MQA Deep Dive - I published music on tidal to test MQA

[PierreV]

Hybrid-MQA? What is that? For what I know, the whole concept of MQA was for saving wideband, to be able to stream at lower bit rate, so why would somebody by a CD with MQA?

Because (dishonest) marketing works?

 

[Raindog123]

@GoldenOne To add to your list of manufacturers' voices. Auralic's position & statement on MQA (Nov 2020)

 

[SKBubba]

- It will change the proportion now that Spotify will add CD equivalent FLAC
- we can't use it as a conclusion that they "don't care", but more than they "seem to not care about paying more for having CD Quality"

Probably so. Tidal HiFi is available for the same price as Spotify via Best Buy in the US, though, which was a pretty good deal until they decided to go all MQA. Wonder how many Spotify users switched? Or how many even know about it?

ETA: Wonder How many Tidal users would pay more for an MQA tier v CD quality hi-fi tier like they had before?

 

[levimax]

I switched from Tidal to Qobuz when Tidal switched over all the Warner Music to MQA and dropped the previous versions. Reading this thread I am now wondering what Qobuz is actually streaming. Has anyone "tested" the Warner Music 44.1 streaming titles on Qobuz to see if they are really 44.1/16 or if they are actually 44.1/13 MQA files? I would do it myself but not sure how.

 

[Grooved]

Probably so. Tidal HiFi is available for the same price as Spotify via Best Buy in the US, though, which was a pretty good deal until they decided to go all MQA. Wonder how many Spotify users switched? Or how many even know about it?

ETA: Wonder How many Tidal users would pay more for an MQA tier v CD quality hi-fi tier like they had before?

Some media are also responsible of people having moved to Tidal when they said recently something like "Warner is now offering it's catalog in Hi-Res on Tidal", yes they used Hi-Res, while they should have said MQA and explain briefly that it can be 16/44.1 and so not Hi-Res (even without talking about the processing and if it's lossy or not,...).
- they don't know what they talk about and should stop
or
- they know, but are ready to lie (to simplify things, or to get benefits)

From people around me, and I can't generalize based on that, Spotify users really like the different features and are more enjoying being able to listen the music they like (and new ones) the simplest way, even if they play some music/work in music. But from the ones playing for fun/hobby or working on, a lot have switched to other than Spotify, even for ones that have more time to listen in the car than any other place. I tried it and found that even in the car, in the middle of a traffic jam, you can hear that Spotify is lower quality than Tidal or Qobuz.
And it's a bit part of the problem, because going from Spotify to Tidal, you get a better experience, even if MQA can be a fraud. Something you may not think if Spotify was offering CD quality ;-)

 

[Grooved]

I switched from Tidal to Qobuz when Tidal switched over all the Warner Music to MQA and dropped the previous versions. Reading this thread I am now wondering what Qobuz is actually streaming. Has anyone "tested" the Warner Music 44.1 streaming titles on Qobuz to see if they are really 44.1/16 or if they are actually 44.1/13 MQA files? I would do it myself but not sure how.

Just what I was talking about the previous post ;-)
Tested on "Blood Sugar Sex Magik" just now and it's FLAC 16/44.1, created from the same master (2014 Remastered which is louder and was made hitter the limiter at 0dB a lot of time... but may be more balanced, to test more to be sure) than the FLAC 24/96 they have.
Qobuz have these two, from the same 2014 master.
Tidal had FLAC from original master and 2014 remastered, then added the MQA of both, then deleted the FLAC version, and then deleted the version from original master, leaving only the MQA from 2014 master.

You can checked the some measurements differences on one track that I got here (Qobuz 24/96 and Tidal MQA 24/96 unprocessed so 24/48, will do a processed one) :

 

[SKBubba]

I switched from Tidal to Qobuz when Tidal switched over all the Warner Music to MQA and dropped the previous versions. Reading this thread I am now wondering what Qobuz is actually streaming. Has anyone "tested" the Warner Music 44.1 streaming titles on Qobuz to see if they are really 44.1/16 or if they are actually 44.1/13 MQA files? I would do it myself but not sure how.

I checked Fleetwood Mac - Rumors. It is only available in MQA "24/96" on Tidal. Qobuz has FLAC 16/44.1 and 24/96.

Roon shows the Qobuz 24/96 streaming/playing as FLAC 24/96 and the 16/44.1 as FLAC 16/44.1, no MQA detection on either.

 

[orchardaudio]

This may be a repost but here is a very good video on the topic:

My stance has been that Orchard Audio will not support MQA decoding in our products.

 

[Music1969]

LOL what a legend, he snuck in content that can be used to test MQA

This was hilarious. And very clever.

 

[GoldenOne]

This may be a repost but here is a very good video on the topic:

My stance has been that Orchard Audio will not support MQA decoding in our products.

Thank you for your firm stance against MQA.
Its good to see more and more manufacturers not giving in to it.

 

[nimar]

Thank you for your firm stance against MQA.
Its good to see more and more manufacturers not giving in to it.

Anyone want to buy a Topping D90 MQA? great DAC (true) and MQA is amazing ;-)

To be fair I've been trying to sell this for a few weeks already for non MQA reasons.

