[DelsFan]

...I am listening via a schiit stack, and schiit does not support mqa... When I am listening to master quality tracks on tidal through my desktop and schiit stack, am I getting any improvement over normal hi-fi quality?

 

Seems to me that if MQA is everything it's cracked up to be you should not have to ask!

 
But that's the whole point.  It seems everyone is comparing Tidal MQA to Tidal hi-fi to Spotify to CDs to... whatever, without incorporating an MQA compatible DAC.  (Sometimes saying they've compared several times and can't hear any difference... listening with only their i-Phone and earbuds - but not in this forum).
 
Isn't that like saying (20 years ago), "I don't get all this 5.1 surround crap, I play DVDs that are supposed to have great sound through my integrated NAD 3020 (I just use the "B" speaker output for the two rear speakers) and, truthfully, selecting 2.1 stereo sounds a lot better than when I select 5.1."
If you don't have a surround processor (now, built into your receiver), why would you expect 5.1 or 7.1 or 9.1 to wow you?
 
Or (back in the even older days), if you had a Dolby encoded VHS tape but your VCR didn't encode Dolby, that wasn't a reason to say Dolby didn't enhance the quality and was a crap format/encoding/whatever.  It just meant your hardware wasn't equipped to take advantage of their "technology".
 
How is it different with MQA?  Isn't 2% of the improvement due to (maybe) careful selection of master recordings and 98% of the improvement due to the decoding done by one's MQA compatible DAC?  
 
It seems Danger Toast has asked exactly the right question; it's not like your laptop/macbook/desktop is doing the decoding, correct?
 
Shouldn't we hear from people who've compared MQA vs any other medium, including CDs, using their MQA capable Mytek, Meredian, Bel Canto, Aurender, Cary Audio, etc DAC?  Isn't that what the guys from "The Absolute Sound" and "Stereophile" did, and that's why they rave about its potential?  Or am I missing the point somewhere?
 
(So, I'm looking to start my high end headphone system with a MQA capable DAC with an I2S input - good luck, right?)
 
 
Edit: 
Woops, was typing (and warming lunch) as the previous poster explained much of how MQA really works.

 

[L8MDL]

My position is that any improvement heard with MQA is expectation bias. I know of no certified test, using identical masters, to dispute that.

 

[DelsFan]

My position is that any improvement heard with MQA is expectation bias. I know of no certified test, using identical masters, to dispute that.

But they don't claim to offer anything better than the Master Tapes, that would be dumb.
 
They do claim that streaming from Tidal Hi-Fi MQA and going through an MQA enabled DAC sounds better than any other streaming method, and that it also sounds better than many/most CDs.  This may or may not be true, but respected people who for several decades have made their living writing about high end audio suggest it is - and on the face of it, is not impossible at all.
 
There are plenty of CDs that don't sound as good as the Masters, and even several releases of the same CD (not including remastered or newly released SACD pressings) where one release by one company sounds better than the original release, or some of the subsequent releases.  Some of the early Bob Ezrin produced Alice Cooper albums are one case in point (and if you're into that kind of thing, they were masterfully done - by, at the beginning, a snot-nosed teenager (Ezrin)).  Again, if an SACD sounds better than a "regular" one (RedBook?) it seems not impossible that MQA streamed music could sound better than regular streamed music.

 

[martyp87]

I would like to compare MQA vs DSD but I only have a couple of DSD albums, the most common one "Thriller" is only FLAC Lossless at the moment on TIDAL and I must say the DSD tracks sound so much cleaner (OPPO HA-1 > HD800S). 
 
My other comparison is with Ed Sheeran divide album. Castle on the Hill on MQA and M4A (iTunes Purchased) I honestly can not tell the difference at all. The track still sounds heavily distorted, it sounds great in the car but on my Egg 150 monitors or headphones it sounds awful. 
 
I won't complain though as I don't pay any extra for MQA over FLAC Lossless. If I did, I probably wouldn't bother.

 

[manpowre]

My position is that any improvement heard with MQA is expectation bias. I know of no certified test, using identical masters, to dispute that.

