MQA arrives on Tidal

[johnhunt]

Click on new albums then the Masters tab and show all.
You have to be a Tidal hifi subscriber to see them and use the desktop app.

There is an amount of stuff MQA on tidal that doesn't show on that particular page

 

[Rug Doc]

MQA is very clever imo. It doesn't appear to 'add' anything to the signal, but instead is more faithfull to the original analogue recording because doesn't remove the parts from the recording which are actually fundamentally important in replicating atmosphere and realism in playback, previously compressed out of the way. Instead what MQA does is faithfully contain the information in a 44.1 container - how? - the term 'folding' is used instead of compressing, the important frequency area of realism and atmosphere (which MQA relies on for its SQ improvements) has been folded away and hidden from non MQA dacs, It's a clever way of getting a 196 file into a 44.1 container. The only way of getting the original file back out (whether it be 24bit 192 or whatever) is with an MQA DAC although higher than CD res files are still available without an MQA DAC, with only three of the four folds being decoded... hence Tidals 'Master' moniker.. I have no idea why Tidal has not made Master and MQA available on apps, making it only available on the desktop system as the whole MQA file is very small and easily stored and streamed on mobile devices... I can only hope that it's not because they'll be able to charge more for it...

 

[zekezebra]

MQA is very clever imo. It doesn't appear to 'add' anything to the signal, but instead is more faithfull to the original analogue recording because doesn't remove the parts from the recording which are actually fundamentally important in replicating atmosphere and realism in playback, previously compressed out of the way. Instead what MQA does is faithfully contain the information in a 44.1 container - how? - the term 'folding' is used instead of compressing, the important frequency area of realism and atmosphere (which MQA relies on for its SQ improvements) has been folded away and hidden from non MQA dacs, It's a clever way of getting a 196 file into a 44.1 container. The only way of getting the original file back out (whether it be 24bit 192 or whatever) is with an MQA DAC although higher than CD res files are still available without an MQA DAC, with only three of the four folds being decoded... hence Tidals 'Master' moniker.. I have no idea why Tidal has not made Master and MQA available on apps, making it only available on the desktop system as the whole MQA file is very small and easily stored and streamed on mobile devices... I can only hope that it's not because they'll be able to charge more for it...

I would suspect / hope it's more to do with processing power. I do notice my processer working more when playing MQA files

 

[venton]

There is an amount of stuff MQA on tidal that doesn't show on that particular page

That's my questiom really, how do you know what is and isn't MQA from search results (outside of masters page).
Clearly in time trying to find the album by looking through every listing on the masters page will be impractical.

 

[johnhunt]

That's my questiom really, how do you know what is and isn't MQA from search results (outside of masters page).
Clearly in time trying to find the album by looking through every listing on the masters page will be impractical.

Who knows how Tidal will manage it. Sensibly I would imagine . There is a link to a list complied by users elsewhere in thread

 

[Stunsworth]

Click on new albums then the Masters tab and show all.
You have to be a Tidal hifi subscriber to see them and use the desktop app.

The problem is that doesn't show _all_ the master albums.

 

[adamdea]

You know - there is (and has been available for a while) at least one DAC "firmware" that's basically a busybox-based rootfs, Bluesound Node 2. Just go to their website and download it.

Their software is basically all in /root, in "encrypted" perl. I don't know if they're that stupid or just want to make it harder for script kiddies, but in case you're wondering - it's easy to crack. Even the official documentation warns about this - https://metacpan.org/pod/Filter::Crypto#WARNING .

The interesting part about it isn't their custom software though, it's the MQA decoder. They implemented it as an ALSA rate converter which you can plug into the "rate" plugin. See ie. /etc/asound/asound.conf.simple. The "ssc" converter is provided by /usr/lib/alsa-lib/libasound_module_rate_ssc.so which is an ARM EABI5 shared library, so you either need to compile and run your alsalib-using program on ARM or use ie. qemu-user to run it on x86 or elsewhere. It should be then trivial to put ie. pcm.file as slave to get the output in raw PCM samples (or with WAV metadata).

Probably not of much practical use as there's no encoder, but it still could be interesting to see what it does when passed samples with mqa-encoded metadata (valid song input).

