Hi-res music streaming via Qobuz / tidal MQA in Audirvana and a recent DAC?

Maybe a mqa file will change resolution on the right, that i dont use.

Upsampling us not bitperfect playing, not necessary bad, but you have to know how to set up those sliders to like how it sounds for you, i don’t know how, so i just play bitperfect music.

When i play 16/44 it shows 32/44, 24/96 shows 32/96, when 24/192 it shows 32/192 when dsd64 it shows dsd64, when playing dsd256 and dac cannot, shows pcm 32/352 :grinning:

1 Like

I did a lot of subjective testing when I first purchased Audirvana 2 years ago and you really can’t go just by the numbers. Some CD-quality recordings sound fantastic and some high-res audio files sound like crap. Much depends on the quality of the original recording and the faithfulness of the digital transfer. Streaming in and of itself imposes its own limitations. At first I couldn’t understand why streamed music from Qobuz didn’t sound the same as playback of the identical, purchased files. It took me over a year to get Qobuz to admit that they deliberately limit the bandwidth of streamed files to minimize problems with drop-outs. The result is that the stereo sound stage sounds flat with Qobuz streamed content. If you want true high-res stereo audio, you need to download the files.

With Tidal, you need to understand that MQA is a form of lossy compression for high-res audio. It allows them to stream high-res content without using more bandwidth by applying a mathematical model to approximate the original content. Depending on your hardware support for MQA, you may or may not prefer it to 16/44.1 CD-quality audio. Personally, I think that Audirvana does a better job of simulating high-res audio from CD-quality music than one gets from MQA.

From my own tests, I know I can clearly distinguish between 16-bit and 24-bit versions of the same recording when played back with Audirvana. I cannot distinguish between 24-bit and 32-bit files, however, nor would I expect that any human is capable such an incredible dynamic range. On some players, I find that DSD versions sound more natural, but with Audirvana I’m absolutely unable to tell the difference between PCM and DSD versions of the same recording. My ability to distinguish 96kHz or higher bit rates from 44.1 or 48kHz is variable and depends on the specific recording. Perhaps if I were younger, I might be able to hear more of a difference. Truthfully, very few audio systems, even very expensive ones, accurately reproduce sounds much beyond 20kHz, so by the sampling theorem, 44.1kHz should be adequate. Nearly all modern DACs are capable of high-order digital filtering, so it really comes down to the accuracy with which the original recording was filtered prior to sampling and digitization.

If you want a good test of what your system and your ears are capable of, download one of the Jazz at the Pawnshop albums from HDtracks. These are 24/192 or DSD128 reproductions made from the original tapes of live performances. Listen to the sounds of dishes clanking around you and quiet conversation while the musicians play. You should be able to distinguish with pinpoint accuracy the precise location of every dish, every conversation and every instrument on stage. The music truly sounds live, and that’s what high-res playback is all about. My hearing may not be what it used to be, but so long as I can still isolate individual sounds and instruments, I’ll spend my cash on high-res audio.


If there is really any audible difference, it is most certainly not coming from the bit depth.

What we can hear is differences between how DAC and other components are handling different bit depths and sample rates. This makes them sound different.
But, if say 24 bit signal is temporarily converted to 16 bit and then back to 24bit, there is no more this audible difference that could be detected in double blind tests.

Of course, if we dismiss double blind tests as many do, then placebo effect is a real thing. Too much money spent on implanting Hi-Rez nonsense in our brain, and brain is actually what is “hearing” things.

1 Like

I don’t disagree with you. I’m a scientist and double-blind tests are the absolute best way to look for a statistically significant effect within a population. My own subjective testing of different tracks by their nature couldn’t be fully blinded, yet I did try by making sure I compared tracks of differing bit depth and sample rate that originated from the same master. I won’t discount the possibility that I could discern differences based on the way the DAC handled different bit depths and sample rates. That said, I didn’t try to tell which track was which bit depth, but rather I focused on which tracks most faithfully reproduced the sound stage. Some 16/44.1 tracks sounded fantastic in that regard, and some 24/192 tracks sounded dreadful. A good case in point was Carole King’s Tapestry, released in 1971. I remember the first time I listened to that album, at the age of fifteen, and thought how wonderful the music was, but how terrible the recording sounded. I later replaced my worn vinyl copy with a CD, which sounded no better, and then with a 24/192 version, which still sounds terrible. How ironic that one of the greatest albums of all time is one of the worst mastered of all time. No amount of additional bits can salvage a bad master.

