Big MQA News (Really - Tidal and Software Decoding, etc...)

[Dr Tone]

Am I correct in assuming that the Tidal-based software decoding is still not presenting the full MQA file to the non-MQA DAC? If it is, I'm perplexed. Was this not the sole means by which MQA was set to generate revenue?

 

It looks like the generic profile software decoding isn't outputting the higher sample rates that have been reported as showing on certain MQA certified DACs. We also know that software decoding isn't doing per DAC "sweetening" for a lack of a better word that also gets applies in the DAC implementation of the decoder.

[austinpop]

This confirms that Tidal is passing through the MQA stream when passthrough is enabled, and doing a *partial MQA unfold* when the passthrough is disabled. The partial unfold takes the MQA stream and does one level of unfolding, up to 88.2 or 96Khz. The resulting partially decoded MQA stream is then sent to the output device.

 

We know this because of the blue light, indicating a MQA stream. Otherwise if Tidal were doing a full decode to 96/24 PCM stream, the blue light would not come on.

 

Thanks for this clarification.

 

Has it been confirmed that the Tidal app tops out at 88.2/96 with non-MQA DACs?

 

Your explanation is certainly plausible, and also explains why MQA DACs still see the blue light even when the Tidal app is doing SW unfolding.

[jhwalker]

Thanks for this clarification.

 

Has it been confirmed that the Tidal app tops out at 88.2/96 with non-MQA DACs?

 

Your explanation is certainly plausible, and also explains why MQA DACs still see the blue light even when the Tidal app is doing SW unfolding.

 

I haven't seen it officially "confirmed", but every observation I've read says the same thing: software decoding is limited to no more than 96/24 (some play back at 88.2), hardware decoding is working to full bitrate (i.e., whatever rate the original file was in).

[Melvin]

Can you compare software decoded tracks to Bluesound decoded tracks and see if there is any significant differences?

 

For my own little MQA comparison test I chose to listen to just 1 song from a Yes album I’m quite familiar with and have in my own library at 24/192 resolution. From “Close to the Edge”, the song I used was “And You And I (I. Cord Of Life, II. Eclipse, III. The Preacher The Teacher, IV”. To really hear what MQA is capable of I disconnected my non-MQA DACs from the system and used only the Bluesound Node 2. Bluesound has of course enabled MQA in their product line and the native DAC in the Node 2 is capable to 24/192.

 

Frankly, the difference was easy to detect and immediate. The MQA file has more clarity and richness. Strings are gorgeous and simply sounded more real. Air, decay, separation of instruments all sounded good. I must admit to being surprised at this result and I will compare more in the coming days. After a while I will reconnect my non-MQA DACs and have a fresh listen. Makes me wonder where my preferences will lie in the end.

[austinpop]

I haven't seen it officially "confirmed", but every observation I've read says the same thing: software decoding is limited to no more than 96/24 (some play back at 88.2), hardware decoding is working to full bitrate (i.e., whatever rate the original file was in).

 

Well, I finally got a chance to dig into this for myself. My operational setup is based on an Aries Mini streamer feeding an Ayre Codex DAC, with details shown in my signature.

 

For this session, I brought in my MB Pro and, running the Tidal App in Exclusive mode, connected it to the DAC instead.

 

@jhwalker - you're absolutely right. I could not get any Master track to run at more than 24/96 sample rate. Some ran at 24/88.2.

 

This is a really intriguing glimpse at what one can only imagine would have been a contentious negotiation between Messrs. Stuart et al. and Tidal. This "partial unfolding" approach sounds like a compromise that allows Tidal to claim they provide high-resolution "Masters," while still enabling MQA to hold on to some of their jewels, that you only get with an MQA DAC.

 

I wanted to report on relative SQ listening impressions, but discovered I was comparing Apples and oranges. My Aries Mini chain is optimized to the hilt with isolation on Ethernet, USB, and DC. I tried to apply the same isolation on USB with my MB Pro, but for some reason I could not get the MBP to see the DAC with Intona/RUR in the chain. The best I could do was just RUR.

