen to different DACs and narrow down the differences to that of the filters. However, I can say that the Alpha DAC RS3 may be the finest DAC on the planet for playing 16 bit / 44.1 material. Given the importance of filters at this sample rate, I believe it says quite a bit about Berkeley's advanced filter claims. 

     

    R1-06312-0000.jpgThe bottom line however is how each of these DACs sound in one's audio system. I love custom everything as much as the next guy and I respect those who design products from the ground up. But, the sound quality must be as good as the competition or all is for naught. Comparing the Alpha DAC RS3 to the dCS Rossini and EMM Labs DV2 is on one hand very easy and on the other very difficult. The easy part is listening to the sonic differences. These DACs sound very different. The Rossini has a much fuller yet laid back sound signature compared to the RS3, with the DV2 in the middle but closer to the Rossini with respect to fullness. I'd say the DV2 isn't laid back at all and is likely the most forward of the three DACs. As I said in the DV2 review, I believe it's a transient monster, in that nothing reproduces the sharp edge of an attack like the DV2. However, after listening more through the RS3 I believe the DV2 is missing some of the harmonics present. This is an act of omission on the part of the DV2 and perhaps done in an effort to get the ultimate transient attack. There's no free lunch, and this may be one example. 

     

    The sound of the RS3 is what I consider raw / unaltered, equivalent to shooting RAW digital photos, compared to a touch processed. The Rossini and DV2 are a touch processed, but this shouldn't be taken to mean inaccurate or offensive. All audio is processed inside a DAC. I use the word processed as an adjective because it's the best way I can think of to describe the difference between the RS3 and DV2/Rossini. The Rossini also has a smoothness to its sound that in no way harms the overall presentation. It's one of those things that I've always noticed when listening to dCS components and have found it much less invasive than a tubed product and nothing like a sterile solid state design. 

     

    These really are different beasts that command big money. Readers would be foolish to take my word solely, without listening to each DAC in their own systems. I should also note that I'm running these components using state of the art balanced power transformers from 512 Engineering. These 5 kVA and 4 kVA transformers enable me to hear much more into each component than I previously could without the transformers. Lack of such a fine power source will greatly affect what one hears when auditioning these and all other components. It's a frustrating fact of life with which I have direct experience.

     

     

    Conclusion

     

    cash@2x.pngThe Berkeley Audio Design Alpha DAC Reference Series 3 is the company's best product to date. After overcoming some hurdles in 2018 and early 2019 developing the capability to manufacture the RS3, the company is now delivering these gems within a couple weeks of customer orders. This is great news because the Reference Series 3 is a unique and special design from a company of the same attributes. The RS3 adds to the previous RS1 and RS2 designs with more of what made them great. Years-long development has enabled Berkeley to up its game with the RS3, releasing a DAC that reveals minute textures, tones, overtones, and endless harmonic details not heard through other DACs. The RS3 reproduces the attack, release, sustain, and decay in a unique way that simultaneously reveals subtle details while recreating the immediacy of a musical event without memorialization of said event. 

     

    Music lovers with an RS1 or RS2 should, without hesitation audition the RS3. Original Alpha DAC owners should also consider that this may be the time to take their listening to another level by moving up to the Reference Series 3. I understand very well that this isn't an inexpensive upgrade, going from Alpha DAC to Reference Series 3, but if one is fortunate enough to have the means, it's a no-brainer. Those with any number of other DACs should at least give the RS3 a listen because it's so different from the competition. I really like what the RS3 does in my reference system and unequivocally recommend it to all audiophiles.

     

     

     

    R1-06312-0000.jpg R1-06312-0002.jpg R1-06312-0006.jpg

    R1-06312-0008.jpg R1-06312-0009.jpg R1-06312-0010.jpg 

    All images shot with Hasselblad 503CW, Zeiss / Hasselblad 50mm CFi lens, Kodak Ektar 100 film.

     

     

     

    Community Star Ratings and Reviews

     

    I encourage those who have experience with the RS3 to leave a star rating and quick review on our new Polestar platform.

     

     

     

     

    Product Information:

     

     

     

    Where To Buy

     

    Audiophile Style supporting dealers:

     

    Timothy W. Marutani
    Oakland, CA
    Phone: 510-652-1911
    [email protected]

     

    The Audio Salon
    Los Angeles, CA
    Phone: 310-863-0863
    http://theaudiosalon.com

     

     

    Full US Dealer list - https://www.berkeleyaudiodesign.com/usa

    Full International Dealer List - https://www.berkeleyaudiodesign.com/world

     

     

     

    Other Berkeley Audio Design Product Options

     

     

     

    Associated Music:

     

     

     

     

    Associated Equipment:

     

     

     

    Listening Room:

     

    This graph shows the frequency response of my room before (top) and after (bottom) tuning by Mitch Barnett of Accurate Sound. The standard used for this curve is EBU 3276. This tuning can be used with Roon, JRiver, and other apps that accept convolution filters. When evaluating equipment I use my system with and without this tuning engaged. The signal processing takes place in the digital domain before the audio reaches the DAC, thus enabling me to evaluate the components under review without anything changing the signal further downstream. 

