• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Question - Tidal MQA quality without an MQA-capable DAC

TylerT

Member
Joined
May 17, 2024
Messages
13
Likes
21
Location
Alberta, Canada
Hello,

Recently, I upgraded from a Fiio BTR15 with an MQA renderer to a Fiio K7 without the MQA renderer. My daily driver is a Hifiman Ananda Nano, and the 2W power output from the K7 seems to drive the planar magnetic headphone better than the BTR15 (it could be a placebo effect, but it does seem to have more bass, fullness, or something). From this thread/post: https://goldensound.audio/2021/11/29/tidal-hifi-is-not-lossless/, any Tidal MQA track will have lossy 16-bit 44.1 kHz FLAC quality instead of being lossless like tracks without MQA; that is, any MQA tracks will be lossy no matter the format. Therefore, it’s better to use a DAC with an MQA renderer instead of switching to “high” quality 16-bit 44.1 kHz FLAC since it’s even more lossy than MQA. From what I can understand, MQA unfolding (lossless) and rendering (lossy) by an MQA DAC will be indifferentiable from hi-res FLAC. However, without the rendering, the MQA file will still be compressed and have significant noise, which is worse than lossless 16-bit 44.1 kHz FLAC. My K7 shows that the bitrate is 88.2 kHz when I play MQA files, but this is missing the render process.

My main query is this: With the transition to the Fiio K7, how much will the quality of the MQA tracks be compromised? If I'm correct in my understanding, Tidal will decode the track, but the K7 will not render it, resulting in a lossy and compressed version of MQA. I find myself resorting to Tidal's “high” quality to ensure that Tidal is streaming a 16-bit 44.1 kHz FLAC file instead of MQA, but it seems even this is not truly lossless on Tidal. Additionally, I find it a huge pain to keep switching back and forth between “Max” and “High” quality all the time, depending on the track being played. Is missing the render process perceivable by human hearing? Is the quality noticeably audible, or is it in such a range that my setup (Fiio K7 to Hifiman Ananda Nano) is not even resolving enough to produce? I use Tidal for my music streaming because of the library, user interface, and music recommendations. I’ve considered switching to Qobuz, although the latter will be more expensive (I have a Tidal student subscription) and have a smaller library (a primary deal breaker for me). Due to these reasons, I find myself sticking with Tidal until a better alternative pops up.

Sidenote - personally, I dislike MQA because it’s not open source (which inevitably adds cost to the end consumer), not transparent, and most importantly, I have to buy a specific MQA-enabled DAC (and even then, the DAC keeps bugging out on MQA tracks for whatever reason). While I have my gripes with MQA, as long as the end quality is indistinguishable from hi-res FLAC, I don’t care what the format is (be it Dolby, CD WAV, or FLAC). I also can’t distinguish the difference between hi-res and lossless, so perhaps my hearing isn’t sensitive enough, or my system is not sufficiently resolving; regardless of the reason, I still prefer to have hi-res playback, similar to how knowing I have a faster processor makes me happier even if I can’t notice/utilize the processing speed.

Thank you for the help in clarifying,
Tyler
 
Tidal will decode the track, but the K7 will not render it, resulting in a lossy and compressed version of MQA.
No, that is not how it works.
The first unfolding, convert 48 to 96 and restore the audio in the 24-48 kHz band can be done in software.
The last unfolding, oversampling and/or using the MQA prescribed filters is only allowed in hardware.

As MQA stores the 24-48 kHz audio below bit 17 (the audio origami) it is lossy by design and no soft- or hardware can do anything about that.
Tidal is replacing MQA with FLAC.
Bit more detail: https://www.thewelltemperedcomputer.com/KB/MQA.htm
 
Thank you for the information and for linking such useful resources. After reading up on MQA, I see now that the encoding is done through a lossy origami process, and the dynamic range cannot be restored to the original quality.

After some consideration, I’ve decided to switch to Apple Music (AM), which provides a student discount and has an extensive lossless library. The only bit-perfect playback AM supports is through an external DAC/Amp connected to an iOS device. It seems that the AM app on Android, Mac OS, and Windows cannot bypass OS resampling. My current setup is iPhone 15 Pro Max – Fiio K7 – Hart 4.4 TRRS Balanced Cable – Hifiman Ananda Nano. Many online sources claim that IOS external DAC playback is bit-perfect and not resampled.

