Published: 2. March 2013, No. 106
Peterborough, Ontario, November 14th, 2012 – Bryston LTD (Bryston Limited: Music for a Generation), builders of the world’s finest audio electronics and celebrating 50 years in 2012, has announced the introduction of the BDA-2 outboard digital to analog converter (DAC).
Bold words, indeed. But then, they come from no less an authority than Canadian Bryston, one of the better-known audio brands both in the world of consumer as well as professional audio. Similarly to such British manufacturers as ATC, Harbeth, Tannoy, PMC, dCS, or Dynaudio from Denmark, Bryston Limited is a company equally at home on both sides of the “glass”. It is also one of the few successful manufacturers of electronics; almost all the others, apart from dCS, mainly offer speakers.
Recording studio provenance of audio equipment usually guarantees one thing, in addition to its user friendliness: high manufacturing quality and thus long equipment life. Satisfied owners of twenty years old or older Bryston amplifiers will be happy to confirm that. But what about Bryston DAC owners? Well – they have not yet had enough time to find it out as the BDA-2 predecessor, the BDA-1, was actually the first DAC offered by the manufacturer since 2010.
The new BDA-2 is not meant to be its direct replacement but rather a product line supplement. The immediate reason for its launch was the need to upgrade the USB interface, previously limited to 16-bit, 48 kHz signal. With the new model we get an advanced, modern solution based on an XMOS chip, offering asynchronous transmission of signal up to 24-bit, 192 kHz resolution. USB is the domain of computers and thus difficult to manage. The BDA-2 arrives with its own driver on a flash drive in the shape of a key. During the review, however, it did not yet support Windows 8 that I use. Usually, Windows based computers limit playback to signals to 24/96 without a dedicated driver. Here it is different – the input is completely inactive until we install an appropriate driver from Bryston. Thus, I had no means to test the DAC's USB interface.
The BDA-2 also features an improved power supply section, with two independent power supplies for digital and analog sections and independent voltage controllers for each stage in the digital chain. Likewise, the DAC chip has been upgraded: in place of the previous 24-bit delta-sigma Cirrus CS4398 with 128-times oversampling the new Bryston incorporates superior dual AKM 32-bit chips. Digital inputs are isolated with impedance matching transformers and the signal is over clocked to minimize jitter. The output stage operates in class A and is fully balanced.
A special care has been given to signal upsampling. The technique, very popular a few years ago, consists in re-clocking the 44.1 kHz (or 48 kHz) input signal to get 192 kHz signal. As is known, the sampling rate determines the signal upper frequency range, which is 22.05 kHz for 44.1 kHz and 96 kHz for 192 kHz sampling rate accordingly. Upsampling does not add anything to the signal so the new 192 kHz signal is still bandwidth limited to 22.05 kHz. It is similar with a change in the word length. Although upsampling circuits are called “Sample Rate Converter”, the process also lengthens the word from 16-bit to 24-bit. The amount of information is not increased; just added 8 "empty" bits. Why all this? It was assumed that the upsampled 24/192 signal, although carrying exactly the same information as the 16/44.1, was much more DACs “friendly” in that DAC chips worked better with so treated signal.
Much has changed, however, over the last several years and today we know much more about digital audio. It turned out that the upsampler effectively worked by minimizing jitter – after all, it is an asynchronous converter with signal re-clocking. Doesn’t it remind us of something? How about an asynchronous USB port? Right there…
Upsampling is a complex mathematical process. And according to signal theory, but also proved in practice, it changes the sound. With discovering other ways of eliminating jitter, e.g. by improving the signal, the effect of upsampling on the sound became more and more pronounced and not necessarily positive. Bryston approached the problem very seriously, as the upsampling in the BDA-2 is SYNCHRONOUS – 32, 48 and 96 kHz signals are upsampled to 192 kHz, and 44.1 and 88.2 kHz to 176.4 kHz. The circuit can be turned on or off according to listening preferences.
A selection of recordings used during auditions:
- MJ Audio Technical Disc vol.6, Seibundo Shinkosha Publishing, MJCD-1005, CD (2013).
- Abraxas, 99, Metal Mind Records, MMP CD 0102, CD (1999).
- Anita O’Day, All The Sad Young Men, “Original Collection 50”, Verve (Japan), POCJ-2761, CD (1961/1999).
- Bonney M., Sunny, BMG Ariola, 24931, CD (1995).
- Dave Brubeck Quartet, Time Out, Columbia/Sony Music Entertainment Hong Kong, 83532, No. 55, K2HD CD (1959/2011).
