Bryston BDA-3 D/A processor Measurements

Sidebar 4: Measurements

I measured the Bryston BDA-3 with my Audio Precision SYS2722 system (see the January 2008 As We See It"). As well as the Audio Precision's digital outputs, I used WAV and AIFF test-tone files sourced via USB from my MacBook Pro running on battery power with Pure Music 3.0. Apple's USB Prober utility identified the Bryston DAC as "BDA3 Audio 2.0" from "Bryston," and revealed that its USB port operated in the optimal isochronous asynchronous mode. Apple's AudioMIDI utility revealed that, via USB, the BDA-3 accepted 24-bit integer data sampled at all rates from 44.1 to 384kHz. I didn't test the BDA-3 with HDMI, DSD, or Ethernet data.

The maximum output levels at 1kHz were as specified, at 4.05V balanced and 2.05V unbalanced. Both outputs preserved absolute polarity (ie, were non-inverting). The unbalanced output impedance was low, at 72 ohms at high and middle frequencies, increasing inconsequentially to 82 ohms in the low bass. The balanced output impedances were twice those values, as expected.

The impulse response with data sampled at 44.1kHz (fig.1) revealed that the digital reconstruction filter used by LG was a minimum-phase type, with all ringing occurring after the single sample at 0dBFS. Wideband analysis of the BDA-3's output while it reproduced 44.1kHz-sampled white noise at –4dBFS (footnote 1) indicated that this filter rolled off steeply above 20kHz, with the ultrasonic image at 25kHz of a full-scale 19.1kHz tone suppressed by more than 100dB (fig.2). Note that the distortion harmonics of that tone all lie at –104dB or lower (0.0006%). This DAC is superbly linear.


Fig.1 Bryston BDA-3, impulse response (one sample at 0dBFS, 44.1kHz sampling, 4ms time window).


Fig.2 Bryston BDA-3, wideband spectrum of white noise at –4dBFS (left channel red, right magenta) and 19.1kHz tone at 0dBFS (left blue, right cyan), with data sampled at 44.1kHz (20dB/vertical div.).

Fig.3, taken with data sampled at 44.1, 96, and 192kHz, is a more orthodox means of showing the Bryston's frequency response. With each sample rate, the BDA-3's output conforms to the same gentle rolloff above 10kHz, with the response down by just 0.25dB at 20kHz, but then a sharp rolloff just below half of each rate. With USB data sampled at 384kHz, the output was down by 6dB at the same 80kHz as with 192kHz data, but continued the gentle rolloff to reach –13dB at 120kHz. Channel separation was superb, at 125dB at 1kHz, and still 113dB at the top of the audioband.


Fig.3 Bryston BDA-3, frequency response at –12dBFS into 100k ohms with data sampled at: 44.1kHz (left channel green, right gray), 96kHz (left cyan, right magenta), 192kHz (left blue, right red), 384kHz (left green, right gray) (0.5dB/vertical div.).

Fig.4 shows spectral analysis of the BDA-3's noise floor with a 24-bit, 1kHz tone at 0dBFS (blue and red traces) and –60dBFS (green, gray). Other than a vanishingly small amount of 120Hz hum (at –147dB!), the spectra are free from spuriae. Note that, as the Bryston has such a low level of self-noise, I have increased the vertical scale in this graph from my usual –150 to –160dBFS, and have also done so for the spectra of the Bryston's output as it reproduced a dithered 1kHz tone with 16- and 24-bit AES/EBU data (fig.5). (The result was the same with USB data, fig.6.) The increase in bit depth drops the noise floor by almost 30dB, meaning that the BDA-3 offers close to 21-bit resolution—one of the best I have encountered. Note also that the spectra are free from any harmonic distortion.


Fig.4 Bryston BDA-3, spectrum with noise and spuriae of dithered 24-bit 1kHz tone at 0dBFS (left channel blue, right red) and –60dBFS (left green, right gray) (20dB/vertical div.).


