Thanks to a lovely-sounding assemblage of components—Vandersteen's Model Seven loudspeakers, Boulder 1060 stereo amplifier, and AudioQuest cabling—we were able to take a first listen to Mytek's Brooklyn ($1995), an idiosyncratically styled, compact USB2 DAC/preamplifier/headphone all-in-one that will ship in early December. Touted as "the world's first stand-alone, high-performance DAC with MQA®technology," the 4 lb, MQA-equipped Brooklyn converts up to 32/384k PCM (including DXD), and native DSD up to DSD256 (DSDx4). It also claims a 130dB dynamic range, which should be sufficient for sending your chickens and more running for cover.
At the conference's start, Stuart (above) laid down the bottom line: "You can take an MQA-encoded file, play it anywhere, and it will sound better than CD."
"Our hearing is much more sensitive to timing information than we thought," he continued. "Because of our ingrained sense of survival, human beings are at least five times more sensitive to timing information than to frequency. We can discern time intervals of a few microseconds.
"What works perfectly for instruments doesn't work for digital filters, which smear timing. CD smears timing for 4 milliseconds, which is the equivalent of sound traveling 4 feet. With MQA, we've redesigned the way digital sampling is done. MQA reduces ringing and smearing dramatically, over ten times more than on 24/192 recordings."
What Stuart did not address at the conference, however, was the unexpected delay in MQA implementation. After the announcement at May's Munich High End Show that MQA, Inc. had over 100 strategic partners, everything has been quiet. Rumor had it that MQA decoding was requiring more battery power than was practical for mobile phones to supply.
Asked to comment on the situation, Stuart replied by email:
MQA saw it necessary for a small delay to optimize the delivery platform in order to provide more flexible streaming. MQA has also been preparing decoders for the mobile platform, but the most important aspect to note is that the audio is still amazing and unchanged. MQA has been busy building the supply chain. By working to increase the efficiency of the delivery, with no compromise to the audio, we make MQA a better company to do business with.
We have many customers in the pipeline in addition to the following manufacturers that have recently announced including AudioQuest, Pioneer, Onkyo, Bluesound and Mytek. And you should expect more by [the 2016] CES.
After Jurewicz (above) gave us the lowdown on the Brooklyn, the two men played snippets of several MQA-encoded selections. The first was a beautifully recorded Frank Sinatra track from 1957 whose lyrics were so soppy that they seemed destined for an entire collection of Hallmark cards. (I am not exaggerating!) Then came an equally sentimental, easy listening, DXD piano recording from Morten Lindberg's 2L label.
The icing on the first layer of the cake was part of a brand new duet by Judy Collins and Willie Nelson, entitled "When I Go." Along with it came a glowing endorsement from Alan Silverman, who has been recording Collins for 40 years. Silverman unequivocally states that encoding and decoding his master file with MQA produced a recording that was clearer than the original master.
While Silverman's endorsement is of major importance, it was difficult to listen to the part of the track that was played without conjecturing that the aging voices of both Collins and Nelson had been heavily processed. How anyone was supposed to know that this track sounded "better" than the master was beyond me.
This frustration was compounded by the fact that no one offered before/after comparisons of any tracks. I heard those comparisons at Audio High in Mountain View, CA, before John Atkinson and I filed our first reports about MQA, and was immediately convinced of MQA's efficacy. What I don't understand is why such comparisons were not conducted either here or at previous public demonstrations of MQA that have taken place at a number of venues, including both the 2015 CES and Audio High.
Be that as it may, the next exciting piece of news was that Mytek will take the Brooklyn's as yet unnamed sister A/D converter to the AES convention in New York City, October 29–November 1. The unit, which includes an MQA kernel, will be given to unnamed recording engineers to play with. What will happen after that could very well change recording history. I am certain that John Atkinson and other Stereophile East Coasters will be on hand to report everything that takes place.
At conference's end, Bråtelund served dessert. He took a huge risk by daring to stream, over the Marriott's terrible, soon-to-be redone WiFi network, an MQA-encoded hi-rez file (24/352.8k) of a recent 2L recording, Kim André Arnesen's soppy Magnificat. Actually, I don't know what was riskier, daring to trust a network that kept crashing time after time in room after room, or inundating press with so much sop that we risked drowning in ooze. Thankfully, the selection was short, and the experiment a sonic success, due to the major reduction in streaming bandwidth made possible by MQA. (The MQA process packed the 24/352.8k data into a 24-bit lossless FLAC file sampled at 44.1kHz.) I left the room excited at the prospect that MQA, Mytek, and Tidal are poised to take music playback to a whole new level.
COMMENTS
The most I've spent on a DAC or amp is $600 or so, since I mostly use portable gear. But I would pledge $2000 or more for an MQA DAC, if downloads of enough music are made available at an affordable price. Downloads of old "classic rock" albums with 1 or 2 tracks that I like, priced at $25-$30, would not constitute affordable music. So it's going to be jazz, classical, easy listening .... (sigh)
Bob Stuart -- "Our hearing is much more sensitive to timing information than we thought"
George Carlin -- "This guy is pretty smart"
--------------------------------------------
Bob Stuart -- "What works perfectly for instruments doesn't work for digital filters, which smear timing. CD smears timing for 4 milliseconds, which is the equivalent of sound traveling 4 feet. With MQA, we've redesigned the way digital sampling is done. MQA reduces ringing and smearing dramatically, over ten times more than on 24/192 recordings."
