With a broader stream of MQA albums now available, I thought it would be good to have a place to discuss our impressions of the listening experiences for decoded MQA vs non-MQA hi-res versions.
I would suggest the following guidelines:
This is about the music and the listening experience – keep technical discussions of the merits/science of MQA out of it
For as much consistency as possible, make sure you are listening through Roon and through the same output zone and MQA-compatible DAC For both versions
List the hi-res version source and format you are comparing
List the equipment chain
Since it is likely that the two versions have different sound levels, try to adjust for that before you perform any critical listening
(Finally!) Mention your impressions of the two versions – timing, tone, brightness, warmth, tingles, tears…
I will be posting my own thoughts in the coming days, but my initial reaction is – intrigue. Definitely different. A few albums sound overly bright/crisp. But I really need more time before I can provide any comments with any level of confidence.
Do make sure you check and adjust for sound levels as you do this.
I have found some dramatic differences between some of my cd rips and the MQA versions. Generally so far the MQA versions have been a bit quieter than the version I have already. In once instance by as much as 6 dBA which was obviously noticeable.
However with Rumours I found the MQA version to be 4 dBA louder on average on Gold Dust Woman. This was one of those Albums where I noticed most improvement from MQA on an initial listen without doing any proper AB listening.
This is with undecoded MQA at the moment so you may find something different with decoded.
I would not use volume leveling to attempt this. Fortunately I have a digital preamp which displays volume controls in dB steps so having listened to the tracks once I can adjust the volume by the difference in the averag. I recorded. I’m just using an app on my phone for the dBA reading but making the corresponding adjustment on the amp from the phone readings nails the levels next time so it can’t be too far wrong at least for relative levels.
As @philr says, level matching is critical for accurate AB testing. Level differences as small as a couple tenths of a DB are enough to create preference for the louder version.
Also be careful that, given close level matching, a preference for one version of a given recording over another is likely due more to differences in the mastering decisions - limiting, compression, EQ, mixing, etc - than to MQA per se. For instance, as has been well-documented, many “hi-res” versions of pop albums have been significantly compressed relative to the “low-res” initial releases - see http://dr.loudness-war.info
Some albums from the 2L label are available on Tidal MQA and at the 2L site for free in the Test Bench. For example, I have the Magnificat in CD, 352, MQA download and Tidal MQA.
Plenty of opportunity to spend a weekend cork sniffing.
Whilst you are correct and we won’t know for sure if the improvement is in the masters or in MQA I’m not too concerned as if we are getting some improvements then I’m all for it no matter where they come from.
A couple of point though.
Firstly Warner has done a few 1000 albums in a fairly short space of time so there is not the time for substantial remastering. (Maybe they could eliminate some of the compression though as they run their master through the MQA process.)
If we are hearing consistent changes with MQA across a range of recordings than I think that this is likely due to MQA rather just any remastering.
I’ve not had time to do much critical listening. Most Tidal/MQA albums sound good or great through my Meridian 818v3. (I’m using Sooloos, not Roon) One album stuck out for me: Natalie Merchant’s Tigerlily in Tidal/MQA didn’t sound better than my RBCD FLAC copy. The Tidal/MQA version sounds much softer and almost muffled compared to RBCD.
Yes, I notice MQA tracks(fully decoded) sound a bit softer than the equivalent 16 bit 44.1k FLAC stream on the Tidal app. Example, Madonna ‘Like a Virgin’. If I disable software decoding, i.e, disable ‘Exclusive’ mode it sound a lot less ‘dynamic’, less grip and overall sounding dull. Therefore, whenever possible MQA playback must be properly decoded.
This is quite similar when I encountered playing DSD vs PCM…
I am listening to the mqa files (and some non-mqa hi-res files) through Roon, Tidal and fiies downloaded from 2L testbed.
My mqa system is a MacBook (Retina 12" early 2015), AKG Pro Audio K712PRO Studio Headphones, Meridian Explorer2.
I’ve had the Explorer2 for less than a week; the price is now $US199 (was $299). I wonder if Meridian is discontinuing the model and clearing the shelves, or if they are just “priming the pump” for mqa.
I will post more in a little while – but I got a hint from this community yesterday that I think greatly improved the listening experience. The Explorer will work right out of the box as long as you redirect the output to it.
Fiddling with the MacOS “Audio MIDI Setup” utility did not do anything I could detect (probably because I do not know what I am doing with it).
However, going to the Settings in Roon and Tidal seems to have had a BIG effect:
With Explorer plugged into a USB port on the Mac (and headphones plugged into the Explorer2):
In Roon:
Settings->Audio->Explorer2 USB DAC->Gears->Setings->Playback->[Select]“Use Exclusive Mode”
In Tidal (with a “HiFi” subscription):
Settings->Streaming->HiFi/Master->Built-In Output->[Select] “Use Exclusive Mode” and “Passthrough MQA”
Save Settings and restart the software
If you download an mqa file and put the file in your Roon directory, the file icon will show “mqa” in most views as long as you have selected “Show album format on browser” in the “General” tab of Roon Settings. Roon does recognize mqa files, but I think you need the hardware (DAC) to decode the mqa and turn on the blue light on the Explorer2.
More later - I apologize if you knew all of this before you started reading - but maybe it will help someone who is just starting.
For what it’s worth, a copy of my email to Dirac, detailing early listening experience with Tidal/MQA:
"Early results of listening to Tidal/MQA:
— Dirac works fine with Tidal/MQA tracks, as long as one runs Calibration Tool with higher bit rate and bit depth.
— MQA tracks have smoother, relaxed, less bright presentation compared to Redbook tracks from same album. Timbre seems more authentic as well and soundstage is larger.
— Tidal only offers software “unfolding” of MQA rather than full unfolding if one has a MQA certified DAC (I do not…). Tidal is therefore limited to 88 or 96 kHz.
— In addition to a Cambridge Audio Azur 851D DAC /Pre (non MQA), I have an Audioquest Dragonfly 1.2, that rumor has it will be updated to unfold MQA at a later date.
— Contrary to my praise for Dirac, I plan on reserving full judgement of MQA until further listening and developments, such as Audioquest rumor and ROON “unfolding” of Tidal/MQA tracks.
— MQA may be truly the “Next Big Thing”, or it may fall short of expectations, like Betamax, Quadraphonic, Laserdisc, HD DVD, Elcaset. Time will tell, and yes- I am old enough of remember those “Big Things””
There are some very different masterings of American Beauty out there. My older CD version is quite different from my 24/96 version of a much later remastering.
I posted a link elsewhere on the forum to a discussion re MQA With Bob Stuart, Bob Ludwig and others at the Montreal Audio Engineering Society Here
In this there is a lot of discussion on AB testing regards to the latest evidence in neuroscience as to why things are not as simple as you may think.
Also MQA goes to the best known first generation master for its source. A source that can be authenticated as the original.
It’s a long talk with Q and A but so worth study.
Rob, was Betamax a “Big Idea” or was home recording of video? I go with the latter and VHS did triumph. The significance of MQA is even if it doesn’t succeed, something will!