• 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!

MQA Deep Dive - I published music on tidal to test MQA

Status
Not open for further replies.

RichB

Major Contributor
Forum Donor
Joined
May 24, 2019
Messages
1,946
Likes
2,611
Location
Massachusetts
Page 1 of stereophile this month.
View attachment 134017

The wording is cringe worthy. How does this sound?
Mr. Austin, it is said, to be a flak for MQA. Let's address the lack of professionalism of this article because it is easiest to dismiss.
...
Mr. Austin's article is a missed opportunity.

Please don't attribute the above statements to me, I am only posting what was said. ;)

Clearly, MQA is all hands on deck. It is interest to identify the hands, though.

- Rich
 
Last edited:

pjug

Major Contributor
Forum Donor
Joined
Feb 2, 2019
Messages
1,775
Likes
1,561
I read that useless explanation -- multiple times actually. There is nothing clear about it:

"We actually can show this effect quite prominently when looking/listening to MQA-encoded music that began life as 44.1 or 48kHz. Very obvious examples are pop recordings such as this Bruno Mars album below originally of 44.1kHz sampling rate, fed into the MQA encoder and then unfolded to 88.2kHz within the Tidal software. "

How the heck does he know what master the MQA file came from? And if he does, why doesn't he know if it was 44.1 or 48 kHz?
The .csv list of mqa albums gas a "mixed" column that shows sample rate. I assume this means the sample rate of the original. I don't know how they know this. Anyway I am not sure why you seem to think there is a good chance the original is not 44Khz when all indications are that it is 44. Your spectrum is different but it does not show any ultrasonics from music either.
 

awdeeoh

Member
Joined
Feb 24, 2020
Messages
68
Likes
28
How? Do you mean with the open source downloader? It appears that this tool can only download one album at a time without human intervention. After each album download one must then select another album for download. With 60M tracks, there are conceivably 6M albums, requiring as many human interventions. Do you still think it can be done in less than a month?

what's the use of it being open source, you can modify it if you know how to code, pretty trivial to bot nowadays.
 

awdeeoh

Member
Joined
Feb 24, 2020
Messages
68
Likes
28
The .csv list of mqa albums gas a "mixed" column that shows sample rate. I assume this means the sample rate of the original. I don't know how they know this. Anyway I am not sure why you seem to think there is a good chance the original is not 44Khz when all indications are that it is 44. Your spectrum is different but it does not show any ultrasonics from music either.

maybe a new encode.
 

RichB

Major Contributor
Forum Donor
Joined
May 24, 2019
Messages
1,946
Likes
2,611
Location
Massachusetts
I read that useless explanation -- multiple times actually. There is nothing clear about it:

"We actually can show this effect quite prominently when looking/listening to MQA-encoded music that began life as 44.1 or 48kHz. Very obvious examples are pop recordings such as this Bruno Mars album below originally of 44.1kHz sampling rate, fed into the MQA encoder and then unfolded to 88.2kHz within the Tidal software. "

How the heck does he know what master the MQA file came from? And if he does, why doesn't he know if it was 44.1 or 48 kHz?

You need to know what the labels gave them to encode. I have seen tons of horror stories which I have documented. You don't need MQA to screw things up -- the labels do that routinely.

Regardless, my point was that this is just a bunch of guessing games which contrary to what Archimago said, require fair bit of expertise just to know the vocabulary let alone what is really going on. The notion then that he can be Joe-anybody online to do this analysis is a farce.

The guessing game, I presume, is the exact cause of the noise. For example, the noise found in your RMC-1 review that was guessed to hardware related. In the end, was actually an improperly loaded register that was corrected by software. The inferred cause was not correct but the measurements were valid.

Or put another way, if I find a turd on the carpet it is reasonable to blame my dog without understanding his biology or even the terms for the biology involved in the act. :p

- Rich
 
Last edited:

awdeeoh

Member
Joined
Feb 24, 2020
Messages
68
Likes
28
I think you can still get a copy of the Bruno Mars track that was tested by Archimago.

Internet is your friend, or maybe ask a Russian friend. :)
 

RichB

Major Contributor
Forum Donor
Joined
May 24, 2019
Messages
1,946
Likes
2,611
Location
Massachusetts
Some manufacturers are reporting the fees for MQA addition are substantial.
Roon and anyone else who markets an MQA software decoder ("first unfold") in their software apparently plays a licensing fee per MQA track played back.

