MQA disappointing

I find this discussion utterly fruitless.
One camp claims that there must be an audible difference.
The other camp claims that there cannot be difference.
Hearing is highly subjective and will differ depending on many factors such as

  • ability of hearing
  • out mental state at the time
  • resolution of the system
  • expectation bias
  • methodology of comparison used
  • differences in source material

and many others.
I does not look to me like any of the “evidence” presented (by either side) would hold up in a peer reviewed scientific context.
Can’t we just accept that some people (subjectively) hear a difference and others do not? It is not a contradiction per se.

3 Likes

I agree when the aim of the thread would be to find a consensus. I’m following this thread since the beginning and learned a lot about MQA from many viewpoints. This for me is/was very fruitful.

Do those who can’t hear differences in MQA in this thread hear any improvements with hires?

I don’t think you advance your argument by suggesting hearing or kit resolution being differentiators. I think we at this level all have pretty resolving kit, and the sort of things it has been suggested MQA does wrong such as pitch is something less affected by hearing so that is a red herring. I think a lot of difference is down to the way some people choose to listen. Personally I hear bigger issues between masters or formats (vinyl vs. digital) than between MQA and non MQA files. In short there are much bigger issues to go after in audio because even if MQA fails we will still be left with loudness and watermarking and even the spectre of DRM which can come back with any format. MQA done well is much less damaging than these.

2 Likes

I am not contradicting what you are saying at all. The discussion I find pointless is not the one about MQA from different viewpoints, but the one arguing the other camp is wrong (“I hear a difference” vs. “no you don’t”)

Yes, it is … et vivere, reservate.

But, the camp that claims there must be an audio difference is made up of those who are pro-MQA , anti-MQA or ambivalent. :laughing:

If the master is the same, I confess I don’t. I’ve always considered the hi-res game to be akin to the pixel wars with cameras, higher numbers sells more equipment. If it keeps a stagnating industry going, then so be it.

The xiph.org article about 24/192 downloads, and by extension even higher res recordings makes sense to me.

If the hi-res version (or MQA version) comes as a result of a better master or better post-production, then that’s great.

1 Like

This question is still dealing with absolutes. Life isn’t like this.

@Chrislayeruk thoroughly enjoys MQA and no one can take that away from him. He can hear the difference. It is arrogant to suggest otherwise. @Rhythmatist can hear a difference too. There is no right or wrong.

On occasion I hear differences–some favourable others definitely not–, so my preference can swing either way depending on the release. To me this is no different than comparing any two releases, but my first choice would probably be Red Book or a newer Hi-Res release.

Whether MQA is a good thing for consumers is a completely different discussion.

4 Likes

Good explanation of what MQA does in Sound on Sound.

The following two passages strike me as interesting:

As already mentioned, in its folded state, the streaming audio file has as typical bit rate around 1.5Mb/s (and never more than 2Mb/s, even in extremis). This is within the capabilities of a 3G mobile data service, and it presents no challenge at all for a 4G data link. Consequently, the MQA format satisfies the claim that it offers convenient and practical downloading and audio streaming for genuine hi-res audio files.

One of the original intentions of MQA is to transport a “like hires” music stream over a bandwidth constrained pipe and unfold it after delivery. This is clearly useful in mobile applications, probably less of a win in a home environment, where bandwidth is abundant.

However, this exacting level of time-domain performance is only possible when MQA is employed as a complete end-to-end system, encompassing both the original A-D sampling and encoding within the mastering process, as well as the consumer access and decoding stages. In this way, and with the implicit and precise knowledge of all digital filtering processes involved in the chain, the system’s overall impulse response can be made almost perfect. For example, MQA claim that the total impulse-response duration is reduced to about 50µs (from around 500µs for a standard 24/192 system), and that the leading-edge uncertainty of transients comes down to just 4µs (from roughly 250µs in a 24/192 system).

The full potential of MQA’s time domain performance can only be leveraged if MQA is applied / present already at the A/D stage. This is obvious (but often forgotten in the discussion). It is impossible for MQA to compensate for deficiencies of an A/D stage whose characteristics are not known.

How much of these improvements in the time domain actually materialize if MQA is only applied to the D/A process and whether they are audible is a different question.

Not really. There is no known A/D converter that goes right to MQA. At least not that I know of. So I have no clue what the heck they are talking about,

From what I understand music labels will sent a Hi-Res master copy to MQA Ltd to do encoding. This excludes the possibility that time domain correction is actually applied during the recording session (at the A/D) I believe once encoding is done by MQA in-house, some kind of post time domain correction is then applied based on characteristics of music, a kind of DSP.

Effectively, this is not really a true ‘end to end’ process unless time domain correction is applied to the recording at the studios (A/D). My understanding is any time domain correction applied may not survive through studio mixing and DSP, so this always done after the mix down to the stereo copies to a certain degree.

There have been thousands of posts on MQA, so apologies if this point has already been made.

Personally, I find it nearly impossible to hear subtle variations when switching from one source to another, if there is any lapse of time, even a minute or so. Such as listening to an MQA file in Tidal, and then listen to the same non-MQA file within Tidal.

However, one way to compare MQA and non-MQA almost simultaneously and in real time is to use the Mytek Brooklyn+ and with the remote, turn MQA coding on and off. An engineer explained to me that such a comparison is actually biased in favor of MQA. Since the MQA file has the MQA coding added to it, that file should, in theory, sound worse with the coding turned off, as compared with the original file without any MQA coding at all.

I already knew that I couldn’t a difference in Tidal when stopping the playback of an MQA file and then starting the playback of the original non-coded file. When making the same comparison, with an MQA file with coding, and turning MQA coding on and off in real time, I also couldn’t hear any difference. And the latter test was actually biased in favor of MQA, since theoretically an MQA coded file with the coding turned off should sound worse than the original file.

I gave up on MQA at that point and never looked back, so to speak.

But they got paid if you bought the Brooklyn+. You are still a customer!

Interested to know, how you know this as he states he doesn’t?
Where did you find the information to support this?

I think the point they are making is the following:
MQA
The improved time behavior of MQA (blue) vs. PCM (red) is for an end to end system. If the recording was made using PCM and mastered using PCM, it is in the “blurred” (red) state after production. No amount of encoding / decoding with an timewise extremely accurate system will restore the red curve to the blue curve. It will (extremely accurately) remain what it is.

With my own eyes. He sells banner space on his website to hi-fi manufacturers.

What is the graph representing, an Impulse response? To me it looks like a weird flattenned impulse response before and after a minimum phase apodizing filter is applied. Such a filter would screw up timing of certain frequencies to remove pre ringing not fix the time response unfortunately.

And why does the blue have jaggies at the top?

This Discussion Have Since It Started The 8 of May 2018!!!
Showed 10,9K Writing’s 617 Hallo!

Stop Wasting Time On MQA and Listening Too It Or Chose CD, DSD…Instead!

Listen to music and enjoy instead!

I have a question!
A lot of people on Roon Community writing Red Book/Books?
When I buy a CD, so does I ask if the store have xxxxxxx latest CD!
Do you ask if they have xxxxxx latest Red Book?
Just of curiosity!

Love & Respect

@alex_wood, @Chrislayeruk, there’s also this, if you’re curious. And a bit of context, from the same source, higher up in the same thread.