 

In this setup, my Aries Mini chain sounded WAY better than my MB Pro chain. So my cunning plan to compare "partially unfolded" MQA files on the MB Pro with the same regular 24/96 files in my library didn't pan out.

 

I listened to 3 songs I own at 24/96 that are also on Tidal Master:

• North Country II by Ola Gjeilo
  
• El Cuarto de Tula - Buena Vista Social Club, and
  
• Wond'ring Aloud - Jethro Tull, from the Steve Wilson CD/DVD release
  

In all cases, the native 24/96 versions on the Aries Mini chain sounded better, but as I said, the system differences made this a meaningless comparison.

 

The other comparison I did was on the Aries Mini chain, I compared my own 24/96 files with the MQA stream that the Aries Mini could pull from Tidal, which was the 24/48 folded (encoded) files. Here again, the 24/96 files were clearly better (as they should be). Equally significantly, the Tidal 24/48 files did sound slightly better than my own 16/44.1 versions. I couldn't find a way to get at the HiFi (16/44.1) version of these files on Tidal.

 

I did notice that that the Tidal files sounded different. I don't know what they're doing during the MQA encoding, but it's almost like it's being remastered. This was especially puzzling with the Tull song, as both are supposed to be the Steve Wilson remaster.

 

So what are my takeaways?

1. First of all, let's not lose sight that this is a huge step forward! Even without an MQA DAC, Tidal is now streaming 24/48 content. That is something to celebrate, since it's at zero added cost to the subscriber.
   
2. It will be interesting to see if MQA licenses the rights and code to do partial unfolding to vendors other than Tidal (like Aurender, Auralic, Sonore, SotM, etc). This would be the biggest win for those of us who do not plan to replace their non-MQA DACs at this time.
   
3. It will be interesting to evaluate Auralic's "private software based MQA decoder." Assuming it does not come under legal challenge from MQA.
   

[witchdoctor]

For my own little MQA comparison test I chose to listen to just 1 song from a Yes album I’m quite familiar with and have in my own library at 24/192 resolution. From “Close to the Edge”, the song I used was “And You And I (I. Cord Of Life, II. Eclipse, III. The Preacher The Teacher, IV”. To really hear what MQA is capable of I disconnected my non-MQA DACs from the system and used only the Bluesound Node 2. Bluesound has of course enabled MQA in their product line and the native DAC in the Node 2 is capable to 24/192.

 

Frankly, the difference was easy to detect and immediate. The MQA file has more clarity and richness. Strings are gorgeous and simply sounded more real. Air, decay, separation of instruments all sounded good. I must admit to being surprised at this result and I will compare more in the coming days. After a while I will reconnect my non-MQA DACs and have a fresh listen. Makes me wonder where my preferences will lie in the end.

 

Thanks very much for posting this comparison, very helpful to know. I really prefer the MQA versions of files I have listened to as well.

[crenca]

I did notice that that the Tidal files sounded different. I don't know what they're doing during the MQA encoding, but it's almost like it's being remastered. This was especially puzzling with the Tull song, as both are supposed to be the Steve Wilson remaster.

 

The more I listen the more I coming to this conclusion, that it has a "remaster" effect that has both pros and cons sonically and is an "improvement" sometimes and sometimes a step back (though as one poster said this might simply be acclamation to the previous masters/mixes). In any case it as often as not "subtle", particularly compared to the tracks I own (usually high res) even straight up, let alone through HQ Player up-sampled to my preference (usually DSD x 4 with minamal phase filtering - such as Mark Knopfler's "Sailing to Philly" album).

 

So, now that I have finally had a chance to evaluate a bit of MQA sonically (via Tidal software decoding and my "legacy" DAC(s)), the question to the Audiophile Press is this:

 

 

WHAT WERE YOU THINKING WITH THIS OUT OF CONTROL HYPE AND PROMOTION?!?!????