Fig.5 Bryston BDA-3, spectrum with noise and spuriae of dithered 1kHz tone at –90dBFS with: 16-bit data (left channel cyan, right magenta), 24-bit data (left blue, right red) (20dB/vertical div.).


Fig.6 Bryston BDA-3, spectrum with noise and spuriae of dithered 1kHz tone at –90dBFS with 24-bit data via USB (left blue, right red) (20dB/vertical div.).

With its very low noise and correspondingly high resolution, the Bryston's reproduction of an undithered 16-bit/1kHz tone at exactly –90.31dBFS was essentially perfect (fig.7): The three DC voltage levels described by the data and the minimum-phase behavior of the reconstruction filter are clearly visible. With undithered 24-bit data, the result was an almost noise-free sinewave (fig.8), despite the very low signal level.


Fig.7 Bryston BDA-3, waveform of undithered 1kHz sinewave at –90.31dBFS, 16-bit data (left channel blue, right red).


Fig.8 Bryston BDA-3, waveform of undithered 1kHz sinewave at –90.31dBFS, 20-bit data (left channel blue, right red).

Fig.2, taken into a high 100k ohms, has already indicated that the BDA-3 offered a very low level of harmonic distortion; repeating the analysis with a full-scale low-frequency tone into a punishing 600 ohms confirmed that indication (fig.9). In fact, though the third harmonic is the highest in level, it still lies at just –100dB (0.001%).


Fig.9 Bryston BDA-3, spectrum of 50Hz sinewave, DC–1kHz, at 0dBFS into 600 ohms (left channel blue, right red; linear frequency scale).

All of the measurements so far were taken without upsampling. Repeating them with AES/EBU data and upsampling selected produced no differences. (The upsampling wasn't operative with USB data.) However, when I used an equal mix of 19 and 20kHz tones, sampled at 44.1kHz and with the peak level of the waveform reaching 0dBFS, the aliasing images of the two tones at 24.1 and 25.1kHz, which had been present without upsampling (fig.10), disappeared as expected when upsampling was selected (fig.11). And unlike with the upsampling function in Bryston's earlier BDA-1, the noise floor remained clean. Either way, actual intermodulation distortion was vanishingly low in level.


Fig.10 Bryston BDA-3, HF intermodulation spectrum, DC–30kHz, 19+20kHz at 0dBFS into 100k ohms, 44.1kHz data (left channel blue, right red; linear frequency scale).


Fig.11 Bryston BDA-3 with upsampling, HF intermodulation spectrum, DC–30kHz, 19+20kHz at 0dBFS into 100k ohms, 44.1kHz data (left channel blue, right red; linear frequency scale).

Tested for its rejection of word-clock jitter using an AES/EBU connection and 16-bit J-Test data, the BDA-3's output was free from data-related spuriae, and all the odd-order harmonics of the low-frequency squarewave were at the correct levels (fig.12, green trace). With 24-bit data, the harmonics disappeared as expected and the noise floor was free from spuriae of any kind (fig.13).


Fig.12 Bryston BDA-3, high-resolution jitter spectrum of analog output signal, 11.025kHz at –6dBFS, sampled at 44.1kHz with LSB toggled at 229Hz: 16-bit AES/EBU data (left channel blue, right red). Center frequency of trace, 11.025kHz; frequency range, ±3.5kHz.


Fig.13 Bryston BDA-3, high-resolution jitter spectrum of analog output signal, 11.025kHz at –6dBFS, sampled at 44.1kHz with LSB toggled at 229Hz: 24-bit AES/EBU data (left channel blue, right red). Center frequency of trace, 11.025kHz; frequency range, ±3.5kHz.

