CES 2017: MQA announces TIDAL Masters, and more
Mar 3, 2017 at 6:27 PM Post #526 of 702
 
In a 44 kHz data stream there is no decoding. Some MQA files at that level may offer some benefit from having been re-digitized or in some way processed from whatever masters exist (there is wide debate about whether that can truly be called "remastering") but many others will not, and in general that is not what people are excited about. Please see this article: MQA Decoding Explained (Audiostream).  As for his Pioneer XDP player, again, that is far too small a market for Tidal, and I'm not even certain UAPP would or could develop for it. IMHO so we'll have to wait until more devices support at least 24/96.


I believe you are wrong here. A "full blown MQA recording" is encoded and stored in PCM at 44kHz. It can be uncompressed/unfolded back up to a higher spec using software or hardware... but the container is compatible with any android OS.

The 44kHz file stream goes to the hardware decoder in the Player and theere is unfolded into its slightly-lossy 192kHz glory.

Please quote directly from Bob Stuart ro MQA website on the actual MQA file itself, not the original source or ultimate destination resolution.
 
<snip>

 
I believe you're missing my point, which is only that

(a) at 44 kHz the file is not decoded/uncompressed/unfolded — which you repeated here, and I did not say it is incompatible with Android or iOS — and, 

(b) after listening carefully to Bob Stuart it is debatable whether each and every file at that unfolded stage is anything special, though some may be, and,

(c) most important — since this was the original topic — in my uninformed opinion there are unlikely to be many users of 16/44.1 devices interested in MQA, which means owners of more powerful devices like the Pioneer XDP plus the relatively small number of Android/iOS users who use external DAC are in total are probably too small a market to interest Tidal in updating their mobile apps for MQA.
 
OK? 
 
Mar 3, 2017 at 6:55 PM Post #527 of 702
   
I believe you're missing my point, which is only that

(a) at 44 kHz the file is not decoded/uncompressed/unfolded — which you repeated here, and I did not say it is incompatible with Android or iOS — and, 

(b) after listening carefully to Bob Stuart it is debatable whether each and every file at that unfolded stage is anything special, though some may be, and,

(c) most important — since this was the original topic — in my uninformed opinion there are unlikely to be many users of 16/44.1 devices interested in MQA, which means owners of more powerful devices like the Pioneer XDP plus the relatively small number of Android/iOS users who use external DAC are in total are probably too small a market to interest Tidal in updating their mobile apps for MQA.
 
OK? 


Yup.
 
Mar 3, 2017 at 6:56 PM Post #528 of 702
 
In a 44 kHz data stream there is no decoding. Some MQA files at that level may offer some benefit from having been re-digitized or in some way processed from whatever masters exist (there is wide debate about whether that can truly be called "remastering") but many others will not, and in general that is not what people are excited about. Please see this article: MQA Decoding Explained (Audiostream).  As for his Pioneer XDP player, again, that is far too small a market for Tidal, and I'm not even certain UAPP would or could develop for it. IMHO so we'll have to wait until more devices support at least 24/96.


I believe you are wrong here. A "full blown MQA recording" is encoded and stored in PCM at 44kHz. It can be uncompressed/unfolded back up to a higher spec using software or hardware... but the container is compatible with any android OS.

The 44kHz file stream goes to the hardware decoder in the Player and theere is unfolded into its slightly-lossy 192kHz glory.

Please quote directly from Bob Stuart ro MQA website on the actual MQA file itself, not the original source or ultimate destination resolution.
And some more detail from Bob on just how they can "steal space" in the 44kHz file to encode the data
 

the detail is that no MQA signal is ever 24bit(or I really misunderstood something crucial if it is). only the container is 24bit. Lavorgna misses that "detail" entirely with in his blog, and Stuart didn't feel the need to correct him in his comment for some reason
wink_face.gif
.
they work in a 13+3+whatever configuration to allocate data as bit values, the first 13 are good old PCM signal down to 13bit, the rest is the MQA salsa basically coding extra samples in the bit values. the bit usage can change depending of setting decision at the encoding if I got that right. probably depending on the target sample rate, maybe the amplitude of "signal" contained in the ultrasonic part of the master, ...
so you get a signal(not the file, the signal!!!!) in the audible range going from 13bit to ... I'm not sure, I'll go with 17bit as I think it's the biggest value I've seen in the patent.
if you happen to get a 13bit signal folded in a 24/48 MQA file and you use no software decompression at all, you end up with a 13/48 song that weights as much as a 24/48 song. not really ideal ^_^. that's why there really is no point in using MQA files outside of, at the very least, some software decoding. yes it's music in PCM container, and yes it works but it's lower resolution that a basic PCM album of the same file size. 
with at least software decoding you then will tend to get back at least part of the ultrasonic content resulting in let's say for the sake of my made up example, 13/96 music(or maybe 17/96 if that was the encoded bit depth target, I'm not very clear on that and couldn't find any clear answer), played from the container that was the 24/48 file. it's mainly and in a complicated way, a trade of bits for samples. not spontaneous generation of extra data pulled out of a digital magic hat.
 