 

[orchardaudio]

I also saw some references to this earlier in the thread (maybe it was on the stereo.net.au forum) as MQA does not allow a full decoder with a digital output. However, any manufacturer that has MQA decoding has access to that data in the digital domain, be it I2S or some other protocol. Because this data has to be sent between the DAC chip and decoder digitally.

 

[mkawa]

According to one report, streaming subscriptions as of Q1 2000 were:

Spotify: 128 million
Apple Music: 72 million
Amazon Music: 56 million
Tencent Music: 44 million
Google Play / YouTube Music: 24 million
Deezer: 8 million
Pandora: 4 million
Others: 64 million

Among “others,” Tidal claims 3 million subscribers, or about 0.7% of all streaming subscribers.

So 99.3% of music streaming subscribers do not have access or listen to MQA. In fact, 97% don’t even seem to care about CD quality lossless streaming.

So basically MQA is a rounding error in the marketplace. It will not be the differentiator that drives Tidal to overcome Spotify.

yah, tidal was honestly pretty screwed and waiting to run out of venture capital before the big infusion recently. i don't see them getting anywhere near even youtube music in the next 5 years though. if lossless is the only thing they campaign on, they are doubly screwed, since those of us who care about that are exactly as you said, a rounding error in the overall marketplace.

i do think it's still good to expose this closed alternative to lossless as, frankly, the fraud that it is, but don't get too much underwear in too much of a twist because spotify lossless is pretty much going to be the nail in the coffin that takes all of our business from tidal and amazon music HD, and spotify absolutely does not want to pay MQA a red cent.

 

[Grooved]

I also saw some references to this earlier in the thread (maybe it was on the stereo.net.au forum) as MQA does not allow a full decoder with a digital output. However, any manufacturer that has MQA decoding has access to that data in the digital domain, be it I2S or some other protocol. Because this data has to be sent between the DAC chip and decoder digitally.

Whoa! Is this 100% asked if you build a MQA DAC ? That's a thing I wanted to be sure.
Actually, user can circumvent it at least for first unfold, I already did it staying in the digital domain, but if they ask for that, it would really said they don't want you to access to the digital stream after the other unfolds.

 

[Raindog123]

any manufacturer that has MQA decoding has access to that data in the digital domain

I think that's the question... How many manufacturers implement the second unfold in a standalone chip (FPGA, DSP or such) separate from a DAC chip? Sounds like the majority of the official licensed MQA implementations are based on integrated decoder+MQA [reprogrammable filter] chips - eg, ESS. But as I personally have no visibility, I am not sure.

While @mansr has mentioned that he reverse-engineered the processing in its entirety, is there an IP core from MQA, Ltd that does it "officially"?

 

[Jimbob54]

I switched from Tidal to Qobuz when Tidal switched over all the Warner Music to MQA and dropped the previous versions. Reading this thread I am now wondering what Qobuz is actually streaming. Has anyone "tested" the Warner Music 44.1 streaming titles on Qobuz to see if they are really 44.1/16 or if they are actually 44.1/13 MQA files? I would do it myself but not sure how.

Roon would tell me if they were mqa on qobuz. I've seen one example (a 2l title I believe), nothing mainstream.

 

[mansr]

I think that's the question... How many manufacturers implement the second unfold in a standalone" chip (FPGA, DSP or such) separate from a DAC chip? Sounds like the majority of the official licensed MQA implementation are based on integrated decoder+MQA [reprogrammable filter] chips - eg, ESS. But as I personally have no visibility, I am not sure.

While @mansr has mentioned that he reverse-engineered the processing in its entirety, is there an IP core from MQA, Ltd that does it "officially"?

MQA provide a software library for ARM and XMOS based devices, possibly others too. Some implementations do the resampling (if you can call it that) in software, others other program the coefficients into a DAC chip. The end result is the same with either approach.

 

[nimar]

I think that's the question... How many manufacturers implement the second unfold in a standalone" chip (FPGA, DSP or such) separate from a DAC chip? Sounds like the majority of the official licensed MQA implementation are based on integrated decoder+MQA [reprogrammable filter] chips - eg, ESS. But as I personally have no visibility, I am not sure.

While @mansr has mentioned that he reverse-engineered the processing in its entirety, is there an IP core from MQA, Ltd that does it "officially"?

You can assume all the none ESS (with built in MQA) designs are doing it this way, eg. All the AKM DAC's with MQA, and ESS without build in MQA. I could be mistaken but based on the design it looks like this is being done in the XMOS XU216 which is also why they only do MQA via USB, unlike the ESS design which allows it over any digital input.

Edit. Mansr beat me to the punchline.

 

[Atanasi]

I wonder what MQA's publishing agreement is like. Is there an anti-reverse-engineering clause to forbid test signals?

 

[Sgt. Ear Ache]

- It will change the proportion now that Spotify will add CD equivalent FLAC
- we can't use it as a conclusion that they "don't care", but more than they "seem to not care about paying more for having CD Quality"


that's also a wrong conclusion, because like explained in a previous post, most of ABX tests are biased as long as you make them with people not knowing what to listen in the music to hear the differences, and/or don't have the system to do it, and by saying that, I don't say you need a millionaire system, but at least a well positioned system.
.

uh, right.