     

    Chris-Alexia-Series-2-inroom-response-before-and-after-DSP.jpg.feaa6cf23ba92764f7594ab55bfd936d.jpg



    User Feedback

    Recommended Comments



    GREAT write-up Chris. I love how they are sticking to their guns, despite no doubt losing sales to those who need/want the latest and greatest three letter acronym flavor of the month. Not an easy thing to do - similar perhaps to Einstein's insistence that running more than 4 transistors per channel mucks up the sound and thus lowers the resolution and transparency to the source. I'm sure there's many other similar companies that chose to stay on their straight and narrow path.

     

    Good for them!

     

     

    Share this comment


    Link to comment
    Share on other sites
    40 minutes ago, ted_b said:

    ?  DSD is hardly a flavor of the month.  Been around for years, and a recording format chosen by some very accomplished folks in the industry.  Just sayin'.  :)

    Unless it uses a R2R ladder or similar multi-bit scheme which I don't believe it does, whatever chip it uses is converting incoming PCM data to a one bit stream and then modulating it. This would be internal DSD conversion. Best not to get caught up in an acronym discussion. I know this is an over-simplification but my guess is that the sound quality is more in the power supplies, analog components and engineering implementation.

    Share this comment


    Link to comment
    Share on other sites
    48 minutes ago, ted_b said:

    ?  DSD is hardly a flavor of the month.  Been around for years, and a recording format chosen by some very accomplished folks in the industry.  Just sayin'.  :)

     

    You can sum up DSD in two words. Nobody cares.  That being said I applaud them for sticking to their guns. After all lots of us still get good results with a Pacific Microsonics Model One.

     

    Chris did a nice job here.

    Share this comment


    Link to comment
    Share on other sites
    8 minutes ago, Rt66indierock said:

    You can sum up DSD in two words. Nobody cares. 

    Nice comment...thanks for quoting me. 

    Share this comment


    Link to comment
    Share on other sites
    29 minutes ago, The Computer Audiophile said:

    Don't be a jerk.

     

    Chris on March 4th I made a 25 minute presentation to a Southern California marketing company about high resolution audio. Part of the presentation was the numbers aren't there for DSD.

    Share this comment


    Link to comment
    Share on other sites
    Quote

    I could hear not only the drum head but I heard deep down into the shell of each drum. 

     

    Often times when I have a revelation like this it turns out I can hear it with other components, not just the one where I discovered it. So do you hear this on your Rossini now?

     

    BTW, I hear not only down into the drum, but also the echo from the floor below so I can tell how far it is off of the floor 😃

    Share this comment


    Link to comment
    Share on other sites
    17 minutes ago, Rt66indierock said:

     

    Chris on March 4th I made a 25 minute presentation to a Southern California marketing company about high resolution audio. Part of the presentation was the numbers aren't there for DSD.

    None of that matters with respect to your comment. Ask Chad (Acoustic Sounds) about how many of his customers want DSD. In the grand scheme of life it’s a small niche, but then again so is golf, given that a tiny percentage of Earth’s 7.7 billion people care about it. 

    Share this comment


    Link to comment
    Share on other sites
    38 minutes ago, The Computer Audiophile said:

    None of that matters with respect to your comment. Ask Chad (Acoustic Sounds) about how many of his customers want DSD. In the grand scheme of life it’s a small niche, but then again so is golf, given that a tiny percentage of Earth’s 7.7 billion people care about it. 

     

    Just the golf equipment market is over $8 billion annually worldwide. 

    Share this comment


    Link to comment
    Share on other sites
    1 hour ago, ted_b said:

    Nice comment...thanks for quoting me. 

     

    I use DSD sometimes as a sound effect for some harmonicas and a few resonator banjos. But it is very hard to work with in a studio.

    Share this comment


    Link to comment
    Share on other sites
    2 minutes ago, Rt66indierock said:

     

    I use DSD sometimes as a sound effect for some harmonicas and a few resonator banjos. But it is very hard to work with in a studio.

    Ummm, lemme think....I don't care.

    Share this comment


    Link to comment
    Share on other sites
    18 minutes ago, The Computer Audiophile said:

     

    You can sum up golf in two words. Nobody cares.