With its extremely low levels of noise, harmonic and intermodulation distortion, and its superb resolution, the Bryston BDA-3 offers measured performance that is as good as digital can get.—John Atkinson

Footnote 1: My thanks to Jürgen Reis of MBL for suggesting this test to me.
Bryston Limited
US: Bryston Service USA
30 Coventry Street
Newport, VT 05855
(802) 334-1201

georgehifi's picture

This could be one that pulls me away from Multibit, very nice measurements JA must be one of the better ones I've seen.
Pity/shame Bryston didn't see fit to include the digital VC that the AKM dac is capable of. Hopefully in the BDA-3 MkII??

"Output Volume (PCM, DSD) The AK4490 includes channel independent digital output volumes (ATT) with 256 levels at 0.5dB step including MUTE. This volume control is in front of the DAC and it can attenuate the input data from 0dB to –127dB or mute. When changing output levels, it is executed in soft transition thus no switching noise occurs during these transitions. It takes 7424/fs from FFH (0dB) to 00H (MUTE). The attenuation level is initialized to FFH by initial reset. Register setting values will be kept even switching the PCM and DSD modes."

Cheers George

MusicEar's picture

Using the internal volume inside the DAC means DSD has to be converted to PCM(according to AKM data sheet) which negate the effect of DSD recordings. Besides, it is preferred to use the pre-amp volume as opposed to digital to improve the SNR for low level details.

georgehifi's picture

DSD does nothing for me because there's no music available that I like that is "Native DSD".
Give me DXD (pcm) if in the future it can get the music content I like.

Cheers George

MusicEar's picture

Check out pure DSD recordings direct from DSD masters

strettonufo's picture

Really curious how playing back music files in all formats supported by the DAC via HDMI compares with playback via USB. Disappointed that the review didn't get into this.

ToeJam's picture

Thanks for the Bryston review. If it had MQA I'd consider purchasing it.

Will you soon be reviewing the Mark Levinson components used in this review? I'm interested in the 536 and 526.

enrique majluf's picture

Hi everyone, maybe it's silly what I'm going to ask. I request help to know if the Bryston BDA-3 being turned off can transmit video passthrough HDMI thorought. That is from a blu ray connected via HDMI to the BDA-3 and then from this to the TV / LED?


sfage's picture

“Hey man. What’s that box do?”

Let’s pretend it’s the late 60s, early 1970s. Bell-bottoms, long-ish hair, mom and dad had the console stereo and television extravaganza replete with the obligatory liquor cabinet. Rock and roll, right? Colour television? Fancy! Then, we all had that one friend that was an audio nut. So we called him / her up and asked where to begin. “Well, since it’s the start of the signal chain, you’re going to need a good turntable with the best needle you can afford.” So, the two of you got in the car and drove down to the audio shop. You’ve decided on a turntable, and now you begin to go down the signal chain. Preamplifier? Yup, I understand that. Then there’s another box and all it has on it is a power switch, a light bulb and not much else. “Hey man. What’s that box do?”

Flash forward to today, and guess what? Not much changes.

I recently purchased the Bryston BDA-3. Was it a financial tick? Sure. Good stuff costs money. But, look at it this way: a lot of people think of a DAC as an add-on. An after-thought. Well, it’s not. It’s like that 1973 turntable. It’s the start of the signal chain. It basically “is” that turntable.

The price is up there... but it won't sink your battle ship.

The BDA-3 is relatively attractive as far as DACs go. It has a serious face plate and chassis that’s made out of real metal. So far so good. It has a bunch of lights on it. Tells you what it’s doing. Groovy. There is a metric tonne of connectivity. There’s a whole bunch of HDMI, USB, optical, coaxial and other stuff. I can’t think of what one could possibly need beyond that.

Is it eye candy? I dunno. The point of the exercise, here, is what the thing does. Right? “You must ween yourself off the eye-candy roller-coaster, grasshopper.”

The BDA-3 is musical. That’s the heart of the matter. It’s very musical. The sound stage, depending upon the recording, is really excellent. For example, I was listening to Macy Gray’s “Stripped” last night. The presentation of that recording is old school. It’s “as if” the engineer had two microphones and a quarter inch mastering reel at 15ips (the truth is, it’s a binaural recording). But, that’s it, baby. I was going to say “awesome” but that’s not really 1970. In the 1970s, the word awesome was taken literally. Hmm. Wait. The sound stage is awesome. Never mind. I’ll stick with that.