I think there is a small subset of us who have exceptional hearing. We account for perhaps less than a percent of the audiophile population and so we’re lost in the statistics of double-blind tests, but those of us with perfect or near-perfect pitch who can discern the difference between a trumpet and a French horn, or between a single and two French horns, can hear subtleties in music that most people can’t. The only thing that matters to me is that I can hear the difference between 16-bit and 24-bit depth on most recordings, and that the latter sounds much better to me. I’ll acknowledge that some of that could be due to the differences in how my DAC handles the different bit depths, but what does it matter if the differences allow me to hear things I cannot hear on the CD? What difference does it make if the result is a deeper, wider, more natural sound stage?

1 Like

I would hazard a claim that most of the differences you hear are not related to the sample rate or bit-depth but rather the quality of the recording and mastering.

1 Like

You missed the point that I was very careful to select CDs and high-res downloads that were made from the same original master. Yes, the quality of the original recording, the manner in which the tracks were mixed, the use of dynamic range compression and the quality of the master are much more important than is bit depth, but unless the master was doctored differently for the CD and the high-res versions, there shouldn’t be any difference other than bit depth.

I’ll give you the point for sample rate for the reasons I already gave, but if bit depth weren’t a factor, the SACD, DVD audio and HD-CD formats would have never been invented. Even as the engineers at Sony and Philips worked on developing the compact disc, they acknowledged that the format was a compromise.

Some of the very best recordings of all time were made in the 1950s using simple condenser mikes, pentode-based tube amplifiers and multi-track Apex reel-to-reel tape recorders. Unfortunately, the tapes themselves have deteriorated with time. The substrate stretches and the magnetic media is subjected to slow but steady gamma ray bombardment. The original tapes are fragile and unlistenable, but they’re being remastered digitally and through the wonders of digital signal processing, the original quality can now be restored. They’re being rereleased as high-res audio, DSD audio and, yes, on CDs. Several of my tests were done using these remastered gems. My testing wasn’t double-blind, but it was as random and blind as I could make it, with tracks from identical masters but with different bit depths and sample rates.

Not everyone hears music in the same way, and some people have unique abilities. A voice like Frank Sinatra’s comes along once in a lifetime, if that. Itzhak Perlman, Joshua Bell and Julia Fischer are great violinists, but they still can’t match Jascha Heifetz. True perfect pitch is exceedingly rare and I’ve only met one person in my lifetime who actually had it. Relative pitch is much more common, but still occurs in only a fraction of a percent of the population. It would be interesting to perform a double-blind test of bit depth and sample rate on a population of only people with perfect or near-perfect pitch. I would be shocked if you could compare their results to that of the general population.

My hearing isn’t what it used to be, but I can still pick out the individual musical instruments in a high-quality recording. Sometimes there is very little difference between the quality of a CD and a high-res format, but all too often, the difference is as stark as that between color and black and white. To someone who’s color blind it might not matter, but to many of us, it does.

1 Like

Correction: I was wrong, sorry, because in reality the behavior of the media player changes according to the upsampling setting: if upsampling is disabled then the data in the lower right corner does not change, if it is activated it changes to adapt to the real one.

Yes, the complexity of the topic is extraordinary because it involves a large number of factors (technical, scientific and artistic) that are largely still undefined or unknown.
Even I no longer have perfect hearing, quite the opposite: I suffer from a tinnitus (a phenomenon that medicine still does not know from what it originated).
So how do I decide which recording is the best (for me)?
In a completely unreliable way from a scientific point of view. If listening in dim light to Thelonious Monk or Canned Heath or Glenn Gould it seems to me they could be present (albeit from a not optimal listening point) then that is my ideal musical reproduction (regardless of any objective factor).

Finally I understood what is the sense of downloading files from Qobuz (paying twice for the same service in essence)!
I asked myself several times.
However, I do not find that it is a correct practice, especially as it is not clearly declared to customers, indeed at all.

People do this sort of thing all the time when they take out a mortgage to buy a house. They end up paying the bank several times the purchase price in interest, but they are able to live in the house while paying it off. During that time, the bank owns the house and in effect rents it back to the purchaser. It’s a rent-to-own scheme, but packaged as a loan.

