Tidal adds hi-res audio streams with 'Tidal Masters' (with MQA)

Discussion in 'General Audio Gear Discussion' started by weldp, Jan 5, 2017.

Tags:
  1. Dino

    Dino Friend

    Friend
    Joined:
    Jun 21, 2016
    Likes Received:
    1,452
    Dislikes Received:
    2
    Trophy Points:
    93
    Hearing background sounds more vividly at low to (somewhere around) medium volume is a classic effect of greater dynamic range compression, IME. I am not saying that is what you are experiencing @TheIceman93 . This is the just the first thing I think of when I hear that description.

    Unless we can compare the exact same mastering with and without MQA, I don't think we will really learn anything. I don't know if the tools (comparing EAC logs or null testing) to verify that the mastering is identical (or not) will apply with MQA involved.
     

  2. trung225

    trung225 Acquaintance

    Joined:
    Oct 22, 2015
    Likes Received:
    98
    Dislikes Received:
    6
    Trophy Points:
    18
    I registered for Tidal today, and noticed that I have owned at least 6 or 7 albums "Master" from Warner Classis as flac 24-96, which was bought at qobuz 7 months ago and definitely the same master as Tidal MQA version. So tomorrow I will test them to know what the fuss MQA is and will report back.
     
  3. Merrick

    Merrick Friend

    Friend
    Joined:
    Jan 6, 2016
    Likes Received:
    2,934
    Dislikes Received:
    4
    Trophy Points:
    113
    Location:
    Portland, OR
    I suspect none of these are unique masters. Most likely these are albums that are already available in 24/96 and now they've gone through MQA compression.
     
  4. Dino

    Dino Friend

    Friend
    Joined:
    Jun 21, 2016
    Likes Received:
    1,452
    Dislikes Received:
    2
    Trophy Points:
    93
    I have similar suspicions, @Merrick

    On the other hand, MQA is promoted as being data compression / (optional) expansion + higher fidelity. That makes me uncertain that nothing would be done to alter the sound. There is that playback process that is supposed to correct for the ADC used in the transfer. I am skeptical of that, but who knows. An easy way to make a change (and to a lot of people an improvement) is to add some dynamic range compression/peak limiting. It seems like SACD mastering tended to do that frequently. Although they didn't often have Hi-Res PCM transfers laying around in the early days of SACD.

    (I paid attention to MQA in the early days. I have not kept up, other than reading this thread.)
     
    Last edited: Jan 7, 2017
  5. Merrick

    Merrick Friend

    Friend
    Joined:
    Jan 6, 2016
    Likes Received:
    2,934
    Dislikes Received:
    4
    Trophy Points:
    113
    Location:
    Portland, OR
    I still have yet to see a clear, concise outline of exactly what the MQA process does to a file. At first I thought it was just a new compression scheme for hi-res, but now I'm hearing all this stuff about correcting for ADCs and being matched with individual DACs and I'm lost as to what the hell MQA is actually doing.
     
  6. Dino

    Dino Friend

    Friend
    Joined:
    Jun 21, 2016
    Likes Received:
    1,452
    Dislikes Received:
    2
    Trophy Points:
    93
    That is basically why I didn't keep up with reading about MQA. The more I read the less I knew.
     
    Wfojas, bazelio, Kattefjaes and 7 others like this.
  7. Vansen

    Vansen Friend

    Friend
    Joined:
    Nov 19, 2016
    Likes Received:
    371
    Dislikes Received:
    1
    Trophy Points:
    63
    Location:
    NYC-ish Area
    This goes well beyond any MQA BS or encoding, but is a good example of why some people should never be allowed to master an album:

    Both tracks are Only Women Bleed from Alice Cooper's Welcome to My Nightmare.

    tidal.com/track/68714508

    tidal.com/track/3387192
     
    Changeling likes this.
  8. Stride

    Stride Acquaintance

    Contributor
    Joined:
    Nov 26, 2016
    Likes Received:
    74
    Dislikes Received:
    0
    Trophy Points:
    13
    Location:
    Sydney
    Can someone explain how this MQA process works? Please correct me if I am wrong but from my understanding MQA are the digital product of sound engineers putting some DSP to an original hi-res digital file to correct ADC timing errors that get signed off by recording executive and/or recording artist saying that it sounds really really good, then gets compressed to a CD size format. So I am assuming Tidal MQA files are just tonally different then redbook versions digitally before it even gets to the DAC?