I think that's probably what the chap on CA has done on the thread I linked earlier (can't remember whether it was on this thread or the other one)

 

[Timcat]

That's my questiom really, how do you know what is and isn't MQA from search results (outside of masters page).
Clearly in time trying to find the album by looking through every listing on the masters page will be impractical.

If you are using an MQA DAC, it should tell you when it is receiving and decoding an MQA file. My Mytek Brooklyn has a blue "light" which comes on when it is receiving and decoding an MQA file.

 

[Jim Audiomisc]

I have a few CDs with the visible waveform crushing of HDCD, so I have thought of making a HDCD decoder that does the expansion without abothering to detect the flag. These are Malaysian pressings and I suspects someone touched the gain, obliterating the flag sequences

If you want to do that, the source code in the test HDCD programs I wrote is available from
http://www.audiomisc.co.uk/software/index.html

The RISC OS version does 'blind' peak correction assuming HDCD encoding. The Linux (ROX) version finds the codes and uses them to control decoding. Both versions have 'C' source code. Beyond that, I'm a lousy programmer, so apologies in advance for the awful style and lack of comments.

 

[Jim Audiomisc]

Just under 200 MQA albums have been added since the weekend. Since they're unlikely to be completely new masters I'm guessing that someone - Tidal or MQA - have pushed existing masters through the MQA encoder.


https://docs.google.com/spreadsheet...-Oo2MjIa3orv9DKZfwiRQKmTAA/edit#gid=945476039

That raised the curious question of: How on Earth would they have "Authenticated" about 200 albums in a matter of a few days? Indeed, how have they even traced exactly what ADCs and pre-processors were used so quickly? It seems more like, get out each tape and run it though the MQA decoder with some standard presettings.

 

[Jim Audiomisc]

You know - there is (and has been available for a while) at least one DAC "firmware" that's basically a busybox-based rootfs, Bluesound Node 2. Just go to their website and download it.

Their software is basically all in /root, in "encrypted" perl. I don't know if they're that stupid or just want to make it harder for script kiddies, but in case you're wondering - it's easy to crack. Even the official documentation warns about this - https://metacpan.org/pod/Filter::Crypto#WARNING .

The interesting part about it isn't their custom software though, it's the MQA decoder. They implemented it as an ALSA rate converter which you can plug into the "rate" plugin. See ie. /etc/asound/asound.conf.simple. The "ssc" converter is provided by /usr/lib/alsa-lib/libasound_module_rate_ssc.so which is an ARM EABI5 shared library, so you either need to compile and run your alsalib-using program on ARM or use ie. qemu-user to run it on x86 or elsewhere. It should be then trivial to put ie. pcm.file as slave to get the output in raw PCM samples (or with WAV metadata).

Probably not of much practical use as there's no encoder, but it still could be interesting to see what it does when passed samples with mqa-encoded metadata (valid song input).

Thanks for this, I'll have a look. The programming is probably way above my head, but it may help me a bit.

 

[onlyconnect]

Wrote up a few observations on my blog:

http://gadgets.itwriting.com/3540-more-on-mqa-and-tidal-a-few-observations.html

One of the things I find interesting is that the Tidal master of Remain in Light seems to be different and superior to any of my CDs. I don't know if for example HD Tracks has an equally good master.

Tim

 

[Jim Audiomisc]

MQA is very clever imo. It doesn't appear to 'add' anything to the signal, but instead is more faithfull to the original analogue recording because doesn't remove the parts from the recording which are actually fundamentally important in replicating atmosphere and realism in playback, previously compressed out of the way. Instead what MQA does is faithfully contain the information in a 44.1 container - how? - the term 'folding' is used instead of compressing, the important frequency area of realism and atmosphere (which MQA relies on for its SQ improvements) has been folded away and hidden from non MQA dacs, It's a clever way of getting a 196 file into a 44.1 container. The only way of getting the original file back out (whether it be 24bit 192 or whatever) is with an MQA DAC although higher than CD res files are still available without an MQA DAC, with only three of the four folds being decoded... hence Tidals 'Master' moniker..

Not quite, m'lud.

As both analysis and Julf's spectral measurements indicate, it does, indeed, 'add something' to the 44.1k. Indeed, it *has* to give you something that differs from plain 44.1k downsampled in order to function. The 'folding' will add HF components that are anharmonic with the LF. The 'bitstacking' adds patterns to the least significant bits which are then potentially altering plain LPCM replay. It may also reduce the resolution of LPCM replay because it has lost the lowest bits that would otherwise be devoted to LPCM indication.