- floating.point, Free Falling, Piotr Sczepaniak, CD-R (2010);
- Frank Sinatra, Only The Lonely, Capitol/Mobile Fidelity, UDCD 792, gold-CD (1958/2002).
- Handel, Arias for Cuzzoni, Simone Kermes, Lautten Compagney Berlin, dyr. Wolfgang Katschner, Edel Classic, 16422BC, CD (2009).
- Jimmy Giuffre, Western Suite, Atlantic/Warner Music Japan, WPCR-25160, “Atlantic 60th”, CD (1958/2006).
- Marek Biliński, Mały Książę, Bi.Ma., BiCD-09, CD (2010).
- Megadeth, Countdown to Extinction, Capitol/Mobile Fidelity Sound Lab, UDCD 765, gold-CD (1992/2006).
- Michael Jackson, Thriller. 25th Anniversary Edition, Epic/Sony Music Japan, EICP-963-4, CD+DVD (1982/2008).
- Miles Davis, Kind of Blue, Columbia/Sony-Legacy, 480410, “Master Sound”, Super Bit Mapping, CD (1959/1995).
- Muse, The Resistance, Warner Music Japan, WPZR-30355-6, CD+DVD (2009).
- Slow, Art of Silence, Proa Records, 194000, CD (2011);
- The Tokens, The Lion Sleeps Tonight, RCA/BMG, 66510, CD (1994).
- The Shadows, Greatest Hits, EMI, 81982, CD (2004).
Japanese editions available from
How to make the musical event taking place either in the recording studio or live at stage and then its recording and reproduction resemble each other as much as possible? Or how to best transfer musician’s ideas so when played back at home they’d be closest to what the artist had in mind? These are fundamental questions in the audio world. Playback of a recording at home is in fact the final stage of a long journey during which anything can happen. Is it even possible? Is it possible to “bring” musical events to our home in 1:1 scale? I think it’s pretty clear that this a myth, something we will never be able to achieve. Mainly because music recording, mixing, mastering, media preparation and manufacturing are all creative process, where there is no single “proper” way or “golden” rule, and everything is based on taste, skills, capabilities, money, objectives, series of events and coincidence.
Playing back the recording at home is the final act and the final interpretation of the music material. It cannot be approached differently if, of course, we want to understand the presentation. I am not concerned with the silly “what the poet had in mind” (the sound engineer/producer/musician/editor/engineer/coincidence), but with how WE understand what is in front of us, what WE think while listening to the music track. I think that’s exactly what Jason Victor Serinus had in mind when he wrote in his editorial Music? Or Sound?:
“Beyond all those specifics, when you take a deep breath and let the music flow over you, does what you hear make musical sense or does it seem unbalanced? Does the music move you, conveying the emotion you sense the composer intended to communicate? How does it make you feel?”
(Jason Victor Serinus, Music? Or Sound?, “Stereophile”, February 2013, p. 3.)
What needs to be known is that similarly to people involved in music recording and production, each audio device also interprets the signal being sent to it. We are used to think of this process as distortion. And that’s right – in the end it comes down to distortion. However, since the signal is distorted by EVERYTHING, including ourselves, sometimes it is worth looking at it from another angle, as a somehow desired effect, as a result of conscious choices, as an “author’s compromise”.
If I were to illustrate it in some way, the BDA-2 DAC from Bryston would be a model example. Costing pretty good money, the device has its own “sound” or, rather, it modifies the signal after its own fashion. It can be assumed that it is a common feature of high quality audio devices designed for both home and studio systems.
Traditionally, studio equipment has been received by the audiophile community as being sterile, dry and devoid of emotion. At the same time precisely defining sound attack, dynamic and selective. Based on my own experience, I must admit that this is largely true. It does not, however, apply to all devices as there has also been a group of expensive products whose sophisticated presentation differ from that stereotype.
Like the sound of the Bryston. Be that a result of contact with environment completely different than the pro audio world or simply its design maturity, what’s important is that in the reviewed DAC I heard both extensive experience and personal choices related in a much nuanced way to the so-called “accurateness”, i.e. “objective” balance.
Narrowing its description to several basic terms, I would say that its sound is very resolved, fairly well saturated and it keeps fantastic timing. Blurring of attack time, lack of a perfect sense of punch is a problem of many products, regardless of their price. I understand the trade-offs leading to that. However, once we listen to something like the Bryston, where a music track has its own internal logic resulting not only from its tune, a sequence of phrases, timbre, etc., but also from its rhythm, subcutaneous pulse, we will always look for that in other devices. And even if we agree to give up this aspect in exchange for something else, a small room will remain in our mind, where behind closed doors that memory will be waiting for its moment. And that will come, sooner or later.