Improvement over what are you referring to ?
If its improvement over 16/44/48, MQA is definetely better variant as it has lower noise floor with 24 bits and thereby higher dynamic range. About 44/48 and 96/192 khz discussion, it extends the audible hertz from way above audible area of humans that can pick it up at 20-22khz, where most of us are at 18-19khz. Some claim we can "feel" the higher tones above audible range.
 
Personally I find the FLAC 24/96 files more accurate in sound reproduction, but I only own 1x 2015 album that is currently MQA. I have found another album now that I might purchase once I get my laptop back from guarantee issue.
 
I think the issue with MQA is definetely that we dont have encoders available for the mass to test and review content. Why its not available is a damn good question!

 

[DangerToast]

Hi @DangerToast, 
 
Take a look at this link for proper MQA explanation:
http://www.audiostream.com/content/mqa-decoding-explained
 
MQA has 3 unfolding levels, 2 of them are software decoded, while last level is hardware decoded only by MQA certified DAC's. 
The first level is 24/48 which has better noise floor than 16/44/48.
 
The second level is 24/96 which is software decoded in Tidal only when you see the "MASTER" lower right in Tidal. BUT, you have to change your driver to support this, so if you are on windows, right click your sound icon in notifications, and change the properties for your playback device. Also disable any enhancements, AND set volume in driver to 100%, and the windows volume to 100%. Once you change your driver, depending on your setup, you should see 24/96 show up on your DAC.
 
The third level is 24/192 which is only for hardware MQA enabled DAC's. This is where the big fuzz is right now as MQA want each DAC provider to pay MQA group for the DACs that are going to be MQA certified. There is a MQA light (blue?) that is enabled when the third level is unfolded on the DAC. Also MQA doesnt provide an MQA encoder for people around and DAC providers to actually test and review the MQA encoding/decoding process.
 
Personally I set my driver to 24/96 as I usually play those files from Foobar2000, and I also listen to some MQA on Tidal, and those times I listen to new content I end up with normal HIFI, but I dont bother to change to the 16bit version which the HIFI is.
 
@L8MDL, completely unnessasary comment as people need to ask if they dont fully understand things which is the sole purpose of forums right? 

 


Thanks for actually answering my question, because of you I'm actually getting some 24/96 goodness out of Tidal. I didn't know you had to change the bitrate in windows. One follow up question: when it asks me to specify a bitrate in windows, I went ahead at chose 32/192. I assume I'm just establishing a "ceiling" here and when I listen to masters on Tidal it's actually feeding dac/amp 24/96, correct?

Also, I didn't mean to wade into an MQA civil war that seems to be happening on this thread, but here's my two cents: If MQA gets me some 24/96 goodness on Tidal (as opposed to regular CD quality), that's a good thing. I guess it sucks that you cant get the full 24/192 without the proper dac, but it's still an improvement.

 

[tjw321]

  Improvement over what are you referring to ?
If its improvement over 16/44/48, MQA is definetely better variant as it has lower noise floor with 24 bits...

I believe that the lower 8 bits are used to encode the MQA information so, if anything, it'll have a higher noise floor than 16/*.

 

[castleofargh]

indeed, the MQA file is presented in a 24bit container, but part of the data isn't bit depth information. @manpowre is wrong about that. again I can only suggest to go look at the patent as it's the only place where they never play on words to pretend like the format is more than it is. it's not the best read of the year, but it's factual information for a change.
each amplitude is a 24bit value, but only from 13 to 17bit(depending on various decisions/requirements) contain the actual music in the audible range, the rest if left undecoded is noise. and if decoded, then it's replaced by dither(nicer, harder to notice kind of noise). of course it can still be played and seen by the DAC as 24bit, just like we can take a 16bit file, output it to 24bit in our player by adding extra zeroes and the DAC will say it's a 24bit file. but the song's resolution is not 24bit. on MQA it absolutely never is!