The *MQA 'meaning'* of these changes is 'hidden' from a non-MQA DAC. But the changes made to what otherwise would be plain downsampled LPCM aren't. They are simply misinterpreted, altering the result. What is unclear is the details of the changes and what audible results they may have, etc.

 

[mattgbell]

That raised the curious question of: How on Earth would they have "Authenticated" about 200 albums in a matter of a few days? Indeed, how have they even traced exactly what ADCs and pre-processors were used so quickly? It seems more like, get out each tape and run it though the MQA decoder with some standard presettings.

I think you're right.

It has also been pointed out that for much (most?) recent pop music, it will not be possible to identify a single ADC, as the recording/producing of each instrument will have been done separately, using different equipment and often even in different studios.

 

[adamdea]

I think you're right.

It has also been pointed out that for much (most?) recent pop music, it will not be possible to identify a single ADC, as the recording/producing of each instrument will have been done separately, using different equipment and often even in different studios.

good point.

 

[Julf]

One of the things I find interesting is that the Tidal master of Remain in Light seems to be different and superior to any of my CDs. I don't know if for example HD Tracks has an equally good master.

Based on spectrograms of the 24/96 version of Remain in Light that I purchased (quite a while ago) I would treat it with suspicion. A clearly dropping HF, with a discontinuity at 22 KHz, and something that looks like random noise above that.

 

[onlyconnect]

Based on spectrograms of the 24/96 version of Remain in Light that I purchased (quite a while ago) I would treat it with suspicion. A clearly dropping HF, with a discontinuity at 22 KHz, and something that looks like random noise above that.

Sure, but the ultra HF content in itself doesn't interest me that much (since I can't hear it)!

Tim

 

[mark-hants]

I think you're right.

It has also been pointed out that for much (most?) recent pop music, it will not be possible to identify a single ADC, as the recording/producing of each instrument will have been done separately, using different equipment and often even in different studios.

Maybe "MQA" have come up with an algorithm that has a go at fixing a track that has already been mixed.

Or maybe they can only do tracks where the original multi-track masters still exist?

Here's a thought. Perhaps what MQA does in the decoding is to pulse the DAC in such a way that it produces an additional sine wave simultaneously which supplements and/or cancels the flaws in the "master wave"?

Maybe a little bizarre but it would explain is why it has to be on the playback side, too, and why it has to be hardware based.

 

[Werner]

Indeed, how have they even traced exactly what ADCs and pre-processors were used so quickly?

I never believed they would realise it to the letter of their claim, except perhaps for a handful of selected audiophile recordings. What I expect(ed) to happen, at best, was a quick identification of the architecture of the dominant convertor in the music creation path: non-oversampled minimum phase versus oversampled half-band linear phase. This puts the Sony, 3M, and Mitsubishi recorders as widely used in the 80s in one lot, and everything since the advent of delta-sigma ADCs in another lot.

The first lot has no pre-ringing (*), some non-linear phase distortion, and virtually no aliasing (**). One could reconstruct such a recording with a linear phase half-band filter, and possibly some phase correction.

The second lot has pre-rining (*), no phase distortion, quite a lot of aliasing. One could reconstruct with a minimum phase filter cutting below Fs/2, i.e. Meridian's flavour of apodising.

Beyond this it get overly complex very quickly, for little or no gains. Except in the wizz bang marketing dept, of course.






(* Not that this is of importance.)

(** I once looked at the AA performance of the old Sonies: with real music this is remarkably fine.)

 

[Stunsworth]

That raised the curious question of: How on Earth would they have "Authenticated" about 200 albums in a matter of a few days? Indeed, how have they even traced exactly what ADCs and pre-processors were used so quickly? It seems more like, get out each tape and run it though the MQA decoder with some standard presettings.

I don't know.

Thinking about it, the albums on that spreadsheet are ones reported by members of the general public - it's not a list provided directly from Tidal. It could be that they've all been there from day 1 but were only added to the spreadsheet when a user came across them.

I've noticed myself that quite a few of those albums don't show up on the 'What's New > Master' screen, Also the albums on that screen vary from time to time.