    In DAR's article, there was suppose to be a benefit for people who do not own MQA-DACs. I do hear differences between MQA and Redbook songs on Tidal when I disable Tidal's decoding through the Bimby. I am undecided on whether it is better.
     
  9. Vorlon

    Vorlon Acquaintance

    Joined:
    Jan 3, 2017
    Likes Received:
    88
    Dislikes Received:
    0
    Trophy Points:
    18
    Location:
    Finland
    The thing with high resolution audio is that most of the masters used for them generally have higher dynamic range than the CD versions. That is the prime thing in making them sound better. That´s also why vinyl usually sounds better than digital.

    Most music these days is compressed to hell and back to make it as loud as possible from portable sources with limited amplifying power. I don´t hear much difference when comparing older CD versions (just compare Metallica´s early CD versions with anything from 2000 onwards) with equal dynamic range high res audio.

    EDIT: I might be quite wrong on the the dynamic range thing, my bad (I had that impression of the HDTracks.com stuff that I´ve bought though, for example the Michael Jackson stuff is way higher DR than the remasters on CD these days)
     
    Last edited: Jan 7, 2017
  10. Merrick

    Merrick Friend

    Friend
    Joined:
    Jan 6, 2016
    Likes Received:
    2,934
    Dislikes Received:
    4
    Trophy Points:
    113
    Location:
    Portland, OR
    I'm a fan of hi-res but I would not say that most hi-res has a higher DR than the CD.
     
    Stride, SSL, Dino and 2 others like this.
  11. Dino

    Dino Friend

    Friend
    Joined:
    Jun 21, 2016
    Likes Received:
    1,452
    Dislikes Received:
    2
    Trophy Points:
    93
    If a customer sought out and paid for what is claimed to be better sound quality, then it would seem logical if Hi-Res and high DR tended to go together. But that is not the case, unfortunately. You can see that here: http://dr.loudness-war.info/
     
    Stride, Merrick, Vorlon and 2 others like this.
  12. trung225

    trung225 Acquaintance

    Joined:
    Oct 22, 2015
    Likes Received:
    98
    Dislikes Received:
    6
    Trophy Points:
    18
    My impression about Tidal MQA vs Flac 24-96. It's is just non-scientific observation, so please take it as a grain of salt.

    My test setup: Laptop > Mutec MC-1.2 via USB > Schiit Gungnir Multibit via BNC Cable > JBL LSR305s via XLR cable. (speakers) or > Schiit Vali > HD650 mod (headphones).
    Tidal MQA tracks are played by Tidal Desktop app, output WASAPI exclusive mode. Flac 24-96 tracks are played by Foobar and Jriver 20, output WASAPI too.

    To compare output level between Tidal MQA and FLAC 24-96, I only use Sound Meter app on my smartphone. The smartphone placement is near my right speaker with distances from it to noise sources (laptop fans and HDD fans) are 1 meter and 1.5 meter rspectively. At this position, my test room noise is around 16-19db (no music playing).

    Test tracks: Five album by Herbert von Karajan (Bruckner 7, Mozart 38, Dvorak 9, Promedean Concert, Sibelius 2). All of them are officially and newly remastered album for the 25th anniversary of the death of Herbert von Karajan on 2014, so I think they are the newest remastered version. My FLAC 24-96 albums are bought from qobuz, and also available from hdtracks, prestoclassical. The timing between two version is also the same.

    First of all, when using the low-file Sound Meter app on my smartphone, I confirm that MQA version and FLAC 24-96 version has the same dynamic range. Loudest peaks are the same. For example, the "Orpheus in the Underworld" Overture tracks has 2 peak moment at 67dB and 69dB, and it is true for both of them. The lowest volume is around 22dB-23dB for both version. The similar situation is also happened with other 4 albums. So, at least with some albums from "big" recordings company like Warner Classics, I can tell that MQA process does not alter the dynamic range when using the same master file.