 

[manpowre]

  indeed, the MQA file is presented in a 24bit container, but part of the data isn't bit depth information. @manpowre is wrong about that. again I can only suggest to go look at the patent as it's the only place where they never play on words to pretend like the format is more than it is. it's not the best read of the year, but it's factual information for a change.
each amplitude is a 24bit value, but only from 13 to 17bit(depending on various decisions/requirements) contain the actual music in the audible range, the rest if left undecoded is noise. and if decoded, then it's replaced by dither(nicer, harder to notice kind of noise). of course it can still be played and seen by the DAC as 24bit, just like we can take a 16bit file, output it to 24bit in our player by adding extra zeroes and the DAC will say it's a 24bit file. but the song's resolution is not 24bit. on MQA it absolutely never is!

I knew about this already, and considered to include it in my post, but I didnt want to dig too deep into the technical aspects of what actually happens on decoding level, but for the average user setting the driver to 24 bit / 96khz is the closest we can get to increased quality compared to 16/44/48 through streaming. The fact that MQA has a noise floor of 13-17bits is why we can hear a difference between real 24/96 and MQA 24/96, and also why we are discussing this heavily. Also dont forget that MQA advertises this as 24/48, 24/96 and 24/192. and this is what the average users will see as numbers.
 
@DangerToast, just set the driver to the closest you expect to play back. setting it to 32 bits doesnt mean you get better quality, in fact it might reduce quality as your DAC thinks it has to create sound based on 32 bits instead of 24 bits. Most modern DACs perform very very well in 24 bits mode, and extremely well in 16 bit mode.
 
@castleofargh, what you mentioned here is exactly why I dont like MQA, as the MQA group doesnt provide encoders for people to use and play with to test and review the output of the encoding/decoding process. I can only guess, but I think this is the sole main reason to why MQA group dont provide encoders for people!

 

[gregorio]

 
[1] There are plenty of CDs that don't sound as good as the Masters, and even several releases of the same CD (not including remastered or newly released SACD pressings) where one release by one company sounds better than the original release, or some of the subsequent releases.  
[2] Again, if an SACD sounds better than a "regular" one (RedBook?) it seems not impossible that MQA streamed music could sound better than regular streamed music.

 
1. That's not possible. If we're talking about commercial CDs (rather than bootlegged copies burnt to CD-R), they can only be manufactured from a master. If a release sounds different to another release, then it cannot be the same CD, it has to be a different master!
 
2. If an SACD sounds better than a redbook version, it's because it's a different master. It's not at all impossible that MQA could really sound better than a CD but it would have to be a different master. In which case we are comparing different masters rather than different distribution formats.
 

  [1] If its improvement over 16/44/48, MQA is definetely better variant as it has lower noise floor with 24 bits and thereby higher dynamic range.
[2] About 44/48 and 96/192 khz discussion, it extends the audible hertz from way above audible area of humans that can pick it up at 20-22khz, where most of us are at 18-19khz. Some claim we can "feel" the higher tones above audible range.

 
1. The noise floor of the container format is irrelevant when talking about 16bit or higher because no commercial recording I know of has a noise floor lower than about -60dBFS, which is 10bits. The noise floor we are going to hear is the noise floor of the recording, not the noise floor of the format which is 6 bits lower in the case of a 16bit container, 16bits lower in the case of 24bit container or 22bits lower in the case of a 32bit container. It's the same as: If you put 20ml of whisky in a 50ml glass you don't get more whisky if instead you put that 20ml of whisky in a 200ml glass.
 
2. More typically, with good young adult ears it would be 16-18kHz. I don't know of any reliable evidence to support the claim of being able to "feel" or sense the ultrasonic sound present in music.
 

  [1] ... but for the average user setting the driver to 24 bit / 96khz is the closest we can get to increased quality compared to 16/44/48 through streaming.
[2] The fact that MQA has a noise floor of 13-17bits is why we can hear a difference between real 24/96 and MQA 24/96, and also why we are discussing this heavily.