For the ability to define individual instruments, events and echo in the whole mix is here outstanding. It can be can be appreciated both on exceptional versions of outstanding records, such as We Get Requests by The Oscar Peterson Trio on the stunning UDC First Impression Music release, as well as stuff that no self-respecting music lover, no reputation-conscious audiophile would be seen dead listening to, for example: Sunny by Boney M., The Lion Sleeps Tonight by The Tokens and The Chordettes by the group of the same name. Regardless of music genre, its “weight” and value, the ability of the Canadian DAC to render individual elements in time, to prevent their blurring, translated into an incredible sense of the “here and now”, usually achieved by appropriate color manipulation. Here, one has a sense of “becoming” of events, not their mechanical, lifeless playback.
I have mentioned the stereotype concerning studio equipment; one of its aspects is a belief that such devices are characterized by strong, sharp treble and low, contoured bass. As it usually happens, that stereotype came from the observation of reality and automatically extrapolating it to all devices of this type. The Bryston case is really very interesting. Its highest reble is slightly hidden, just like its bottom end. A comparison with the reference player showed it immediately. Does that mean that the BDA-2 favors midrange, that its sound is colored in the fashion of tube devices? Absolutely not! What I'm talking about is a part of this device’s own nature on which the engineers, the way I see it, worked really hard.
In order to do justice to what I heard, I have to refer to an example related to recording sessions’ techniques. Human voices and acoustic instruments are recorded by microphones. That also applies to electric instruments in combination with specialized amplification systems (e.g. guitar amps), such as electric guitars, Hammond organ, and others. It has become customary to treat the microphone as a “window” through which we “see” what is in front of us. It seems to me, however, that what would be much closer to reality is a comparison with a magnifying glass. The microphone magnifies or reduces the recording element – it all depends on how far/close it is located from the sound source (and is a result of microphone properties). The vast majority of recordings are made from a very short distance; a few inches from the piano, voice or trumpet is the norm. This translates into a larger instrument, into weighing down its color, discernible even after frequency correction. Such are almost all jazz recordings from the 50s and 60s, e.g. (for illustration purpose only) Western Suite by Jimmy Giuffre or Anita O'Day’s All The Sad Young Men. It’s even clearer, however, on new productions, even very good – for example The Art of Silence by the duo Slow.
Bryston BDA-2 shows these recordings closer to what we would hear live. It delicately lightens lower midrange, gently withdraws bass, and shows everything in a slight distance. We are used to a close sound, 'tangibility' has become one of the attributes of a good sound – and that’s OK. However, the Canadian device redefines what is in the signal, as if re-constructing a real event, trying to jump over the recording and production process. It's an amazing experience, worth enjoying even if you believe (as do I) that the main task of any audio component is a faithful rendition of what is on the disc/file. The BDA-2 slightly distances the sound, at the same time offering lots of details, very good definition and differentiation. It is a dense sound, which is better audible with a media files player as the source.
Recordings made from a considerable distance, such as most classical music, were also slightly distanced. The sound was a bit smaller than from the reference player. But it was not as pronounced as in the case of close miking; these effects were not symmetrical.
The DAC behaves slightly differently when connected to a media files player. The sound becomes deeper and warmer; higher midrange slightly quieter. Low midrange and some bass are strong, full, saturated, making all the events on the stage seem large. Not as big as from the reference CD player but large nevertheless due to this weighing down the bottom.
The sound is not quite as precise as before but we get something that is rather rarely associated with the files, i.e. tangibility, reliability. Almost all the file players, apart from those most expensive, suffer from a kind of anorexia and detachment from reality. As if they could not convey emotions contained in the recordings. Thus they would not meet the basic postulate suggested by Jason Victor Serinus, cited above. Bryston goes against the current with that slight coloration, with shifting the accent down. Recordings are a pleasure to listen; the benefits of moving to 24-bit are audible, be that in the form of a deeper color or by adding the instruments bodies some "seriousness". Differences in sampling frequency were not as clear as could be but apparently one can’t have everything.
Listening to audio devices from the point of view of our system, in other words wondering whether to buy them or not, it is worth first examining our conscience and think about what we really expect from a given product. What is important in the assessment, of course, is whether the sound is of a high order, in any price range, but in the end what really matters is how comfortable we feel with the device.