 

 

Sorry to yell, but they really and truly droped the ball on this one it appears. Besides John Darko and our own Chris, the rest of the "it sounds like" evaluations (at least the ones I have read) have all been no holds barred "it's the greatest SQ improvement since the invention of the wheel"...

[rickca]

Perhaps, but I have a strong suspicion that the actual $costs$ are not on Tidal at all - rather the lables are funding and pushing for the adoption of something (anything) that gets them some kind of DRM protection (not matter how trivial) of any music software of 16/44 or higher resolution. Tidal is simply looking for some exposure and perhaps a few more subscribers based on the hype of SQ improvement. In other words it is an "access change" and not a price hike in the near/middle future...

It's becoming clear to me that the primary motivation/business case for the labels to convert their libraries to MQA is to fend off piracy. I agree that eventually MQA will be the sole legal way to distribute high res, whether it's streaming or downloads. They realized an advance propaganda campaign engaging the audio press in enthusiastically promoting how much better it sounds was a prerequisite for acceptance by the audiophile community. So let's hope it actually does sound better.

 

I'm not saying the audio press is a bunch of dupes in a conspiracy. I trust the integrity of some reviewers. But it sure does sell a lot of newspapers. It gets the eyeballs you need to attract advertisers. So everybody in the industry wins. Let's hope it's also a win for the consumer.

 

What bugs me is that the software 'MQA core decoder' as implemented by A+ V3 just unfolds to 24/96 and doesn't give you the real raison d'etre sound quality improvement of MQA ... that requires the customized filter of an MQA-enabled device.

 

I just hope we can continue to stream non-MQA 16/44 from places like Qobuz for a long time to come.

[DGA]

Mixed results. Could someone listen to both versions of Chicago Transit Authority? I love this album, tape hiss and all. The hiss is gone but so are all the incredible dynamics of Does Anybody Really Know What time is? on the MQA version. Is it just me?

[austinpop]

Mixed results. Could someone listen to both versions of Chicago Transit Authority? I love this album, tape hiss and all. The hiss is gone but so are all the incredible dynamics of Does Anybody Really Know What time is? on the MQA version. Is it just me?

 

Just listened to this in both Hifi 16/44.1 and Master 24/48, since I don't have an MQA DAC.

 

Have to agree with you on this song. So weird to hear the hiss gone! It does feel there is a couple dB difference in level, with the Master softer. But it does feel like the dynamic range is compressed.

 

It really does seem to vary.

 

Listen to Riders on the Storm. The Master version sounds so much smoother and sweeter, but is that just the MQA encoding, or a different master?

[DGA]

 

 

Listen to Riders on the Storm. The Master version sounds so much smoother and sweeter, but is that just the MQA encoding, or a different master?

 

Interesting. Blacker background, more 3D and individual instruments are more localized. Sounds good. Am I hearing improved encode/decode or a remix though?

[abrxx]

Interesting. Blacker background, more 3D and individual instruments are more localized. Sounds good. Am I hearing improved encode/decode or a remix though?

 

Probably both!

[crenca]

Probably both!

 

I am leaning to more remix and that MQA black box "magic" might be doing very little:

 

http://www.computeraudiophile.com/f8-general-forum/mqa-friend-devil-sonic-evaluation-31182/index2.html#post621172

 

 

see my last post about Bob James "A New Cool" album...

[witchdoctor]

It looks like the generic profile software decoding isn't outputting the higher sample rates that have been reported as showing on certain MQA certified DACs. We also know that software decoding isn't doing per DAC "sweetening" for a lack of a better word that also gets applies in the DAC implementation of the decoder.

 

 

So what? I'll bet you couldn't even tell the difference of MQA sample rates in an A/B test. I would not spend even $1 for a dac just to hear a sample rate higher than 96/24

[witchdoctor]

General MQA observation- the tracks I have listened to appear to have better micro and macro dynamics. I think the choice of an amp you use is critical. The amp needs to be able to have the reserves necessary to mirror those dynamics and support them in their glorious detail.