 
1. Setting to 24bit would not increase quality as there is nothing below about 10bit to increase, as mentioned above.
2. Even if MQA were only 13bit, that is still in excess of about the 10bit max of any commercial recording and as that 13bit noise floor is "shaped", unless they've made a huge foul-up with their algorithm, that noise floor should be completely inaudible beneath the recording noise floor. As with the whisky analogy, 20ml of whisky will not taste any different if drunk from a 50ml glass than from a 100ml glass, unless it's actually a different whisky!
 
G

 

[jwbrent]

For an appreciation thread, there is a lot of dissent here. Is it even possible to create a thread about MQA without the cynics, skeptics, and objectivists piling on?

I, for one, am happy with my TIDAL HiFi account. I do hear improvements when I listen to most Master files, especially in the upper frequencies. My only disappointment is the slowness of Master releases.

I use a MacBook and a Mojo for my TIDAL Masters listening. I feel most of the benefit of hi res files comes at 24/96, so having an MQA supported DAC is not meaningful to me.

I do wonder whether Audirvana users feel MQA files sound better than when using TIDAL's MQA software.

 

[headfry]

For an appreciation thread, there is a lot of dissent here. Is it even possible to create a thread about MQA without the cynics, skeptics, and objectivists piling on?

I, for one, am happy with my TIDAL HiFi account. I do hear improvements when I listen to most Master files, especially in the upper frequencies. My only disappointment is the slowness of Master releases.

I use a MacBook and a Mojo for my TIDAL Masters listening. I feel most of the benefit of hi res files comes at 24/96, so having an MQA supported DAC is not meaningful to me.

I do wonder whether Audirvana users feel MQA files sound better than when using TIDAL's MQA software.

thanks for your reply - I also enjoy the sonic/musical improvement of Tidal Masters. I'm not sure that there's much of an improvement in sq using Audirvana 3+ over Tidal desktop, but recommend trying the free trial; I do appreciate the following benefits:

1) allows use of AU (AudioUnits) plugins - so I can eq my Grado's to sound the way they should! I always felt the
desire to eq and now I have it! Using Marvel Voxengo geq - free and excellent equalizer!

2) eliminates the endless buffering issue of certain Masters albums - such as Yes Close To the Edge,
Fragile and Relayer

3) frequent updating of the software - excellent support!

For me, the two players sounded fairly similar but the above points, especially the ability to eq won me over!

 

[jwbrent]

thanks for your reply - I also enjoy the sonic/musical improvement of Tidal Masters. I'm not sure that there's much of an improvement in sq using Audirvana 3+ over Tidal desktop, but recommend trying the free trial; I do appreciate the following benefits:

1) allows use of AU (AudioUnits) plugins - so I can eq my Grado's to sound the way they should! I always felt the
desire to eq and now I have it! Using Marvel Voxengo geq - free and excellent equalizer!

2) eliminates the endless buffering issue of certain Masters albums - such as Yes Close To the Edge,
Fragile and Relayer

3) frequent updating of the software - excellent support!

For me, the two players sounded fairly similar but the above points, especially the ability to eq won me over!

Thank you for your feedback. Yes, having an EQ would be a nice addition.

With the recent cash infusion by Sprint, I'm hoping the desktop app will be further improved ... a directory for Master titles and artists are at the top of my list. Maybe they'll add an EQ, too.

 

[headfry]

I've been enjoying James Taylor's Gorilla (1975), Tidal Masters/software decoded
and comparing it to the HIFI version.

No contest, the Masters is obviously much, much better - and in the ways many/most Masters
tend to be better

Masters: coherent & consistent
open window on the performance
natural 3D soundstage and specific imaging
much more refined

comparing the HIFI version to Masters is like comparing a demo version of the song with the final

the HIFI version is flattened, brasher, blurrier, lacking in bass definition&weight and flattened soundstage
with large,mostly undefined imagery

anyone who says they can't hear a significant improvement through Tidal-decoded MQA
makes we wonder why!!!!~!

 

[Left Channel]

@headfry at 1x decoding 24/96 I still can't hear a difference. But thanks for your reply last time:

https://www.head-fi.org/f/threads/c...masters-and-more.831291/page-45#post-13393711