 |
| cf. As We See It - More on MQA :-) |
MQA: A Review of controversies, concerns, and cautions
As the title indicates, this is a review of much of what I have presented here over the years as well as accumulated knowledge by folks who have taken time to dissect and divine some of the inner workings of this "format". I salute the audiophiles who have worked to figure these things out (many of them references in the footnotes in the article) - those who have had the courage to "call a spade a spade", those with the perseverance to ask tough questions, and pursue meaningful answers.
I would also like to thank Chris Connaker for reaching out and having the courage to publish the article. I don't know what the "politics" are like behind the scenes in the audiophile world, but I cannot imagine there not be pressures to conform to the "official" sanctioned message especially dealing with a product originating from a veteran audiophile company, endorsements of the major print magazines, and with the apparent backing of major music labels in the Industry.
Hope you all have a great week ahead!
Enjoy the music...
Addendum: October 2024 - A Dive Into... audio podcast summary of MQA with more recent updates.
Hi Arch,
ReplyDeleteMy mother used to say, if you want to get rid of a smell you use soap, not deodorant, i.e. tackle the cause not the symptom. MQA, deodorant for music.
Send my regards to mother.
DeleteShe didn't raise no fool! :-)
Great reading, my main problem with MQA is the DRM it might bring to the table. Almost like a trojan horse sneaking inside the industry and our gear.
ReplyDeleteFor example Sony has gear with MQA, quite unusual because Sony was always very stubborn to adopt 3rd party codecs. Now don't get me wrong I'm a bit of a fanboy for Sony but its the same Sony who made CDs that installed a Rootkit, for DRM... -_- So if Sony is pushing for MQA I'm sure it's not only about the sound quality if you get my point.
I understand your need for anonymity, in a world of legal conflicts it's better to be anonymous.
Anyway I'm always happy to read you blog Archimago, now don't do a NwAVguy on us! :-D Your readers actually worry about you believe or not. Best regards!
Rafael Lino (JourPT)
Sony is not the innovative company we all admired and held in hi regard decades ago. Have a look at their website extolling the virtues of hi res, employing the same misleading graphs and myths, eg hi res is like analog (as if performance like analog recording/playback equipment is a a step forward from CD). And this sort of shrill rubbish from one of the inventors of the CD just demonstrates the empty shell of what was once a technical leader.
Deletehttp://www.sony-asia.com/microsite/hiresaudio/
Thanks for note Rafael and Prep.
DeleteYeah, Sony is not what they used to be. A real shame given their legacy of so many innovations - especially CD / 16/44, some fantastic ES hardware also. These last few years, it just seems like they're "phoning it in" with products and jumping on bandwagons like the misleading advertising material.
I believe there is a time and place for anonymity. Obviously, if there is a need for maintaining a personal relationship, I would have no issues with being completely open. But I do believe that as a tech blog, and as one who has criticized the audiophile industry as being way too driven by personalities, it would be too much if who I am actually mattered! What I do and what I can express as truth matters much more.
I hope NwAVguy is doing alright.
Hi Archimago. Your article is great and if the readers are smart, they'll stop using MQA at all, pushing for (if anything more than CD) 24/48 FLAC instead :)
ReplyDeleteThanks Honza.
DeleteWe'll see what the consumer ultimately does with this. I hope that at least the audiophiles are well educated about what MQA is about and can do what we can...
Great work! The next weeks will be more than interesting. All the MQA pro fighters and the HiFi magazines can no longer ignore the facts, and will have to take a clear side now. :-)
ReplyDeleteAmazing how divisive MQA is, eh, Techland?
DeleteI think you're right. In the days ahead, what will the supporters of MQA do? I hope at least it will allow those who have been unflinchingly positive about MQA to reflect on how they evaluate what is to be recommended and how "quality" is to be defined and evaluated.
Hi Arch
ReplyDeleteFirstly, congratulations on producing well argued and evidenced based article. Also, kudos to Chris and Computer Audiophile for publishing it.
It is unfortunate that the industry is getting behind MQA, or trying to. We all know that the glory days of the industry have long since passed but to push all this pseudoscience purely around ulterior motives for the possibility of DRM is in my view unconscionable.
And you are right, it is like a cancer spreading through the industry and hanger ons (like the subjective press out to make advertising revenue). Just like how your long running thread on the Steve Hoffman site was deleted, even though the contributors were by large, robust discussions rather than nastiness. I guess the cancer is affecting that site's host too as he seeks to become relevant again to an industry that has shunned him for several decades.
