MQA: Meridian Delivers New Format
I hinted at the coming announcement from high-end equipment manufacturer in yesterday’s post. Today, the veil has been removed and we’re learning more about what MQA is. First, there’s the acronym. It stands for Master Quality Authenticated and is supposed to be a new method of capturing and representing digital audio in a much smaller package. The website states, “A revolutionary British technology is bringing a whole new meaning to the sound of music. MQA will change everything.” The new scheme is called “encapsulation” and is based on the way we all hear sound.
This is all starting to make sense now. A month ago, I attended Robert Stuart’s (and Peter Craven) paper session entitled, ” A Hierarchical Approach to Archiving and Distribution”. Little did I know that it would branded MQA and available as a commercial product complete with encoders and decoders. Leave to the brainiacs at Meridian to come up with a new way of storing and retrieving music.
The essence of this new approach is described in the abstract of the paper. Here’s what it says:
“When recording, the ideal is to capture a performance so that the highest possible sound quality can be recovered from the archive. While an archive has no hard limit on the quantity of data assignable to that information, in distribution the data deliverable depends on application-specific factors such as storage, bandwidth or legacy compatibility. Recent interest in high-resolution digital audio has been accompanied by a trend to higher and higher sampling rates and bit depths, yet the sound quality improvements show diminishing returns and so fail to reconcile human auditory capability with the information capacity of the channel. By bringing together advances in sampling theory with recent findings in human auditory science, our approach aims to deliver extremely high sound quality through a hierarchical distribution chain where sample rate and bit depth can vary at each link but where the overall system is managed from end-to-end, including the converters. Our aim is an improved time/frequency balance in a high-performance chain whose errors, from the perspective of the human listener, are equivalent to no more than those introduced by sound traveling a short distance through air.”
So what does that mean for the future of high-resolution audio? It means smaller audio files that contain all of the information of very large sound files produced through the use of very high sampling rates. As they say on one of my favorite TV programs, how do they do it?
I’ll go into the details in tomorrow’s post but as a teaser, think about all of those empty data bytes that you’ve been downloading when you purchase a high-resolution audio file. As sample rates increase to 192, 384, 768 and even 1536 using 24-bit or 32-bit words, audio files are getting huge! What if you could scale the encoding of a piece of music according to how much information is present at each frequency range and at each stage in the production/distribution chain? Obviously, there is not a lot of sound in the ultrasonic region…even if you agree with me and accept that partials at 40 kHz make sense. So why do we need to through 24-bits of information to storage amplitudes that are down 70-90 dB? What Robert and Peter have come up with is a way to maximize the use of digital data where it makes the most difference…in the audible range and to tailor the amount of data at each stage in the distribution process.
But this new approach to capture and distribution doesn’t address the fact that the original source recordings…the input to the new “encapsulation” process…still don’t have sufficient fidelity to tax 96 kHz/24-bit PCM. The burden will still be on the producers of the music to get their act together and release records worthy of the MQA technology.
The promise of MQA lies in the uniqueness of the approach. This is data compression without loss. Remember these very same individuals brought us MLP (Meridian Lossless Packing) 16 years ago. This isn’t really unexpected.
To be continued.
+++++++++++++++++++
I’m still looking to raise the $3700 needed to fund a booth at the 2015 International CES. I’ve received some very generous contributions but still need to raise additional funds. Please consider contributing any amount. I write these posts everyday in the hopes that readers will benefit from my network, knowledge and experience. I hope you consider them worth a few dollars. You can get additional information at my post of December 2, 2014. Thanks.
What’s the difference to a FLAC file – based on 24/96 (or 192)??
I don’t have that information yet but I wouldn’t be surprised if they chop 30-50% off of the files.
From… these marketing talks, it seems MQA is a lossy format specified for high-res music sources.
I am not so sure how this will fly to people… With just 1Mbps stream there will be a lot of stuffs being thrown away.
Won…it’s much more than that. This will not be “lossy”…unless you think throwing away a whole bunch of packets with nothing but zeros in them. These are people that I trust to do the right thing. I’ll reserve judgement but it sounds very interesting.