This would be interesting data to confirm.

If Roon is paying per track, that could be a problem for the lifetime subscription.
It also might explain the price hike from $499 to $699. Monthly might be absorbable by Roon but lifetime could be the gift that keeps on taking.

The argument that fees do not increase costs and impact pricing is absurd, but those not skilled in economics or its terms can be easily fooled. :p

- Rich
 

RichB

Major Contributor
Forum Donor
Joined
May 24, 2019
Messages
1,946
Likes
2,611
Location
Massachusetts
C'mon. This is bad faith marketing BS. The consumer is paying for it. Just because your subscription rate wasn't raised, that doesn't mean there wasn't an opportunity cost, as I'm sure you understand.



I disagree. But I understand how that perspective doesn't fit the MQA proxy talking points on how MQA is free to end-users. lol

The Roon lifetime subscription price was raised from $499 to $699.

- Rich
 

Raindog123

Major Contributor
Joined
Oct 23, 2020
Messages
1,599
Likes
3,554
Location
Melbourne, FL, USA
You also had bad luck at picking a streaming ecosystem that picked a bad fight with Tidal.

Yep. With a caveat of that - Auralic’s - ecosystem being just fine when I picked it five years ago. Tidal was barely taking off with the “CD quality” streaming, and no one even heard of MQA. Things went south (for some Auralic users) only after MQA got on its feet at Tidal... And that’s exactly the point - things might look fine… for a while.
 

Werner

Active Member
Joined
Mar 2, 2016
Messages
109
Likes
135
Location
Europe
When I look at the level of the aliased image "allowed through" by the MQA filter I see peaks at -96dB and when averaged, well below -100dB.

That's not the correct way of reading an FFT. The level of the ultrasonic images has to be integrated over the bandwidth they occupy, some 24kHz in this case. The result of that sum is what matters, and it will be substantial when compared to the payload signal, and easily within your system's capabilities, loudspeakers excepted.
 
Last edited:

pjug

Major Contributor
Forum Donor
Joined
Feb 2, 2019
Messages
1,775
Likes
1,561
We will just have to wait for confirmation.

Again, less than -100dB.

[edit: I have been making it sound like the decoder does this upsampling. Seems to make more sense to me that they would add the mirrored ultrasonics using the same technique but prior to or part of the encoding process. Then the decoder can treat these like other MQA files]

I'm pretty sure you are clear on this, but just so it is clear, this has nothing to do with the 2nd unfolding. The point made by @Werner is that when MQA files are made from 24/44 originals, the decoder (1st unfold) automatically upsamples with a weak filter to produce that mirroring. This gives the appearance of ultrasonics but since there were none to begin with I would call them fake.

Anyone can do the same. Here is what I get when I upsample a 44Khz version of the track using no filter (so worse than what MQA does):
1622980133964.png
 
Last edited:

raistlin65

Major Contributor
Forum Donor
Joined
Nov 13, 2019
Messages
2,279
Likes
3,421
Location
Grand Rapids, MI
A fundamental misunderstanding of basic economics. You are paying, whether it is visible or not.
And the upcoming Tidal pricing tiers includes a premium for MQA.
This would be interesting data to confirm.

If Roon is paying per track, that could be a problem for the lifetime subscription.
It also might explain the price hike from $499 to $699. Monthly might be absorbable by Roon but lifetime could be the gift that keeps on taking.

The argument that fees do not increase costs and impact pricing is absurd, but those not skilled in economics or its terms can be easily fooled. :p

- Rich

According to public financial records, MQA has been losing money. See page 10

https://find-and-update.company-inf.../MzI4NzY4NTAzMmFkaXF6a2N4/document?format=pdf

So unless MQA fails and goes bankrupt, the people who have taken on that loss will want to recover it from licensing fees, which will be passed on to users somehow.

.
 