One positive though, there is a lot of scepticism out there about MQA and its claims. You would expect that with those that have a bit of technical understanding of audio, particularly digital audio, but it is seen even among some of the more usually gullible sections of the audiophile community.
Just an anecdote, a couple years ago Bluesound (which make excellent music streamers) introduced MQA as part of a software update. Unfortunately the update meant that the bass and treble sliders were no longer available for any music. The complaints they received were loud and clear, their customers did not want to trade a functionality which hardly anyone uses anyway, for the "benefit" of MQA. After much justification by Bluesound they eventually gave in, though they did introduced MQA again but managed to keep the other functions.
Anyway, one of the reasons I checked your blog at this point in time is that someone has posted the article in the Steve Hoffman site, see link below. Let's see how long it takes before it is deleted...
http://forums.stevehoffman.tv/threads/mqa-controversy.734334/
Audio industry would do great step forward in distributing 24/48 FLAC, pure PCM from studio masters. That would help much more than fiddling with MQA, HDCD and other "enhancements".
DeleteThat post on the Hoffman site has already been deleted. They hate free speech there unless it is woo.
DeleteThanks Prep for the note about the Bluesound.
DeleteI wasn't aware of the loss of bass/treble tone controls with the MQA addition initially. Not surprising since MQA would not be keen on anything that can be viewed to change whatever "studio sound" they're aiming for... So are you able to use the bass/treble sliders when playing MQA, or is it deactivated and only used with non-MQA playback?
Alas, missed the Steve Hoffman post :-). Boy that was quickly taken down!
I honestly wonder what the fear is over on that site whether it's Hoffman himself or the "gorts". It's not like they're designing MQA hardware and as far as I can tell, Hoffman isn't remastering music in MQA, right? Unless he's friends with Bob Stuart we might not know about...
In fact, embracing discussion of MQA at the Hoffman forum might actually be a good thing for his image and that of the site. As one who's "looking out for the people" and independent of the interests of the larger Industry; a discussion forum for the "perfectionist specialist audio" folks who truly want the best quality audio available.
Oh well...
Hi Arch
DeleteThat is a good question regarding the Bluesound. I didn't know whether the bass and treble sliders work on MQA tracks as I don't use the tone controls. I gave it a try now and yes it works.
That is also a good point re Hoffman and being independent. I actually met him once, seems like a nice guy, and from his posting he doesn't come across a someone who would be that interested in the MQA format. Perhaps you are right that it is more the gorts. I think I mentioned before an off-line exchange I had with one of their more senior gorts which was a bizarre conversation to say the least.
Tone controls with MQA on the Bluesound work by applying a biquad filter (with parameters derived from the bass/treble settings) to the output of the core decoder, then restoring the LSB with the renderer instructions.
DeleteInteresting, and I think this needs some detailed explanation. My technical understanding of this is limited, so with regard to MQA and all its claims: Is this a legal workaround to enable EQ in the digital domain with MQA active, or is it more like cheating, with MQA only working partially anymore?
DeleteThanks for the details Mans. Good to see Bluesound gave the consumer what they wanted and is able to use the Core decoded data.
DeletePrep: Sounds like an "interesting" discussion with the senior "gort" :-).
The LSB copying is done by the MQA library. I doubt they disclose how it actually works.
DeleteThanks again Archimago for a great article exposing the ongoing problems and dangers with MQA.
ReplyDeleteI’ve been wondering why in your blind test I nevertheless preferred all three MQA samples because I heard more details in them than in the straight hi-res files.
In your February 3rd post about « Blurring » you pointed out that MQA’s filter introduces frequency dependent phase shifts and delays (up to 50 µs by 20kHz). Maybe then I was responding to added rather than revealed « deblurring » in the MQA files.
What I mean is that if a voice for example and a piano operate mostly in a different region of the spectrum, it is possible that frequency-dependent timing errors could be heard as more « separation » of the two, and so as somewhat clearer or more detailed. This could explain why MQA advocates hear a varying improvement in their listening test since this is entirely dependent on the music’s content.
One way to illustrate this is for example listening to a very thight Renaissance choir where the goal is sometimes one voice singing a chord (figuratively speaking) as compared to an opera choir where each voice has more character and is sometimes heard as a collection of soloists. Opera choirs are sometimes very loose in their simultaneity, a reason also being that they have to act and move.