I hope Merridian has thought carefully about how they will licence the decoding part of this IP. If the decoding technology is either free or open source and easy to implement (I like they are adopting FLAC as a container) and it lives up to the hype of high quality sound in an efficient compact container then it will become popular or even dominent. Hopefully they can make money on the encoding side at the mastering stage, etc? If they go proprietary from end to end then it will fail in the marketplace. For example WMA was free and has gotten reasonable support but WMA Pro required licensing and thus its use was stunted and it is essentially a non existent format. As a person who has over 6,000 albums of FLAC files on my music server, I think the FLAC container is the way to go.
From the very limited information that I’ve seen and the paper that I’ve read, they will be licensing technology to other hardware and software companies just like Dolby or DTS. Im fact, MLP became Dolby TrueHD through a licensing arrangement. FLAC is fine but imagine the same quality of sound at 96/24 in a package half the size. It raises the bar for streaming as well. This could redefine things. But the restriction is on the production of audio that meets the new standards…this isn’t going to happen.
Licensing IP is fine but it has to also deal with the realities of the Internet, Open Source, and a myriad of other issues. Users gravitate towards freely transportable media and away from closed systems. Think MP3 which is wildly successful compared to SACD which was a disaster (as you have pointed out Mark).
As a user and proponent of the Logitech Music Server (originally the Squeezebox Music Server) which in my opinion is still the best (beats the pants of DLNA for instance in speed and featuers) reasonable cost end to end music server and client technology (remember I showed you the Raspberry Pi Squeezebox client that streamed 192k/24b), MQA will not get implemented unless it is free to license. Apple lossless is included in LMS but WMA Pro was never included. LMS lives on precisely because it was open source from the beginning and was able to out live the corporate botch job when Logitech acquired and then abandoned the technology.
I remember a long ways back in the early days of the computer revolution when I worked for a large software company and I happened to attend a very early presentation of Adobe’s PDF document format. Adobe fully intended to charge for both the PDF document creation tools AND charge for the client readers. I told them that they had to give the client away for it to be success. They of course balked at my prediction. History supports me in that PDF eventually was opened up and the clients are free and widely available.
I just hope Meridian really thinks this through if they really want this to become a dominent technology.
How do you expect a company like Meridian to make money from their development if not through licensing?
I would expect them to make money in the ENCODING stage as I said before! The DECODING stage is where it will either fail or succeed based on how freely and unincumbered they make it. Apple Lossless is free at least on the decoding end I believe and has a reasonable usage share. Originally, WMA Pro orignally could only be licensed and has negligble share!
FLAC trumps them all and is simply a container just like MKV is a container. You put whatever codec you want inside these containers. For instance you could put an Ogg Vorbis compressed file inside a FLAC container which wouldn’t make it lossless anymore! From what I understand you’ve written MQA is a new codec that can fit inside a container such as FLAC. From what I read at the Meridian site, it would appear there is a forking process that will allow an existing FLAC playback device to play an MQA file at a lower rez level and that same file will play at a higher rez level on a MQA device? Hopefully all will be revealed in the coming months?
They are enhancing the sound when the appropriate technology is available (the MQA decoder) and leaving things in the FLAC or PCM world when it isn’t. MQA is much more than just a compression technology…but it isn’t a technology that’s going to restore older recordings to high-resolution standards.
i suspect there are already a number of MQA capable DACS on the market, upgradable via firmware. this has been in development for a long, long time.
They’ve been working on this for a long time and have made the rounds to the labels and creatives…I’m sure they have equipment for doing this but I would be surprised if it’s widespread already. You have to engage with the labels first.
It looks as if Robert Harley has had a chance to experience this new format (MQA) that Meridian has to offer he seems to be very impressed with it. http://www.theabsolutesound.com/articles/robert-harley-listens-to-meridian-mqa/ We will have to see if “those in the know” embrace it and unify behind it before some other format is introduced as being so-called better and the spiral downward begins again. From the looks of things this upcoming CES will truly be the High-Res event of 2015 I am really looking forward as to what’s to come time will tell.
