• 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.

Moby is MQA.
The problem is that the 'hifi' version is simply MQA but with MQA flagging removed. meaning any tools to check for MQA flagging will say it is not MQA. But actually it is.

Here is the deltawave comparison between the MQA version and Hifi version of Moby - Porcelain:
1619023025146.png


The audio content of the file is 100% bitperfect identical.
The only difference is the "Master" track has MQA flagging. That's it
But the hifi version is still the MQA release just disguised as regular FLAC.

The same happened with my tracks, the "HiFi" version had no MQA flagging at all and nothing would recognise it as MQA. But it was 100% bitperfect to the MQA release and was not the same as my master.

@Glasvegas @snowsurfer Tidal is serving MQA and falsely describing it as "Lossless CD Quality"
 
For my own edification, in layman's terms, does the bolded description of a lossy encoder mean that it is recording change from previous sample, not absolute value of the current sample?

So as a simple conceptual example:

10, 8, 4, 3, 6, 7 would be encoded as: 10, -2, -4, -1, +3, +1?
That's still lossless. Lossy would be encoding it as "start high, drop to a bit less than half, then go almost all the way back up."
 
With the mass replacement of redbook on Tidal, surely a lot of these MQA are made from 44.1K or 48K originals. So then the spectra have a gap between the music and the fake ultrasonics, like this (from https://audiophilestyle.com/ca/reviews/mqa-a-review-of-controversies-concerns-and-cautions-r701/)
View attachment 125479
The gap here is unimportant. It merely serves to emphasise that everything above it is a rolled-off mirror image of the spectrum below. The reason for the gap is that the original recording was filtered at ~21 kHz, nothing to do with MQA.
 
Moby is MQA.
The problem is that the 'hifi' version is simply MQA but with MQA flagging removed. meaning any tools to check for MQA flagging will say it is not MQA. But actually it is.

Here is the deltawave comparison between the MQA version and Hifi version of Moby - Porcelain:
View attachment 125480

The audio content of the file is 100% bitperfect identical.
The only difference is the "Master" track has MQA flagging. That's it
But the hifi version is still the MQA release just disguised as regular FLAC.

The same happened with my tracks, the "HiFi" version had no MQA flagging at all and nothing would recognise it as MQA. But it was 100% bitperfect to the MQA release and was not the same as my master.

@Glasvegas @snowsurfer Tidal is serving MQA and calling it lossless

Thanks for that, as @mansr pointed out the tool I am using may not be identifying all MQA files, but it's the only thing I have right now. :)
 
The gap here is unimportant. It merely serves to emphasise that everything above it is a rolled-off mirror image of the spectrum below. The reason for the gap is that the original recording was filtered at ~21 kHz, nothing to do with MQA.
Yes but it makes very easy to identify MQA files that are in no way hi resolution.
 
Not Found
The requested URL /tmpFiles/elib/20210417/20456.pdf was not found on this server.
OK, then try to google: "A Hierarchical Approach for Audio Capture, Archive, and Distribution", or reach for it directly in AES. This paper can be downloaded for free from AES.
 
The same happened with my tracks, the "HiFi" version had no MQA flagging at all and nothing would recognise it as MQA.
So where does MQA flagging take place if it's not in the PCM data? I thought it was precisely the LSBs that trigger MQA decoding.
 
They did, it is called FLAC and even Apple has conceded this.
I don’t think that FLAC can encapsulate a high resolution audio stream that’s backward compatible with a standard one (CD). That’s what I read into his comment.

PS: unless it’s MQA, that is.
 
So where does MQA flagging take place if it's not in the PCM data? I thought it was precisely the LSBs that trigger MQA decoding.
Im not entirely sure. @mansr is the best person to ask about this. He has made a tool that adds mqa flagging to any file
 
I don’t think that FLAC can encapsulate a high resolution audio stream that’s backward compatible with a standard one (CD).
How so if it already does? FLAC not only matches CD fidelity, it can exceed it if necessary. FLAC is just compression, the data insider of the container can be exactly the same as a CD, but with the benefit of being smaller if needed.
 
How so if it already does? FLAC not only matches CD fidelity, it can exceed it if necessary. FLAC is just compression, the data insider of the container can be exactly the same as a CD, but with the benefit of being smaller if needed.
I think that if you put a 24/192 audio stream in a FLAC container it cannot be played by something that can only play 16/44.1 unless it knows how to downsample it. I might be wrong, but I‘ve never heard otherwise.

An MQA stream can present itself as 16/44.1 and play as such on any legacy device, while being uncompressed to supposedly higher resolution on an MQA-capable device.

I’m not commenting on whether MQA can actually do that in any meaningful way other than simple upsampling, but to my understanding that’s the goal.
 
No, but nobody needs such a feature. What problem does it solve?
The same problem as DD+ with Atmos being backwards compatible with stereo, I guess. You get more, provided you have the means of decoding it, and it’s still functional if you don’t.
 
Would be interesting to do another analysis on your tracks from Spotify. What does the lossy compression do with those test signals? What is the difference? Does the encoder also go haywire?

A naive application of vorbis to the OP's "Original Masters" 16/44 file:
EDIT: original is 24/44

Code:
┌─ [dc@xps]:[~/Dropbox/audio_testing/mqa_test]
└─>  oggenc Original\ Masters/441\ ORIGINAL.wav -q 10 -o 441_orig.ogg
Opening with wav module: WAV file reader
Encoding "Original Masters/441 ORIGINAL.wav" to
         "441_orig.ogg"
at quality 10.00
        [ 99.8%] [ 0m00s remaining] -

Done encoding file "441_orig.ogg"

        File length:  8m 00.0s
        Elapsed time: 0m 05.6s
        Rate:         85.9589
        Average bitrate: 270.4 kb/s

┌─ [dc@xps]:[~/Dropbox/audio_testing/mqa_test]
└─>  oggdec -b 16 -o 441_orig_ogg.wav 441_orig.ogg
oggdec from vorbis-tools 1.4.0
Decoding "441_orig.ogg" to "441_orig_ogg.wav"
        [100.0%]

And the resulting spectrograms:
top is OP's original, middle is OP's "MQA Encoded" version, bottom is the vorbis-to-wave version. Gray is -100dB.


spectrograms.jpg
 
I think that if you put a 24/192 audio stream in a FLAC container it cannot be played by something that can only play 16/44.1 unless it knows how to downsample it. I might be wrong, but I‘ve never heard otherwise.
I suppose FLAC supports optional blocks, so there can be 44.1/16 as a standard stream and high-resolution data as an optional block. This would be backwards-compatible if players skip blocks they don't understand instead of barfing.
 
I still fail to see the reason people are so aggressively against MQA. It's a format made to deliver high quality music at acceptable bandwidth via streaming services by making intelligent tradeoffs on inaudible wavelengths and it achieves this goal, no?
 
Status
Not open for further replies.
Back
Top Bottom