    Them’s fighting words around these here parts. LOL🤪

    Share this comment


    Link to comment
    Share on other sites

    @The Computer Audiophile --

     

    Do you see DAC's on a broader scale ever (still) progressing sonically over the years, and if so what in general do you assess to be the reason(s) for this - DAC chip development, jitter suppression, the analogue output stage, PSU, other, a combination of all/overall implementation? 

    I guess a follow-up question to this could be: would you rather acquire an older, used "reference" DAC, or a new mid-level priced DAC (that, just for the sake of simplicity and to support this thought experiment, we assume costs about the same as an older, used reference DAC)?

    Share this comment


    Link to comment
    Share on other sites

    Chris, I find it interesting you made no mention of MQA playback and how the DAC handles it. Berkeley were big promoters of MQA and how wonderful they said it sounded a few years ago-essentially they agreed it sounded as good as the original, or even better. Even to the extent of coming online and arguing with MQA detractors about it.

    They also emphasize their implementation of MQA in the product description. 

    Share this comment


    Link to comment
    Share on other sites
    3 minutes ago, firedog said:

    Chris, I find it interesting you made no mention of MQA playback and how the DAC handles it. Berkeley were big promoters of MQA and how wonderful they said it sounded a few years ago-essentially they agreed it sounded as good as the original, or even better. Even to the extent of coming online and arguing with MQA detractors about it.

    They also emphasize their implementation of MQA in the product description. 

    My thoughts on MQA, both the technology and company, are well known and needn't be discussed in this review. 

    Share this comment


    Link to comment
    Share on other sites

    Chris,

     

    I assume that you fed the DAC with the AES from the W20SE?  Did you use the Alpha USB from another source?  Just curious about architecture.  

     

    Bob

     

    PS thanks for the links to the music!  

    Share this comment


    Link to comment
    Share on other sites
    32 minutes ago, bobfa said:

    Chris,

     

    I assume that you fed the DAC with the AES from the W20SE?  Did you use the Alpha USB from another source?  Just curious about architecture.  

     

    Bob

     

    PS thanks for the links to the music!  


    Here are three common ways I used the RS3. 

     

    Sonore Signature Rendu SE optical > USB > Alpha USB > AES > RS3

     

    W20SE > USB > Alpha USB > AES > RS3

     

    W20SE > AES > RS3

    Share this comment


    Link to comment
    Share on other sites
    10 hours ago, The Computer Audiophile said:

    My thoughts on MQA, both the technology and company, are well known and needn't be discussed in this review.


    Great review Chris and I’d love to audition, but from the prices you quote, why the huge price jump from series 2 to 3?

     

    But given the prominence given in the review to 1. the fact that Berkeley eschew USB, DSD and Ethernet and 2. their technical and sonic chops, their inclusion of and advocacy for MQA is surely worthy of comment.  One simple conclusion that could be drawn from this is that their expertise tells them MQA is worthy of inclusion.  There are alternate conclusions of course, but the inclusion of MQA by engineering led companies like Berkeley and DCS is a powerful counterpoint to forum opinion.

    Share this comment


    Link to comment
    Share on other sites
    2 hours ago, Norton said:


     There are alternate conclusions of course, but the inclusion of MQA by engineering led companies like Berkeley and DCS is a powerful counterpoint to forum opinion.

    Maybe. Or it just means they had customers demanding it and decided to include it. Not everything carries a special meaning.

     

    In any case, even if a company like Berkeley likes it, I don't give their opinion any more or less weight than that of any other experienced audiophile. There's one or two people at such a company that make such a decision, and I don't see any reason to automatically defer to their opinion. Especially when there are other "engineering led" companies that have come to the opposite conclusion....Dueling experts....

    Share this comment


    Link to comment
    Share on other sites
    54 minutes ago, firedog said:

    Maybe. Or it just means they had customers demanding it and decided to include it.


    That could well be a general case for the inclusion of MQA, but it’s an unconvincing one in the specific case of Berkeley who seem not that interested in responding to customer demand in the case of USB, DSD or Ethernet, although I guess at this rarefied level “demand” for anything is likely to mean a handful of individuals.  Inclusion of MQA is an odd outlier, given Berkeley’s approach, unless they see something of genuine merit in it, or else a more cynical explanation is attributed.  The former seems the simplest and therefore most likely explanation.

     

     

    Share this comment


    Link to comment
    Share on other sites



    Create an account or sign in to comment

    You need to be a member in order to leave a comment

    Create an account

    Sign up for a new account in our community. It's easy!

    Register a new account

    Sign in

    Already have an account? Sign in here.

    Sign In Now



×
×
  • Create New...