Sidenote – Some FLAC songs on Tidal and AM seem to be different masters provided by the same artist. I’ve read online that the artists simply provide the same masters to all online streaming services (ignoring MQA for Tidal) and that any/all lossless streams should sound identical regardless of the service. However, I’ve noticed that certain FLAC-only songs on Tidal are of lower spec quality than the same song on AM (and vice versa). For example, Natsutourou by Hachi on Tidal has a max quality of 16-bit 44.1 kHz FLAC; on AM, the highest quality available is 24-bit 96 kHz FLAC. My guess is that Hachi (vTuber artist) provided Tidal with a CD quality master first, then worked with Apple to provide a hi-res remaster of the same song. Why would the artist intentionally provide a lower-resolution FLAC file to Tidal if she already had the original hi-res FLAC? Just something interesting I’ve noticed. While I can’t tell the difference between the two tracks, I do prefer streaming on AM just knowing it’s most likely a higher quality master.
 
Thank you for the information and for linking such useful resources. After reading up on MQA, I see now that the encoding is done through a lossy origami process, and the dynamic range cannot be restored to the original quality.

After some consideration, I’ve decided to switch to Apple Music (AM), which provides a student discount and has an extensive lossless library. The only bit-perfect playback AM supports is through an external DAC/Amp connected to an iOS device. It seems that the AM app on Android, Mac OS, and Windows cannot bypass OS resampling. My current setup is iPhone 15 Pro Max – Fiio K7 – Hart 4.4 TRRS Balanced Cable – Hifiman Ananda Nano. Many online sources claim that IOS external DAC playback is bit-perfect and not resampled.

Sidenote – Some FLAC songs on Tidal and AM seem to be different masters provided by the same artist. I’ve read online that the artists simply provide the same masters to all online streaming services (ignoring MQA for Tidal) and that any/all lossless streams should sound identical regardless of the service. However, I’ve noticed that certain FLAC-only songs on Tidal are of lower spec quality than the same song on AM (and vice versa). For example, Natsutourou by Hachi on Tidal has a max quality of 16-bit 44.1 kHz FLAC; on AM, the highest quality available is 24-bit 96 kHz FLAC. My guess is that Hachi (vTuber artist) provided Tidal with a CD quality master first, then worked with Apple to provide a hi-res remaster of the same song. Why would the artist intentionally provide a lower-resolution FLAC file to Tidal if she already had the original hi-res FLAC? Just something interesting I’ve noticed. While I can’t tell the difference between the two tracks, I do prefer streaming on AM just knowing it’s most likely a higher quality master.

Apple Music can play in lossless and hi-res on some Android DAPs with SRC bypass (based on the sample rate displayed on my Hiby DAP). You do lose access to the Atmos mixes unless your device has Atmos built-in, which no DAP has AFAIK.
 
I’ve read online that the artists simply provide the same masters to all online streaming services
I think that might be true sometimes, but no-one knows how often it happens. There are many anecdotal examples where this is not the case. In general the safest approach is to assume you know nothing about which master you are fed by a streamer. Consequently you can never compare a song from two different streamers and assume it's the same master or has been through the same process.
 
Apple Music can play in lossless and hi-res on some Android DAPs with SRC bypass (based on the sample rate displayed on my Hiby DAP). You do lose access to the Atmos mixes unless your device has Atmos built-in, which no DAP has AFAIK.
Do you know if the SRC bypass also bypasses the Android internal mixer dither? After further research, it seems that even when the sample rate remains unchanged, as long as the internal mixer is not disabled, it dithers the data before sending it to your external DAC. As long as the mixer is active, even if it does not ‘mix’ any inputs or convert your sample rate to match OS, it will still convert to float, mix (add nothing), dither, and reconvert to an integer.

The iPhone automatically does this even if the sample rate is not converted to match iOS; I suppose this is the closest to bit-perfect we can get at the moment (on mobile phones).

I think that might be true sometimes, but no-one knows how often it happens. There are many anecdotal examples where this is not the case. In general the safest approach is to assume you know nothing about which master you are fed by a streamer. Consequently you can never compare a song from two different streamers and assume it's the same master or has been through the same process.
True! Although I wonder how often songs get remastered for streaming purposes, it doesn't seem like a valid reason to remaster tracks for different streaming services. It would make more sense that the artists simply provide the same master until the day comes when it needs to be remastered.
 
I wonder how often songs get remastered for streaming purposes, it doesn't seem like a valid reason to remaster tracks for different streaming services
Unless something has changed recently, tracks are not specifically remastered for streaming.

Take a song from 1977. A stereo master is created from mixing down the multitrack. A different version is created specifically to cut LPs to cope with bass limitations etc. These 2 masters go on the shelf after copies are made from them. Now imagine a massive temperature controlled warehouse full of thousands of tapes.

In 1985 they are recovered and used to make a CD (but they use the LP master by mistake). In 1991, the correct master is digitised, but this time with better quality ADC and dither and the CD re-released. There's a 2002 version released with tons of compression. Then a complete remix from the original multitrack is made in 2015. hi-res versions (96k and 192k) are available for download. This release comes with different studio takes of the song "previously unreleased". In 2020 an Atmos version is released along with a new 180g LP version (needing a special master to handle the mono-ing of bass).

