MQA explanations wanted - fundamentalists need not apply

(Peter Lie) #21

The existence of multiple Tidal Masters of the same album indicates that this album was originally available in more than one Hi-Res masters. Both went through the MQA process to result in two Tidal Masters. Whether they sound different depends on the original masters.

It should be noted that they can have different regional licensing restrictions.

(Jeff) #22

Bob was there? He knows how every picture looked in real life?

A few mastering engineers of the original masters have indicated they were not involved in a “white glove” treatment. In fact we have had a few go as far as to say that what they hear from MQA isn’t how they wanted it when they mastered it.

To take it a step further, there isn’t white glove treatment being applied to bulk back catalogs. Just a generic “lens correction” in most cases. That’s money the studios aren’t going to spend. For MQA advertising of course they discuss the full mile and in special cases probably do do it.

Personally I don’t apply generic lens correction to my photos it does more harm than good.

It’s pretty simple, if you want true to what the mastering engineer heard buy the hi-res pcm. PCM captured from analog tape at high res doesn’t include these destructive filters in the audible range that MQA is “fixing”. Nor does it require extensive filtering on playback like redbook to reduce pre and post ringing.

If you want what Bob thinks it should sound like having never been in the room, use his DSPing buy MQA. Note that the MQA will have thrown away original data to achieve the oragami compression when applying it’s DSP.

Moving forward:

Modern digital recordings have so many analog to digital events and sometimes digital only events the best MQA could do is average or at best not do any “deblurring” without doing more harm than good.

(Daniel Beyer) #23

There are several sites out there that have investigated what happens during the various MQA stages. The rendering section is made up of an upsample and a particular filter selection. This filter selection, not the upsampling, is the “secret sauce” embedded into MQA DACs.

The MQA instruction information is not embedded with the music pre se, it is sent on very specifc bits in the file ( bit 8 I think), much like header information, so that the programs know where to look for the information.

In fact, it is possible to either truncate the MQA stream to 16 bit or replace it with non music information and still have the MQA DAC light and original resolution display.

An interesting thread to read on this subject would be:

(Jeff) #24

Just a follow up on this. This generic lens correction is probably why we have so many indicating that they like MQA on some albums/tracks and don’t like it on others.


But, how could this be possible. Bob promised us, that each MQA album is the best possible quality, sounding as it sounds in the studio on recording day, authenticated by the artist, the sound engineer and Bob himself.
So it is not possible that two different MQA versions can exists, especially not with different sound.


Looks like the authentication happens on the first 16 bit; this is done so it can be ported into a MQA CD, which is 16 bit at 44.1k standard. The only thing is MQA CD don’t have the upper 8 bit to reproduce the ultrasonic range like a 24 bit MQA file.

That’s fine, but if one hook up to a MQA DAC, you will probably see 176.4k not 44.1k! So, it is telling a truth? I know Tidal has album which is MQA 16bit 44.1k, when I playback using Roon, the signal path shows ‘to MQA 88.2k’.

(Daniel Beyer) #27

I don’t have one to try, but I suspect that you will see the original encoded rate. MQA DACs display the original master resolution, not what is actually being decoded, as far as I understand it.


It’ s all a play with numbers to please audiofools. Because higher numbers are always better!
Like the great philosopher Nigel Tufnel said: ‘These go to eleven’

(Peter Lie) #29

This is a feature of MQA to survive 16/44.1 AirPlay transmission.

(Anders Vinberg) #30

It isn’t “generic” lens correction, the correction is based on measurements of the lens. Adobe Lightroom contains a database of such lens measurements, and unlike audio recordings the lens and camera are identified in the file metadata.

Wrt “it does more harm than good”, few agree with you. Indeed, many modern cameras do such correction in the camera. And not just moderate cost cameras like Fuji, but even Leica.

In addition, Adobe Lightroom corrects for limitations of the printer, ink and paper combination, based on a similar measurement database. This corresponds to the DAC profile correction in MQA.

The concept is quite straightforward and widely accepted in photography.

(Andrew J Shepherd) #31

Assuming the lens aberrations are visible…

What if the photographer/cinematographer specifically chose that lens and its aberrations for stylistic qualities? Would the auteur then want a third party to “correct” those qualities with technology years or decades later?

Or what if the lens aberrations were not a stylistic intent of the creator but just a technical limitation of the time? Yet, the work became part of the cultural heritage with those stylistic qualities or artifacts intact. The same question about “correction” holds.


(Jeff) #32

No one said Lightroom lens correction is generic, my comment was an analogy to MQAing done on bulk catalogs.

I’ve used Lightroom for years, and always preferred DXO’s lens correction to it. Now I’ve moved to Capture One. I fully understand the specifics of it.

(Anders Vinberg) #33

Wrt the auteur wanting control of the corrections — isn’t that what “authenticated” is about, the blue light?

BW was a technical limitation, and colorizing old BW movies is questionable from an artistic perspective but widely done for commercial reasons. We can bemoan that. But the lens on the Leica Q was built with optical distortions, to keep the cost and size and weight down, and in-camera digital correction used to produce Leica quality images. I don’t see a problem with that. But if you prefer, Leica will be happy to sell you more expensive gear.

And you can play vinyl, which has no digital artifacts and no digital corrections, and is lossless.

(Anders Vinberg) #34

Yes, this is a fundamental problem with the whole correction idea.
Seems more applicable to old recordings.

(Jeff) #35

And I have to assume captured at lower sample rates where the analog filters needed to do so in the ADC caused the “issues” MQA is trying to fix with deblurring.

With modern captures at high sample rates or even with AD converters that are delta sigma based the problem should no longer be an issue at all. Maybe someone else can weigh in on this.

(Andrew J Shepherd) #36

But, somewhat ironically, analog low pass/anti aliasing filters do not exhibit pre ringing.


Why MQA is bad and Roon (shouldn't bother) (shouldn't be bothering) shouldn't have bothered with it :)

Is there info anywhere that tells us details about the sources for two MQA Masters appearing in TIDAL?

(Mark) #38

Agreed, and not only that your speaker-room interface has many, many orders of magnitude more effect on how a recording is presented to you


Maybe the artist asked for authentication and how the sound was in the studio has a schizophrenic disorder and gave two different answers :smile:

(JohnV) #40

Maybe one CAN serve two masters!

ok ok a bit of a stretch…