    Secondly, I can tell that MQA version and FLAC 24-96 version, though derived from the same master, sounds different. The difference between them are very obvious, several times bigger than difference between FLAC 24-96 played by Jriver and Foobar. The first thing I have noticed is MQA version sound leaner, cleaner and smoother. Because it sound leaner, one can feel more air in them than playing normal FLAC 24-96 version, and sometimes it is easier to notice obscure details from the mix. For example, on 4th movement of Bruckner Symphony No.7 album, from 2m08s is the passage when Woodwind and Violins plays the main theme with Cellos and Contrebasses play pizzicato as support. Because of leaner and cleaner sound in MQA version, it is very easy to track the melody line of pizzicato, which is pretty muddy in old remastered CD version, and only really viable with FLAC 24-96 using good DAC like Gumby. Sound good? Yes, but only for the first time. What's annoying is when looking back to when and where EMI (now parts of Warner Music) records this album. This album was recorded in 1970, and in Jesus Christus Kirche, Berlin, and no way in hell the music plays by full orchestra in the church can sound that cleanly. And I also feel this process also reduce some microdetails, especially reverb detail for the sake of smoothing. Seems like MQA process is some sort like Cedar remastering, which in the effort of clean the sound also remove useful information.

    Even worse is the weird things happens sometimes with imaging. That is not the problems with some recordings, but very apparent on some. The example again is pizzicato section in 4th movement of Bruckner 7 albums. With MQA version, the imaging is off. Let me describe, with only Cellos and Contebasses plays, the sound from pizzicato must come from middle to right speakers. And the main melody, which plays by Woodwind and Violins, must come from middle to left speakers and occupy a bit higher, because Woodwind musician sits higher than Violin, Cello. With FLAC 24-96 version, the imaging is good. But in MQA, the sound from pizzicato is far lower than the main melody, and I feel there are almost no sound in between. That is so annoying and unfaithful to the real situation.

    Too bad but I feel MQA will soon become mainstream. Firstly, record company will never pass the opportunity to make music lover buy the same music over again. Secondly, while sounding a bit unnatural and weird imaging, the overly smooth of MQA file will attract so many people, especially with not-so-good gear. For example, someone with Fiio E17k will like the easy-to-noticed details through MQA presentation, and one with lowly and harsh sounding ES9018 DAC will like the extra smoothing aspect of MQA. And not everyone has obcession with imaging and natural presentation.

    Personally, I don't oppose MQA idea but don't follow it closely. I don't use music streaming for critical hearing as I prefer owned music for myself rather than renting them. But when comparing side by side with FLAC 24-96 on the same master, I feel that that MQA is dud. First, it didn't correct ADC and record process as advertisement in any mean. No good record and mastering can let the weird imaging happens. Second, I sound too cleanly to the point of feeling unnatural. Third, in related to poor sounding and bad imaging, there will be a huge loss if record companies stops selling Hires FLAC and only provide that MQA "Master" to download customer.
     
    Garns, Wfojas, Grahad and 10 others like this.
  13. drfindley

    drfindley Secretly lives in the Analog Room - Friend

    Friend
    Joined:
    Sep 25, 2015
    Likes Received:
    1,096
    Dislikes Received:
    1
    Trophy Points:
    93
    Location:
    Palo Alto
    Great writeup @trung225! The moment I heard that DSP is applied to the file is the moment I wanted nothing to do with MQA. I have a probably unpopular opinion that DSP of about any kind kills the naturalness of music. What you're describing sounds like my experience with DSP: "fixes" some things, but kills others.
     
    Stride, landroni, murray and 3 others like this.
  14. SSL

    SSL Friend

    Friend
    Joined:
    Nov 12, 2015
    Likes Received:
    686
    Dislikes Received:
    2
    Trophy Points:
    93
    This is just me talking out my ass, as I haven't heard MQA, but this just looks like another push to get music under DRM. Yawn.
     
    drfindley likes this.
  15. Merrick

    Merrick Friend

    Friend
    Joined:
    Jan 6, 2016
    Likes Received:
    2,934
    Dislikes Received:
    4
    Trophy Points:
    113
    Location:
    Portland, OR
    Just to expand on this. Most hi-res releases are new remasters. Some remasters have been done with a lot of care, like the recent Rush remasters and Tom Petty for two examples. Some of these are the definitive digital versions. McCartney has released some dynamically unlimited versions of his albums and they sound incredible. There have been some good jazz remasters, etc. These releases to me are audibly more engaging than the red book release, even if the mastering is the same (although I do have some exceptional red book releases too).