I have some pieces of music in 6 or more versions.

Spotify gets a file to stream in 2015, Tidal gets a file in 2016, Amazon in 2020 etc. Which streamer gets which version? If a new version is released do they replace it? It's basically a circle of confusion!
 
In my very limited testing most streaming services seem to carry the same version(s) of albums as each other. Im sure very easy to disprove and all I did was look at some very famous examples. My thinking is the record companies/ distributors just send them all swathes of the same files at the same time. Sure a few smaller concerns/labels/ artists upload piecemeal too.

Its possible given what MQA purported to be from the recording/ distribution end that an MQA version was a different master to any of the other versions out there.
 
I think it also depends on the provenance of the music. For something recorded in 2023 all streamers are likely to be the same. For something from the 70s it's hard to agree which version is the "correct" one!

Thread 'Review of the Michael Jackson's album Thriller between 21 versions including Vinyls, CDs, cassettes, SACD, Streaming, MOFI releases, 360RA' https://audiosciencereview.com/foru...tes-sacd-streaming-mofi-releases-360ra.39616/
 
I think it also depends on the provenance of the music. For something recorded in 2023 all streamers are likely to be the same. For something from the 70s it's hard to agree which version is the "correct" one!

Thread 'Review of the Michael Jackson's album Thriller between 21 versions including Vinyls, CDs, cassettes, SACD, Streaming, MOFI releases, 360RA' https://audiosciencereview.com/foru...tes-sacd-streaming-mofi-releases-360ra.39616/
Well reminded. Was on the back of that I had a look at the different streaming options. The same https://www.audiosciencereview.com/...eaming-mofi-releases-360ra.39616/post-1667228

There may be 5 million versions of that album but the streamers all have the same couple.
 
Do you know if the SRC bypass also bypasses the Android internal mixer dither? After further research, it seems that even when the sample rate remains unchanged, as long as the internal mixer is not disabled, it dithers the data before sending it to your external DAC. As long as the mixer is active, even if it does not ‘mix’ any inputs or convert your sample rate to match OS, it will still convert to float, mix (add nothing), dither, and reconvert to an integer.

The iPhone automatically does this even if the sample rate is not converted to match iOS; I suppose this is the closest to bit-perfect we can get at the moment (on mobile phones).


True! Although I wonder how often songs get remastered for streaming purposes, it doesn't seem like a valid reason to remaster tracks for different streaming services. It would make more sense that the artists simply provide the same master until the day comes when it needs to be remastered.

I have no idea about the dither on Android or iOS. My comment was based solely on the indicated sample rate in the status bar which matched the source with lossless or hi-res tracks.
 
Spotify gets a file to stream in 2015, Tidal gets a file in 2016, Amazon in 2020 etc. Which streamer gets which version? If a new version is released do they replace it? It's basically a circle of confusion!
My thoughts exactly, hahaha.

Its possible given what MQA purported to be from the recording/ distribution end that an MQA version was a different master to any of the other versions out there.
This is my new theory of why some people believe MQA sounds better :)
 
I am still looking for the “best” solution to use EQ to tune my Tidal hiRes tracks. So far I have tried:

1. Roon EQ on my PC: very good parametric EQ, but I just can’t justify the cost of Roon
2. HiBY Music on my Android DAP. It streams my Tidal tunes in 44.1, and the EQ works, but not as good as Roon.

Any other suggestions?
 
I am still looking for the “best” solution to use EQ to tune my Tidal hiRes tracks. So far I have tried:

1. Roon EQ on my PC: very good parametric EQ, but I just can’t justify the cost of Roon
2. HiBY Music on my Android DAP. It streams my Tidal tunes in 44.1, and the EQ works, but not as good as Roon.

Any other suggestions?
USB audio player Pro (uapp) on android
 
It's a shame that no DAPs actually support Atmos...
How would you enjoy the "benefits" of Atmos with a DAP... have you got some sort of crown with a constellation of speakers...? ;-)
 
isnt dolby atmos a free feature in any modern even half way decent phone?

ALSO "Tidal drops MQA for FLAC" is about the most damning thing ever

we're going back to a free lossless carrier over a proprietry codec that requires significant expenditure on hardware for Tidal and the end user for very little benefit

i would also posit there's significant technical support issues on both end especially for the consumer that do not exist with current free codecs

thanks meridian and lenbrooke for this grand experiment
 
It's supported on Apple devices and many Android phones, just no DAPs. And yes, it works on headphones similar to 360RA.
 
Back
Top Bottom