Thanks for the link. I read Robert’s assessment and find his description of the sound of the Modern Jazz Quintet completely off the mark. It was recorded in the 50s so it’s going to be limited by the technology of analog tape. Just because it was captured through MQA encoding and the decoded from the smaller sound file, doesn’t (and shouldn’t) do anything to modify the fidelity of the original. MQA could become a great “codec” like MLP before it…but it doesn’t magically improve the recordings that we already have. If I were to encode my tracks using this technology (which I might do for streaming for example), I would expect the tracks to sound exactly as they sound right now. I don’t want a distribution technology to “polish” my tracks.
Let’s say Meridian succeeds in marketing a new kind of audio files: a “auto-dynamic data compression envelope”, the problem of big buckets for standard resolution music could disappear! This would then help discern true HRA and clarify provenance.
I’m with you…this is huge in the distribution of high-resolution audio…it does nothing for the sound quality.
well, maybe it is me, but I just don’t get it.
1) smaller files? So what? with storage being so cheap today (quick look shows 4TB drives for $150), who needs smaller files?
2) for on-the-go non-streaming listening you say? Heck, between internal storage and the SD card, my A&K 240 has 347GB and 4,602 songs on it. Granted, only a small percentage are hi-res containers, but I use uncompressed FLAC and I still have 123GB free!
3) what about for streaming on the go? Sure, sending smaller files will keep your data bill lower, but most people that use streaming seem to have been satisfied with Spotify and if people are using Apple earbuds, they don’t care about SQ anyway.
4) streaming at home? given the speeds we have today with broadband, if your player of choice has a big enuff buffer, file size shouldn’t matter.
Now the DSP that they seem to be doing to make music sound better appears interesting, but have you seen the price of a Meridian system lately? I don’t think they are going to get any converts just because of this technology.
And as for RH’s comments, when was the last time you saw anyone in the major magazines give a bad review, especially when given a sneak peek at something?
Chris…like it or not we all need smaller files. If Meridian has figured out a way to store the same or better fidelity in a bit bucket that is half the size of the original, they have a winner. It’s not just about storage, it is also about transmission.
The streaming component is a clear winner…not more lossy compression algorithms.
Based on what I read, MQA will need to be applied during recording / mastering. Only an MQA playback system will be able to decode the additional psycho acoustics data. This is like HDCD, where you need to license a technology to for your DAC. No with computer audio, do they (Meridian) expect a player to support this, the DAC or both? I wish them best of luck but if this will not be GPL or open in any other way, adoption will be very limited.
It’s an encode and decode process that shrinks the size of the resulting file AND allows the benefits of higher sampling rates to be included in the smaller files.
Having owned several Meridian components down the years, and attended past events presented by Bob Stuart, I have nothing but the greatest respect for this very English HiFi marque.
The news that an innovative format is forthcoming from Meridian therefore merits serious amounts of my attention.
Although we’ll have to hear what it sounds like in practice, they very fact that this latest concept gets us out of the ever rising bit rate numbers game should be as welcome as a pair of Wilson Alexandria XLF’s under a giant Christmas tree.
Experience tends to suggest that Meridian doesn’t do things by half and is a technologically led company, so my confidence is pretty high that they are going to deliver.
If we are talking about non-lossy compression, there are already viable options. Why another? Is it a return to a simpler encoding time…. like perhaps run length encoding? I’m assuming there are vast quantities ofconsequtive ‘0’ in the data stream……
It’s more than just another lossless compression.
The way I understand this is that, from a data point of view, the compression will be lossy. Wheather this will be lossy from an ‘audio data’ point of view remains to be seen (or heard). I also trust Meridian to deliver a quality protocol. However I can envisage endless discussions, ‘tests’ and reviews in the so called audiophile circles regarding the audibility or not of the new data-lossy compression algorithms. One has only to read through discussions and ‘articles’ pertaining to supposedly audible differences between mathematically lossless compression algorithms like FLAC and uncompressed WAV to imagine what will happen with regards to this new format which, after all, will be mathematically lossy. Time for the popcorn! Btw any lossless compression algorithm like FLAC or ALAC includes run-lengh encoding which gets rid of extraneous zeros so this functionality does not appear to be the jist of MQA encoding. Variable sampling and bit length sound to be the core idea and it does sound very promising.