Has anyone else noticed this?

[Samuel T Cogley]

I'm hoping this is the best thread for this observation. The MQA version of Jackson Browne's Running On Empty appears to be a remix. It sounds great. Soundstage is W-I-D-E compared to what I've heard since '77. But it's radically different including lots of added stuff that was presumably trimmed for the '77 master. Based on the track duration times, even what's on HDTracks is the "original" master, not the remix.

 

EDIT: Ok, I've found several other remixes. It's going to be impossible to compare SQ to the originals. I wonder if master tapes got lost/destroyed and Warner had to remix the session tapes? MQA may change history, but not in the way we were expecting...

[ajay556]

How do i use off-line (no internet while traveling) MQA tracks on desktop application?

How can we tell from file if its MQA format? Can i search for songs for MQA format only?

[arcman]

One thing to consider (maybe far fetched). Mastered for itunes requires a high rez copy at the specs that Apple requests. Apple down samples to their output format. I found some MFiT to sound a little different even among same mastering (Technically, the MFiT version could be considered a separate mastering). Many Labels may just submit the newer iTunes Mastered digital file (which is high Rez) with additional MQA requirements. Thus, same master versions could sound completely different between normal and high rez.

[blownsi]

After several days of listening to my redbook rips, Tidal redbook, and Tidal Masters with more than one "legacy" DAC. I've came to the following conclusion: TM/MQA at it's best seems to be a redbook equivalent new master not a "free" high resolution version of what I already own. I now firmly believe that MQA is a means to generate profit. Profit in the form of selling decoders in new dacs and in the purchase or streaming of new versions of music.

With all that said some of the new masters are wonderful and are absolutely worth the cost, others are just different than redbook, not necessarily better. If Tidal continues to offer this at Hifi prices and offer both versions then this is a wonderful thing. If they force us to accept whichever version they think is better or double the rates to access Masters then I'm out.

[Samuel T Cogley]

I posted some MQA hardware decoding listening impressions here, but this little snippet from a post at the Hoffman forum completely jibes with my experience and says it much better than I did:

 

Fleetwood Mac - Rumors: I've got the original LP and the (I think) 3rd CD re-release. The MQA version of the album heard through a Bluesound Node 2 feeding a Naim Supernait 2 driving a pair of Kudos Cardea Super 10 standmounts sounds airy and hollow in spots; as though there's too much space between the musician/singers and their mics. Same with two different MQA releases of Coltrane's Giant Steps. Specific examples - the MQA'd "Never Going Back Again" sounds cleaner and slightly unveiled compared to the CD re-release (the so-called Super Deluxe release), which is an improvement. On the same album, the MQA'd "Songbird" sounds cleaner and crisper in the same way as tilting up the treble tone control, the problem being that the lower harmonics in the McVie's voice are completely missing compared to the LP and the Super Deluxe CD release, which makes the MQA version notably worse.

 

I called it a level of detail that eventually causes a bit of listening fatigue. YMMV of course.

[crenca]

I posted some MQA hardware decoding listening impressions here, but this little snippet from a post at the Hoffman forum completely jibes with my experience and says it much better than I did:

 

Fleetwood Mac - Rumors: I've got the original LP and the (I think) 3rd CD re-release. The MQA version of the album heard through a Bluesound Node 2 feeding a Naim Supernait 2 driving a pair of Kudos Cardea Super 10 standmounts sounds airy and hollow in spots; as though there's too much space between the musician/singers and their mics. Same with two different MQA releases of Coltrane's Giant Steps. Specific examples - the MQA'd "Never Going Back Again" sounds cleaner and slightly unveiled compared to the CD re-release (the so-called Super Deluxe release), which is an improvement. On the same album, the MQA'd "Songbird" sounds cleaner and crisper in the same way as tilting up the treble tone control, the problem being that the lower harmonics in the McVie's voice are completely missing compared to the LP and the Super Deluxe CD release, which makes the MQA version notably worse.