George Carlin -- "A aa He's full of ****"
How can he lie like that is mind boggling.
"...encoding and decoding his master file with MQA produced a recording that was clearer than the original master." So the encoded file does not sound identical to the original. And as we know, "clearer" could describe any number of effects that might be described as undesirable if we weren't dealing with Hi-Res files.
"clearer" could describe any number of effects that might be described as undesirable if we weren't dealing with Hi-Res files.
In the original MQA presentation, Bob Stuart claimed that the correction of the A/D and D/A converter's time dispersion resulted in an overall recording/playback chain impulse response, from the analog signal reaching the microphone to the decoded analog signal in the playback system, of just 50µs with a leading-edge uncertainty of an incredible 4µs, and zero pre- or post-ringing.
John Atkinson
Editor, Stereophile
1) So if I understand what you wrote, the reason it sounds better than the original can be attributed to correction of time dispersion errors that happened after the signal was generated by the microphone and by the time the information was recorded on the medium used for the master?
2) In order to encode existing masters that did not use MQA technology, would the engineers have to estimate what time dispersion errors were incurred in the recording chain up until the master was created?
3) To change the subject (which, after all, is "Changes"), what's the chance that Apple will stream MQA encoded music, even if they didn't increase the bit rate? What would be the equivalent uncompressed bit rate? (E.g., Meridian claims material encoded with MQA technology at redbook bit rate is at least as good as 24/192K uncompressed streaming.)
In order to encode existing masters that did not use MQA technology, would the engineers have to estimate what time dispersion errors were incurred in the recording chain up until the master was created?
The primary source of timing dispersion is the A/D converter. I am told that mastering houses keep good records on what was used - the population of professional A/D converters is small - so MQA has been modeling those converter errors and applying the appropriate correction. See my article on MQA: www.stereophile.com/content/ive-heard-future-streaming-meridians-mqa.
To change the subject (which, after all, is "Changes"), what's the chance that Apple will stream MQA encoded music, even if they didn't increase the bit rate?
MQA mandates use of a lossless-encoded 24-bit container, and as far as I am aware, even though Apple now asks for 24-bit masters for iTunes, they have no plans to introduce lossless files. But that might change, of course, as they try to differentiate their streaming service from Spotify, etc.
John Atkinson
Editor, Stereophile
The timing dispersion of any pro ADC and DAC on the market is 0 picosec "there is no such thing" of time dispersion, there is filter time latency but that thing only matters in recording and monitoring of overdubs and most properly setup DAW have latency compensation (it shifts the new recorded track the necessary samples back to properly align the timing with the other tracks) so for classical and jazz music it is not relevant.
The timing dispersion of any pro ADC and DAC on the market is 0 picosec "there is no such thing" of time dispersion, there is filter time latency but that thing only matters in recording and monitoring of overdubs...
You are forgetting the fact that when an analog signal is converted to PCM digital, there needs to be a high-order low-pass filter, called an antialiasing filter, to remove all content above half the sampling frequency. Otherwise that content folds back into the baseband producing enharmonic distortion. This filter introduces smearing in the time domain. The aim of this aspect of MQA is to eliminate that smearing.
This nothing do with the conversion latency you mention, which is how long the conversion takes, which is indeed an issue that needs to be addressed when performing overdubs or using plugins in a DAW.
John Atkinson
Editor, Stereophile
What is "smearing in the time domain" explain that to me?
Anti image and anti aliasing filters are phase linear there is no time frequency selective shifting only added latency and both channels are clocked from the same source so no time discrepancy between the channels.
I did design AD front ends for a PA console i never did find any time based problems with digital filters, Please explain to me what "smearing in the time domain" means in mathematical sense sincerely i really wanna know.
What is "smearing in the time domain" explain that to me?
See Bob Stuart's article from the October 2015 issue of the Journal of the Audio Engineering Society at www.aes.org/tmpFiles/elib/20151107/18046.pdf. (It's a free download; you don't have to be an AES member to read it.)
John Atkinson
Editor, Stereophile
As a letter from Alan Silverman, Judy Collins' longtime recording engineer, will explain in the January issue, the snippet of her duet with Willie Nelson that I heard was virtually unprocessed. The voices were not filtered, nor was dynamic range compressed. I think it is fair to say that Silverman remains in awe of MQA, which he calls "one of the most important developments in the history of digital audio because it virtually eliminates artifacts that color and cloud even the highest resolution digital recordings."
| Loudspeakers Amplification Digital Sources |  |
Analog Sources Accessories Featured | |
Music Columns Retired Columns | |
Show Reports | |
Features Latest News Community | |
Resources Subscriptions |
© 2024 Stereophile