However, reading this http://www.whathifi.com/news/meridian-audio-mqa-paves-way-high-res-streaming one does not get the impression that MQA is just a recoding format for transmission but it may be a method to originally record (sample and code) the signal also. Conflicting(?) info at this stage? The info / speculation about data file size similar to CD quality file size and about base (un-decoded) CD quality look interesting.
Check out today’s post.
I agree to a point. I wouldn’t invoke “lossy” with this technology. The output will be authenticated to match the original. What I’m really interested in is the quality of the original.
Here’s a very interesting thread from people that have heard MQA. There’s even a link to the actual patent application!
http://www.meridianunplugged.com/ubbthreads/ubbthreads.php?ubb=showflat&Number=225234#Post225234
Here’s a link to the full PDF of the patent filing. A little easier to read!
http://patentscope.wipo.int/search/docservicepdf_pct/id00000025657679.pdf?download
Thanks.
As I read throught various posts, it would appear that MQA is a much about “Timing Information” as it is about Frequency Response. It also makes me think about so called “Time Aligned Speakers”.
The MQA encoding process as I am coming to understand it reminds me of Mr. Johnson’s HDCD technology (now owned by Microsoft) that would encode 20 bit Metadata in the CD data stream that could be decoded via HDCD aware DACs. Apparently the MQA Metadata is encoded in normal PCM fils (FLAC as well) and is decoded by MQA aware DACs.
A youtube video at MQA launch.
What’s curious to me is that MQA has been shopped around for supposedly several years but this is the first general public notice. To some extent if this pans out to be a breakthrough, then products like Pono are toast or least have to re-jiggger their product – first gen hardware is a doostep?
So, did Neil Young and company know about MQA? I would say yes, because I just now remember seeing a picture on the Internet many months ago where Neil Young was meeting with Bob Stuart. So maybe there is another shoe to drop with Pono that couldn’t be revealed because of NDA and because MQA wasn’t ready to launch on the Pono time table?
Could the Pono player be constructed so that a firmware update in the future could turn on MQA?
Mark, I think there is a lot of digging and reporting to be done here and your just the man to do it!
I’m fascinated by this development and will reach out to Robert at Meridian and see what I can find out.
Here’s the youtube link:
https://www.youtube.com/watch?v=2xpeEn1Mjtk&app=desktop
This all seems like a much more advanced form of psychoacoustical stuff that resulted in MP3. Much more advanced yes but the same general concept. Sure it makes what would be large files easier to distribute and store but at the cost of requiring us to either upgrade or change out our DACS or AVRs with no real improvement in SQ with the high rez downloads we can get now. I am not surprised that Robert Harley was wowed with the sound! I wonder why.
Joe…I’m convinced that this is much more than just more psychoacoustic stuff…they’re getting into new sampling theory and neurological reactions to sound.
Neurological reactions to sound = Psychoacoustics, at least in my book.
New sampling THEORY? Not sampling techniques? Big difference, you know. Are there any Phd’s behind this? Any publications in scientific journals in the public domain?
You can read their patent application and the AES papers by Robert Stuart and Peter Craven…these are very smart people. Their MLP algorithm beat out all other CE companies back when DVD-Audio was looking for an encoding solution.
MLP was a very nice algorithm, as far as I know the first commercial implementation of audio compression based on linear prediction. FLAC and ALAC are in the same boat. Linearly predictive coding itself had been around longer than I care to rmember. I was in the University in the 80s and it was already a long established technique in computer speech processing. So what I’m asking is not the info about the new algorithm but the basis for suggesting this is NEW sampling THEORY. Anyhow, time will reveal all as always.
Nik, you obviously have more in depth knowledge of this stuff than I do. I appreciate the comments. I think what’s going on is the specific application of the knowledge to the problems of high-end audio.