Bryston BDA-2 offers things that other products would be proud of, if only they could boast of them. First of all, it keeps the coherence of presentation, fantastically defining time in music. This in turn leads to a lack of any internal tension given to music by the device. Without slowing down, rounding off, which can be nice but only in the short term, its color is difficult to clearly define. On the one hand, bass and treble can sound really strong, especially with a files player and a computer, but on the other, the presentation is a little distant, which subjectively seems to favor clearer midrange. Yet this is not an intentional color manipulation but rather the result of other choices. Resolution is excellent and only much more expensive digital devices are capable of an internally richer, more nuanced presentation. It’s worth pairing the Bryston with a saturated system; nothing will be taken away and we will get a “spring” that drives all that machinery.
Very interesting effects can be achieved by using upsampling technology. As previously mentioned, it is an operation on the signal that changes it irrevocably (not only due to a change of sampling rate and word length). In the BDA-2 from Bryston it is immediately audible, which reflects well on its capabilities in terms of differentiation or resolution, but also on the fact that the signal is well treated, i.e., has low jitter. Pressing the "upsampling" button adds weight and mass to the sound, directing listener’s attention to midrange and lower. So anyone missing a bit of weight may well try this option. At the same time, however, the sound becomes a bit "plastic" and less defined, with slightly silenced upper midrange. It is of course up to the listener to decide and it’s great to have at our disposal this type of digital filter, shaping the sound. For me, though, the DAC sounded unequivocally better with this button’s indicator off.
This is a very successful device nicely showing what really matters to the people involved in recording studios and on stage, and also having contact with the other side of the "glass" – the end users. A superbly equipped and built DAC that will serve its owner for many, many years.
DACs are usually small. They do not need very efficient and large power transformers, there are no moving parts (e.g. a motor drive), nor do they emit too much heat. Bryston BDA-2 DAC is no exception, although its enclosure has been designed to be as solid as possible. Hence, a standard width thick aluminum front panel and a rigid stainless steel chassis – low but quite deep.
The device is operated with small buttons sporting small LED indicators above. Green/red LEDs! Finally, there are no blue ones! Green indicates the synchronization of input with the transmitter (transport); red show lack of synchronization. As if that was not enough, there is an additional "Lock" LED indicator that serves the same purpose. The buttons are used to change the active input, to switch upsampling on/off and to power the device on/off. The “Lock” LED is located in one of the two rows of LEDs indicating sample rate. The DAC accepts signal up to 24-bit with sample rates of 44.1, 48, 88.2, 96, 176.4 and 192 kHz. But it does not accept DSD signal. There is also no indication of bit depth.
Bryston uses its own proprietary techniques to reduce signal jitter, including reclocking and upsampling. Unlike other manufacturers, it employs synchronous upsampling, separate for the 44.1 kHz sample rate family (orange LED) and the 48 kHz (green).
The device is equipped with a wide array of inputs and outputs. It sports eight digital inputs: two optical TOSLINK, two RCA, two BNC, one AES/EBU and one USB port. The latter is asynchronous 24/192. There is also a digital RCA output and two pairs of RCA analog outputs: unbalanced RCA and balanced XLR. All connectors look very solid and come from Neutrik. There is also a RS232 port for firmware upgrade and a remote 12 V trigger. Mains connector is a classic EIC 16 A. The DAC sits on large rubber feet. It’s worth placing it on feet with ceramic balls.
Bryston is a modern company, no garage (with all due respect and love), one-man manufacturer. Accordingly, its electronics is built using surface mount technology. Only certain components are through hole – mainly capacitors and voltage controllers; in other words: the power supply section.
The printed circuit board occupies less than half of the interior; the enclosure is oversized in order to move a large toroidal transformer away from the PCB and to provide mechanical stability for the circuits.
The inputs are coupled via impedance matching transformers. The USB port features a large US1219 XMOS chip with its own high quality clock. Right next are two more clocks: separate for 44.1 kHz and 48 kHz and its multipliers respectively.
After selecting the active input the signal goes to the upsampler, the SRC43921 from Burr Brown. The next step is a digital to analog conversion circuit. It is based on two large, very nice AKM AK4399 Audio4pro series converters, one for each channel. They work with 32-bit 192 kHz signal. Finally, the analog section – nice, separate paths for each channel. It is fully discrete, built on SMD components, with really large transistors in the output stage. Each stage has separate voltage controllers. The whole power supply is quite complex, with separate circuits for the analog and digital sections. Overall, it is a very solid device.