Time to move on. It’s 2018, right?

My academic background is music composition. I have written for all manner of ensembles in all sorts of configurations. It was my job: make sounds that work, and don’t turn in to a traffic jam for orchestra. While that may not matter to most people, it matters to me. I’m the guy standing there in dress rehearsal with the score in hand and a pencil. So, I am listening to the BDA3… like that.

Yeah, I’m grumpy… and the violin section is annoyed because it doesn’t sound like Mozart. Did they practice their part before they came to rehearsal? Well… umm… you know… NO.

I flipped on Markus Eichenberger and Daniel Studer’s “Suspended”. Clarinet or bass clarinet, and double bass. There are schwack of extended techniques for each of the instruments. Slap tongue, key clicks, harmonic sorts of stuff, col legno battuté, open string and fingered snap pizzicato. All sorts of funky stuff that I can guarantee you didn’t hear on the hit parade in 1973… or the TV commercial: we were all supposed to buy each other a soft drink.

The bass clarinet and double bass are in the room. Like, really… in the room. See, I love stuff like that. It’s like dress rehearsal, the final run-through and the audience is in the lobby sloshing vino ten minutes before the ushers open the doors and start handing out concert programs.

Structured and non-structured contemporary improvisation is not for everyone. However, we’re talking about the way the BDA-3 is going to handle this, right? Yeah well, here’s a really big seller: the BDA-3 is percussive. Slap tongue on the bass clarinet has that really great “thoop”. The percussiveness is as important as the pitch itself. They each have their own place in the sound of the action upon the instrument. The BDA-3 will give you that. All of it. All forms of pizzicato on the double bass is astounding. It makes you want to hear all of that stuff, more. The BDA-3 does percussion and percussiveness really, really well. In fact, I would almost say its this DAC’s strongest point.

Streisand? Barry Manilow? No no no. This is not that. In terms of choice of music, you came to the wrong show, brother-man.

Electronic and electro acoustic music: that’s another matter. I have been discovering some really cool contemporary Scandinavian and German Jazz lately. I found Thomas Strønen on Tidal a while ago. “I liked it so much I bought the company.” Oh wait, no, I bought the CDs. The other thing was shaving. Hippies don’t shave, or get haircuts.

Thomas Strønen’s “Pohlitz” is not like his other music. He’s a jazzer, generally… but this Cd is really cool. It’s not Jazz. It is what you might expect from a lot of electronic and electro acoustic music. There are musical “events” that happen and each of them must be accounted for, together, and separately. The BDA-3 will handle this panoply effortlessly. It separates these events and puts them where they are supposed to be on the sound stage. Each of the events sound good, and most importantly, sound musical. Remember, that’s why we’re all here.

My advice to Bryston is: go with what you know. The thing the BDA-3 does REALLY WELL is percussiveness. Does it do do percussion well? Yup. Really well. But remember, technically, piano and guitar are percussion instruments. The string is struck or plucked, and then it decays. But it’s not that it happens… it’s HOW it happens. It’s fast. The BDA3 is really fast. You know that comic book character that zips around the room and everyone else is slow-mo? The BDA-3 is fast, like that.

Does BDA-3 sing? Yup. You want Streisand? Go ahead. Violins in a section? Sure, why not? It’s awesome. You’ll want to buy the world a Coke. You want to go to the Copa Cabana? Go ahead. The BDA-3 will make you want to get out your tie-dye shirt and a pair of bell-bee dungarees. Maybe some Gene Simmons heels, too.

Shane Fage.

PS: The BDA-3 does not suffer bad recordings well. It won't fix bad recording engineering. In fact, "If any of you cats dropped the brown acid, make your way to the hospital tent, ricky-tick. It's a bummer trip, man."