Mar 3, 2017 at 7:27 PM Post #529 of 702
I believe you're missing my point, which is only that


(a) at 44 kHz the file is not decoded/uncompressed/unfolded — which you repeated here, and I did not say it is incompatible with Android or iOS — and, 


(b) after listening carefully to Bob Stuart it is debatable whether each and every file at that unfolded stage is anything special, though some may be, and,


(c) most important — since this was the original topic — in my uninformed opinion there are unlikely to be many users of 16/44.1 devices interested in MQA, which means owners of more powerful devices like the Pioneer XDP plus the relatively small number of Android/iOS users who use external DAC are in total are probably too small a market to interest Tidal in updating their mobile apps for MQA.

OK? 


So long as Tidal or others deliver the lossless 44kHz stream to the innards of whatever player can do MQA, the player should recognize it.

So its really just up to Tidal to have an "MQA On" button. Not too much work for them at all.
 
Mar 3, 2017 at 8:32 PM Post #530 of 702
So long as Tidal or others deliver the lossless 44kHz stream to the innards of whatever player can do MQA, the player should recognize it.

So its really just up to Tidal to have an "MQA On" button. Not too much work for them at all.

 
With software it's never that simple. Adding a feature is always more like opening a can of worms, especially in this case with the vast number of different Android models and Android OS customizations out there. 
 
Mar 3, 2017 at 10:55 PM Post #531 of 702
  Anyone heard word on the Android and/or IOS apps with MQA enabled?  I have a Pioneer XDP which is compatible from a hardware perspective but no Tidal software yet...

 
I'm eagerly waiting for this too :) I'm using an Onkyo DP-X1 myself. very tempted to get the Meridian Explorer v2 during Can Jam SG next week if it's going for a great price so that I can listen to some MQA goodness with Audirvana Plus 3, which is also due to be out in the near future.
 
Mar 3, 2017 at 11:27 PM Post #532 of 702
https://docs.google.com/spreadsheets/d/10VtON9VjMAt3uyHC2-Oo2MjIa3orv9DKZfwiRQKmTAA/edit#gid=945476039

Done. This lists 1,600 or so. So maybe there are more than 1,000. It didn't seem that way to me going through the masters list on tidal. It's a little depressing to think there's only like 5 or 6 albums in that list that I care about in MQA format, lol 


I did not know it reached 1600 albums, I was checking from Master catalogue :) That's great! I am eager to listen Maria Callas for example!
 
Mar 4, 2017 at 8:00 AM Post #534 of 702
   
With software it's never that simple. Adding a feature is always more like opening a can of worms, especially in this case with the vast number of different Android models and Android OS customizations out there. 


​There are two parts to the implementation and I can image one is harder than the other:
 
1) Adding MQA "passthrough" mode so that an MQA-enabled DAC can decode the stream (should be relatively easy).
2) Add software MQA decoding to the mobile app (relatively harder).
 
Mar 4, 2017 at 1:23 PM Post #536 of 702
  the detail is that no MQA signal is ever 24bit(or I really misunderstood something crucial if it is). only the container is 24bit. Lavorgna misses that "detail" entirely with in his blog, and Stuart didn't feel the need to correct him in his comment for some reason
wink_face.gif
.
they work in a 13+3+whatever configuration to allocate data as bit values, the first 13 are good old PCM signal down to 13bit, the rest is the MQA salsa basically coding extra samples in the bit values. the bit usage can change depending of setting decision at the encoding if I got that right. probably depending on the target sample rate, maybe the amplitude of "signal" contained in the ultrasonic part of the master, ...
so you get a signal(not the file, the signal!!!!) in the audible range going from 13bit to ... I'm not sure, I'll go with 17bit as I think it's the biggest value I've seen in the patent.
if you happen to get a 13bit signal folded in a 24/48 MQA file and you use no software decompression at all, you end up with a 13/48 song that weights as much as a 24/48 song. not really ideal ^_^. that's why there really is no point in using MQA files outside of, at the very least, some software decoding. yes it's music in PCM container, and yes it works but it's lower resolution that a basic PCM album of the same file size. 
with at least software decoding you then will tend to get back at least part of the ultrasonic content resulting in let's say for the sake of my made up example, 13/96 music(or maybe 17/96 if that was the encoded bit depth target, I'm not very clear on that and couldn't find any clear answer), played from the container that was the 24/48 file. it's mainly and in a complicated way, a trade of bits for samples. not spontaneous generation of extra data pulled out of a digital magic hat.