David Harper

Senior Member
Joined
Jul 7, 2019
Messages
359
Likes
434
I'm new to this discussion and I never heard MQA and don't know a lot about it. So correct me if I'm wrong here: my understanding of it is that it's principal benefit is not (in absolute terms) sound quality but rather good SQ in a smaller file size. Is this essentially it? So MQA music should sound
equal in SQ to, for example, uncompressed or lossless redbook CD but in a smaller sized package? So if I don't care about file size then there is no benefit in MQA for me? I chose redbook CD as a reference because after listening to a lot of hi-res I came to the conclusion that it doesn't sound different or better to me. I save my music in 16/44 WAV files and the file size has never been an issue for me.
When streaming music I'm mostly not concerned with SQ I'm just getting into the music. The phrase "critical listening" always seemed odd to me.
It seems like an oxymoron. I listen to music, not equipment. And also it seems that the main criticism of MQA is that it requires a total investment, end to end, in order to realize the benefit? So this results in some seeing it as just another scheme to milk the rubes? And make us buy our music all over again? As I said, I don't know much about this so forgive me if my ideas are completely clueless.
 

pjug

Major Contributor
Forum Donor
Joined
Feb 2, 2019
Messages
1,775
Likes
1,561
So MQA music should sound
equal in SQ to, for example, uncompressed or lossless redbook CD but in a smaller sized package?
No, they say the sound quality is better than the hi-res original but in a redbook size package. Compressed MQA is about the same as uncompressed redbook. Compressed (flac) redbook is a good bit smaller than compressed (flac) MQA. You can get an idea of size differences on the 2L test bench table:
http://www.2l.no/hires/
 

lucretius

Member
Joined
Aug 18, 2018
Messages
41
Likes
59
Good points.

If one forgoes the rendering step (which some object to anyway), unfolded MQA is just LPCM.

Not necessarily. E.g. when you tell Roon that your DAC is a "Renderer only", Roon leaves some code in place in the control stream for (1) authentication; (2) for "upsampling" and mQa filter selection; and (3) indication of the source sample rate. OTH, If you tell Roon that your DAC has "No MQA support", it's not clear what Roon does with the control bit -- it's likely that Roon just pads the bit in that case. So the unfolded MQA will never be like a real 24 bit LPCM stream.
 

Zensō

Major Contributor
Joined
Mar 11, 2020
Messages
2,740
Likes
6,740
Location
California
I'm new to this discussion and I never heard MQA and don't know a lot about it. So correct me if I'm wrong here: my understanding of it is that it's principal benefit is not (in absolute terms) sound quality but rather good SQ in a smaller file size. Is this essentially it? So MQA music should sound
equal in SQ to, for example, uncompressed or lossless redbook CD but in a smaller sized package? So if I don't care about file size then there is no benefit in MQA for me? I chose redbook CD as a reference because after listening to a lot of hi-res I came to the conclusion that it doesn't sound different or better to me. I save my music in 16/44 WAV files and the file size has never been an issue for me.
When streaming music I'm mostly not concerned with SQ I'm just getting into the music. The phrase "critical listening" always seemed odd to me.
It seems like an oxymoron. I listen to music, not equipment. And also it seems that the main criticism of MQA is that it requires a total investment, end to end, in order to realize the benefit? So this results in some seeing it as just another scheme to milk the rubes? And make us buy our music all over again? As I said, I don't know much about this so forgive me if my ideas are completely clueless.
It seems as if you’re primarily concerned with how MQA sounds in normal listening (as opposed to critical listening). If that’s the case, it’s unlikely you’ll hear any difference. So yes, viewed in that context, it’s just another scheme to extract more revenue from unsuspecting audiophiles.
 
Last edited:

lucretius

Member
Joined
Aug 18, 2018
Messages
41
Likes
59
what's the use of it being open source, you can modify it if you know how to code, pretty trivial to bot nowadays.

You do realize that you must login to an active Tidal account to effect this download? IOW, you must leave your footprint.
 

raistlin65

Major Contributor
Forum Donor
Joined
Nov 13, 2019
Messages
2,279
Likes
3,421
Location
Grand Rapids, MI
You do realize that you must login to an active Tidal account to effect this download? IOW, you must leave your footprint.

Yep. There's probably some kind of automatic flag for review if someone reaches a certain threshold of heavy downloading. And probably something buried in the terms of service agreement that allows them to cancel one's subscription for excessive downloads.
 
Status
Not open for further replies.
Top Bottom