I do have a sensitivity to short timing differences, so maybe the MQA files in the blind test sounded better to me because I like more details, whether they are real or not, a kind of euphonic preference I guess…
Keep up the good work!
Thanks Gilles for the note and interesting thought.
DeleteAn interesting idea about whether group delays of this nature might be having this effect. Remember that for those test files, I subjected BOTH the MQA decode and the 24/88 or 24/96 tracks to the same type of oversampling with the MQA-type filter!
Your preference therefore would not have been the filtering used, but there is still potential that the MQA processing did change time-domain characteristics on the encoder side with the baseband material or with the first decode when they expanded the "sub-band" data into the 22.05-44.1kHz or 24-48kHz octave if we accept that the ultrasonic frequencies had an effect on what you heard.
A more pedestrian explanation could be that you had a preference to the bit-depth setting the encoder used and dithering technique when encoding the MQA.
One way you could test is if you take standard 44.1kHz resolution audio and compare if you preferred linear phase or minimum phase upsampling. Do you hear more "separation" by doing that? You might end up preferring the minimum phase setting with the higher frequency phase shifts if this is the case...
Oh, sorry, I had missed the point that the filter was used both times! My mistake, and so my explanation does not work. Maybe as you say there were some time-domain characteristic already encoded. I am not equipped to play with filters because I listen mostly to CDs and SACDs or normal streaming from the internet, so I can't test that hypothesis myself. I looked back at the upsampling filter test you posted in April 2015 (in which I did not participate), but the data is not available any more. Maybe you could make it available again temporarily...
DeleteOr it could be because 3/3 'correct' in a three-sample binary choice test statistically isn't *that* hard to achieve by chance. (p=0.125)
DeleteThanks for this valiant critical labor of love — I've already high-fived and commented (as HalSF) at CA.
ReplyDeleteI'm really disappointed that John Atkinson has steered clear of this article by disingenuously declaring it's all a been-there/done-that rehash and by implying there's something hinky and disqualifying about your anonymity.
I'm still holding out hope that in the long run Stereophile is going to have to honestly grapple with the seemingly fatal shortcomings and downsides of MQA.
I'll dissent however from the criticisms of the Steve Hoffman site for declining (based on past train wreck threads) to host yet another MQA eye-gouging shoutfest flame war discussion, which is what we all know it would be despite the relentless civility of your own contributions. I value the relatively troll-free environment of the SHF site and value it more than yet another MQA thread.
Hi Hal,
DeleteYeah, I think it's fine that ultimately folks steer clear of the MQA debacle. Thanks for your thoughtful contributions as well.
It's good to see that the newest issue of Stereophile is starting to show some questioning around the "benefit" of MQA for audiophiles as per the "As We See It" editorial by Jon Iverson.
While it's a start, it is still disappointing that so many in the "professional" press had to take this long...
Great article. MQA is yet another in a line of audio shucks in search of a reason to exist *besides* DRM.
ReplyDeleteThanks Steven.
DeleteI believe Christoph Engemann has an upcoming article for Computer Audiophile coming in the days ahead that will go deeper into DRM and the ways something like MQA represents that concept.
I would like to thank all the record companies and their mastering engineers for donating all their time and money to enhance all of our music in this wonderful new MQA format, without any thought for gain or profit, on their part. Yeah, right. Not on this planet where greed reigns supreme! DRM will be turned on, sooner, rather than later.
ReplyDeleteHey David,
DeleteYou speak much truth... Let's hope it's not what happens of course!
This is also the more reason for why we fight a "good" fight IMO as consumers, for consumers.
Wow ! The best article on MQA ever published, congradulations ! Those published in Stereophile are so "lossy" compared to yours. Thanks for sharing your expertise and your so well documentend thoughts.
ReplyDeleteAnd yes, it is scaring to see that MQA could take the monopoly of the audio digital format: no matter which DAC you wish to buy, all the data coming in would have to go through in the MAQ cruncher. No, we, consumers never asked for that.
Amazing article.... Awesome.
ReplyDeleteRegards from Brazil
Thanks Denis and Marcelo,
DeleteI think everyone must have heard about the recent TEN publications including Stereophile being acquired by the same guys that run Hi-Fi News & Record Review. It'll be interesting to see how this works out and whether there are changes to the way things are reported when it comes to MQA.
I think HFNRR and Paul Miller have been historically skeptical of MQA. I'm sure there will be significant editorial freedom but it would be interesting to see disparate opinions across the pond with publications owned under the same entity!