 

I called it a level of detail that eventually causes a bit of listening fatigue. YMMV of course.

 

Well, this is getting pretty consistent. We are getting many reports here and elsewhere that all correlate. I called the stereo separation "more" (as in, artificial) and the HF stuff "digititus", etc. Besides the guys who are simply reporting "MQA rocks!" without actually describing what they are hearing and why (mid-fiers?) the picture is coming together.

 

It flat out contradicts the "birth of a new world", non-qualified or nuanced YYUUUUGGGGEEEE improvement from the so called Audiophile press. Perhaps some of this was expected but have any of you ever seen such a disconnect between the press hype/promotion and the actual product (calling the MQA standard a "product" for just a moment)?

[TubeDriver]

I certainly hear the differences as mixed. Some Master tracks sound better than their normal 16/44 version, some sound worse, all sound DIFFERENT.

 

The more I listen the more I coming to this conclusion, that it has a "remaster" effect that has both pros and cons sonically and is an "improvement" sometimes and sometimes a step back (though as one poster said this might simply be acclamation to the previous masters/mixes). In any case it as often as not "subtle", particularly compared to the tracks I own (usually high res) even straight up, let alone through HQ Player up-sampled to my preference (usually DSD x 4 with minamal phase filtering - such as Mark Knopfler's "Sailing to Philly" album).

 

So, now that I have finally had a chance to evaluate a bit of MQA sonically (via Tidal software decoding and my "legacy" DAC(s)), the question to the Audiophile Press is this:

 

 

WHAT WERE YOU THINKING WITH THIS OUT OF CONTROL HYPE AND PROMOTION?!?!????

 

 

Sorry to yell, but they really and truly droped the ball on this one it appears. Besides John Darko and our own Chris, the rest of the "it sounds like" evaluations (at least the ones I have read) have all been no holds barred "it's the greatest SQ improvement since the invention of the wheel"...

[wgb113]

Mac OSX users - how are you getting varying bit rates and resolutions? I tried a bunch of different settings within the Tidal app (latest version) and Audio MIDI but ultimately I always get whatever Audio MIDI is set to showing up on my DAC.

[jhwalker]

Mac OSX users - how are you getting varying bit rates and resolutions? I tried a bunch of different settings within the Tidal app (latest version) and Audio MIDI but ultimately I always get whatever Audio MIDI is set to showing up on my DAC.

 

If you set the Tidal app to "Use Exclusive Mode", it will change the output to whatever the correct sampling rate should be.

[Samuel T Cogley]

Well, this is getting pretty consistent. We are getting many reports here and elsewhere that all correlate. I called the stereo separation "more" (as in, artificial) and the HF stuff "digititus", etc. Besides the guys who are simply reporting "MQA rocks!" without actually describing what they are hearing and why (mid-fiers?) the picture is coming together.

 

It flat out contradicts the "birth of a new world", non-qualified or nuanced YYUUUUGGGGEEEE improvement from the so called Audiophile press. Perhaps some of this was expected but have any of you ever seen such a disconnect between the press hype/promotion and the actual product (calling the MQA standard a "product" for just a moment)?

 

When Pono switched from Meridian to Ayre for the DAC tech (BTW, the Pono has a legacy blue light, just sayin'), most of the early raves about Pono's sound quality were based on a desktop prototype that was allegedly equivalent to what the production unit would sound like.

 

I bring this up because I suspect all the raves about the awesomeness of MQA sound quality probably came from listening to similar prototype gear. It's likely that a $2000-ish MQA-enabled DAC will smoke the little Explorer2 (or the Bluesound NODE 2) in terms of sound quality. To be fair, I think we have to wait to hear what MQA sounds like on a better DAC.

 

After listening to more MQA stuff through the Explorer2, I'm back to the Yggy, Gumby, and iDSD Micro/HQPlayer playing Redbook. No listening fatigue!