 
I think you are getting into the habit of looking at numbers and thinking larger numbers == better sound. Just use your ears to demo MQA and keep in mind this is streaming music, for streaming music, it is some of the best quality around.
 
If you have dedicated files then you'll probably want to use those if you have a ton of 24/96khz or better stuff or DSD stuff. But even then the quality of the original mastering affects greatly it's sound, and again, it's not the same market as people that use streaming services. The advantage of streaming services is people who are discovering new music, not listening to their favorites at the highest quality, and so, although MQA is a lossy mechanism, it still is fulfilling a market segment. 
 
Mar 4, 2017 at 9:35 PM Post #538 of 702
  I think you are getting into the habit of looking at numbers and thinking larger numbers == better sound. Just use your ears to demo MQA and keep in mind this is streaming music, for streaming music, it is some of the best quality around.
 
If you have dedicated files then you'll probably want to use those if you have a ton of 24/96khz or better stuff or DSD stuff. But even then the quality of the original mastering affects greatly it's sound, and again, it's not the same market as people that use streaming services. The advantage of streaming services is people who are discovering new music, not listening to their favorites at the highest quality, and so, although MQA is a lossy mechanism, it still is fulfilling a market segment. 

eheh, I'm not thinking that larger numbers equal better sound, at least not from an audibility perspective(I fail to consistently identify 320mp3 so... ^_^). I just don't want people to get the wrong idea that somehow it's all bonuses and unicorns in a """small""" streaming format. marketing does marketing, and I'm here to be grumpy about everything as ordained by the yin & yang forces in the universe. or maybe I just like to complain a lot. but it's one of those 2 options for sure. 
evil_smiley.gif

for bit depth in particular, in my own tests(non MQA related) 12 or 13bits were about the bit depth limit for transparency even at what I consider loud on rather calm passages, and very much transparent to me at normal listening levels. add a nice dither just in case and all is well. my post didn't mean that not having 24bit would sound bad, I don't believe that. I'm definitely a threshold guy, not a higher is better dude.
wink.gif
 
 
Mar 5, 2017 at 6:58 AM Post #539 of 702
   
I think you are getting into the habit of looking at numbers and thinking larger numbers == better sound. Just use your ears to demo MQA and keep in mind this is streaming music, for streaming music, it is some of the best quality around.
 
If you have dedicated files then you'll probably want to use those if you have a ton of 24/96khz or better stuff or DSD stuff. But even then the quality of the original mastering affects greatly it's sound, and again, it's not the same market as people that use streaming services. The advantage of streaming services is people who are discovering new music, not listening to their favorites at the highest quality, and so, although MQA is a lossy mechanism, it still is fulfilling a market segment. 

To be honest, I did almost 2 weeks of MQA listening from Tidal, and I stopped as I found the MQA decoding to be louder (3-6db depending on albums) and they were obvious same master. Knowing that MQA has a 13 +3 bit dynamic range I suspect (personally suspect) that the encoding process increase the loudness to lift it up from the 13 bit noise floor, and that makes us "think" MQA sounds better.
 
I found a few high dynamic albums with alot of drums and instruments in MQA library from Tidal, and it sounds so much better on 16/44.1 and even 24/44.1 since these are high dynamic albums they sound more like how they mastered them, and not the MQA stream that increases the loudness.
 
At the moment Im working on installing a linear power supply to my ADC as it has some artifacts due to EMI, and once that is done in a few weeks (just ordered a new linear PSU) I will start measuring the MQA albums on tidal to prove my point through FFT spectrum. 16bit vs MQA bits (apparently 24 bits, which we now know is not).
 

Users who are viewing this thread

Back
Top