    However a lot of hi-res releases are heavily compressed. This is probably 80% of the hi-res market to be honest. Some hi-res releases aren't even real hi-res, just upconversions from older 16/44 masters. There's a distressing lack of provenance on most releases, leaving studious consumers to do the leg work by analyzing DR and waveforms. It's BS. The studios should be transparent about this.

    I agree, great write up! More informative than a lot of what I've read on MQA. As for DSP, I think it's best left applied by the individual if they want to suit their taste. Applying it to the file at the source takes the choice away from us and leaves us with a potentially compromised product.

    Agreed. That and streaming.
     
    Wfojas, drfindley, Vorlon and 3 others like this.
  16. Psalmanazar

    Psalmanazar Number 1 Big D Superfan!

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    1,410
    Dislikes Received:
    15
    Trophy Points:
    93
    Thanks! Great impressions of whatever their DSP is doing. Cutting the room mud is seriously annoying for a lot of stuff that was mixed to leave it in.


    This stuff isn't even hi-rez as analog tape as a medium is less dynamic than Red Book and almost all these masters could be printed on a normal CD. The best analog tape has something like the equivalent dynamic range of 14-bit digital and so few productions actually used that; the real dynamic range of an LP is much lower given surface noise. There's just no reason for most of this stuff to exist as many recordings have had good CDs made already from undegraded mix tapes; the labels are just selling a lie especially when the Yggdrasil has the best SNR and only is what? 21-bits? No D-S DAC that can decode MQA will come close to that.

    The best part is no commercially released music recording has taken full advantage of the Redbook spec and you probably wouldn't want one that did either; the most dynamic track found so far only measures with a difference of 36 decibels and that's on a poorly mastered mediocre death metal album: http://dr.loudness-war.info/album/list/dr-max/desc
     
    Stride, trung225 and Gaspasser like this.
  17. trung225

    trung225 Acquaintance

    Joined:
    Oct 22, 2015
    Likes Received:
    98
    Dislikes Received:
    6
    Trophy Points:
    18
    Personally, I also think Redbook is enough for music. But when come to sound mastering and recording, there are a problem: most of sound mastering equipments and softwares nowadays work on 24 fixed (or 32 float) bit system. I don't know why, but I suspect there are possibly some advantages when doing the mixing or remastering in high bandwidth, high bit rate environment. Downmixing and SRC from 24bit-xxxkHz to 16bit-44,1kHz is very painful with all the trade off between aliasing, dithering,... @Serious and I had a small experiment with that downmixing and SRC, and I feel the best is avoiding them whenever possible. So, in theory, Redbook is enough for me, but in reality, if I must choose between 16-44.1 and 24-96 version of the same mastering, no doubt I will choose the 24-96 version to avoid the extra stage in downmixing.
     
    Merrick, Psalmanazar and Serious like this.
  18. peterInVan

    peterInVan Rando

    Joined:
    Mar 27, 2016
    Likes Received:
    0
    Dislikes Received:
    0
    Trophy Points:
    1
    MQA, Tidal, and Roon.... Please please add and publish the metadata for all "remastered" and "MASTER/MQA" albums and be transparant as to where the master source came from and what it was converted to.
     
  19. Psalmanazar

    Psalmanazar Number 1 Big D Superfan!

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    1,410
    Dislikes Received:
    15
    Trophy Points:
    93
    depends on what plug-ins and software they're using. Most of the professional stuff is better than the prosumer and the free SoX is easily much better than most prosumer graphical stuff but command line only. I can definitely hear a difference between SoX and those.
     
    trung225 likes this.
  20. Merrick

    Merrick Friend

    Friend
    Joined:
    Jan 6, 2016
    Likes Received:
    2,934
    Dislikes Received:
    4
    Trophy Points:
    113
    Location:
    Portland, OR
    It is entirely possible that my preference for hi-res is all based on placebo.
     
    Psalmanazar and trung225 like this.

Share This Page