You are here

Tidal MQA tested, worse than FLAC

All about the tools and techniques involved in capturing sound, in the studio or on location.

Moderator: Moderators

Tidal MQA tested, worse than FLAC

Postby DanDan » Mon Apr 19, 2021 11:39 pm

User avatar
Posts: 89
Joined: Tue Feb 10, 2009 1:00 am
Location: Cork Ireland
Dan FitzGerald MIOA  MAES

Re: Tidal MQA tested, worse than FLAC

Postby James Perrett » Mon Apr 19, 2021 11:57 pm

I think that there have been doubts for some time about how well MQA lives up to its claims.
User avatar
James Perrett
Posts: 10647
Joined: Mon Sep 10, 2001 12:00 am
Location: The wilds of Hampshire
JRP Music - Audio Mastering and Restoration. JRP Music Facebook Page

Re: Tidal MQA tested, worse than FLAC

Postby Hugh Robjohns » Tue Apr 20, 2021 10:30 am

I'm watching the video now...

There were/are some interesting technical concepts and ideas involved in MQA which I tried to explain in this article, but fundamentally, MQA is a commercial marketing exercise intended to generate profits for all those involved in file creation and distribution.

The authentication aspect -- the little blue light -- is little more than a checksum to say the file is as it should be. It might offer some consumer and retailer confidence, but that's about it. It's not that reliable or foolproof.

The convenience aspect of MQA is important as a big marketing plus -- sensible file sizes on a par with legacy CD but with claimed better 'resolution' makes it appealing, but that whole aspect is less of an issue these days than it probably was when the concept was conceived.

There's a lot of jiggery-pokery involved in the way the wanted high frequency elements of the signal are rolled into the format.

The interesting bit about MQA for me, was the system's claimed time-domain benefits. Some of the technology used in that respect is already known and proven -- like apodizing -- but some is new and I found it incredibly difficult to get proper explanations about that from any of those involved at MQA. Which surprised me quite a lot. I understand commercial sensitivities, IP protection and the rest. I sign NDAs to protect information like this regularly... but MQA were very unwilling to share technical information or even detailed concepts and in fact I'd go as far as to say that they were intentionally vague and difficult when presented with probing technical questions.

I also know for certain that some of the system demonstrations played to me at their HQ were not as they claimed, and despite lots of promises in advance we at SOS were never allowed to run any tests of our own with our own material...

So in the end I took it all with a large pinch of salt... I'll watch the video and report back on anything I think relevant.

... and now I've watched it. Nothing really surprises me. I do think some of the ultrasonic material tests might be misleading because the signal levels were way higher than the system might expect from 'normal' material, so some of the aliasing artefacts might not be as relevant as they appear... but the noise and distortion and other artefacts from encoding and unfolding in general are disturbing, as are the changes imposed on normal signal content.

One thing I'm not entirely clear about is the claim of MQA being loss-less... My understanding was that elements of the HF fold-down process were loss-less in that those HF components could be restored accurately in a suitable decoder. But the 44.1 MQA file is not, and never could be, 'lossless' in the way a FLAC file is.
User avatar
Hugh Robjohns
Posts: 30705
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound
In my world, things get less strange when I read the manual... 

Re: Tidal MQA tested, worse than FLAC

Postby Martin Walker » Tue Apr 20, 2021 6:14 pm

Fascinating video! (I've just finished watching it too)

It's a shame that the files submitted with lower levels of ultrasonic content were rejected, because we can't now get a better handle on how the MQA process deals with a more 'realistic' hi-resolution spectrum, but overall there does seem to be a lot of obfuscation involved with the MQA process.

User avatar
Martin Walker
Posts: 17496
Joined: Wed Jan 13, 2010 9:44 am
Location: Cornwall, UK

Re: Tidal MQA tested, worse than FLAC

Postby Folderol » Tue Apr 20, 2021 6:28 pm

Some very interesting comments in the Wikipedia article on it.
Two things stand out for me:

Licensing looks like an attempt to screw over everybody for no real benefit.

The 'lossy' word. Whether on not you can hear it, and despite all the hand-waving. If anything in the processing chain is lossy, then the whole file is.
User avatar
Jedi Poster
Posts: 12842
Joined: Sat Nov 15, 2008 1:00 am
Location: The Mudway Towns, UK
Yes. I am that Linux nut.
Onwards and... err... sideways!