Qobuz both rents and sells music, or more correctly, the license to listen to music. The copyright holder always retains ownership of the actual music. However, you aren’t really paying twice for the same service. When you subscribe to Qobuz, you pay for the right to listen all you want, as much as you want to music from their entire catalogue. The only thing deceptive is that they don’t tell you how they limit bandwidth during streaming. You can also download the music and listen to it all you want, so long as you maintain your subscription. However, the key advantage of a subscription remains that you have access to the entire catalog and not just to your purchases.

Qobuz also sells music at competitive prices. You don’t need to be a subscriber to purchase music from them and you can listen to the purchased music for as long as you breathe. As with a house you own versus one you rent, it’s yours to keep, forever.

Qobuz also offers a combination subscription, their so-called sublime plan, that lets you stream their entire catalog of music and purchase music at a substantial discount. Even if you never listen to streamed music, a sublime subscription pays for itself if you spend more than $50 on music purchases a month. Further, a sublime subscription lets you listen to an album in its entirety before you buy it. Otherwise you can only listen to song snippets. It’s actually a very good deal if you’re a serious audiophile who listens to a lot of music.

In contrast, Tidal only rents their catalog of music by subscription. They don’t need to limit bandwidth during streaming because they only stream CD-quality music and MQA, which involves lossy compression. If you want to purchase something you heard on Tidal, you have to go elsewhere and pay full price. Tidal does have a much larger catalog, but it’s hard to get passed their hip-hop-centric portal.


I agree with your very detailed and correct comparison between the two suppliers (unfortunately only two, we hope that they increase to improve quality through competition). I would just like to add that the fact that Qobuz hides the superiority of the purchased files suggests that they themselves realize that the practice is strongly incorrect towards audiophiles. In fact, if the different quality was explicitly declared, it could increase sales. However, they prefer to hide it.

I don’t believe this is the case. They might throttle the the speed and deliver lower resolution file if the bandwidth is not there. Tidal does that too. If the connection is slow, or you’re on mobile, you might get MP3 file even though you subscribe to HiFi tier.

If they say it’s a 24/192 file playing it’s that file. Don’t believe that they have a separate file that has been doctored to show incorrect info. That would be fraud end they would be liable.

1 Like

It’s true.
However, is it a fact that more than one user claims that the audio quality of the purchased files is superior?
The hypothesis that the slower connection implies a drop in quality seems rather remote, considering that the bandwidth is always verifiable.
So just suggestion?

There might be other issues at play here, if the quality is indeed inferior (it could be also just a perception bias).

1 Like

Possible, certainly.
However, I think that some users with the necessary technical skills will sooner or later be able to ascertain whether a decrease in resolution is actually made only in case of insufficient bandwidth.

I’m sure there are limits placed on file quality by the record labels too. As it is, the labels pretty much set the prices. That’s why competition hasn’t resulted in lower prices. There are regional restrictions too. For example, I’m not much of an opera buff, but there’s a wonderful boxed set of all of Marie Callas’ studio recordings available - you just can’t buy it in the U.S. It’s available in the U.K. on CDs and if you have an address in the U.K., you can order it from Amazon. You just can’t order it from or have it shipped to the States. There are regional restrictions and pricing restrictions that the various on-line stores and streaming services can’t get around.

I strongly suspect that the labels place restrictions on the streaming of high-res content. Otherwise, bandwidth shouldn’t be an issue. I have a gigabit connection on my end and, after all, most high-res movies stream at about 50 MB/hr, with a 4k HDR version streaming at double that. I think the bottom line is that the record industry doesn’t want lossless, high-res versions getting away from them. Not at a price consistent with streaming anyway.

BTW, I managed to get a copy of the Marie Callas boxed set last Christmas on sale at 90% off from A&K, in high-res on SD cards no less.

1 Like

Bandwidth is a cost to the streaming service. That’s why Tidal went with MQA as it allows them to stream at 24/48 and deliver “High-Res” that way.

Also music rights holders prefer to rent the music as opposed to you buying it. That gives them a continuous revenue stream, they sell it to you over and over again. When you buy your music it’s only an one time transaction. Sure they’ll try to sell it to you multi times in different formats (CD, high-res PCM, DSD), but that’s it.

1 Like

Steve was not reportedly referring to a bandwidth cost issue from streaming service providers. Instead, he was referring to a possible lack of bandwidth by the customer of the service, a deficiency that would justify, so to speak, the lowering of the advertised definition, and therefore of the quality of the music. The fact that the use of broadband also constitutes an additional cost for streaming operators, if anything, may possibly increase doubts about the effective management of the lowering of the definition.

Regarding the economic interest of the record companies and therefore their commercial strategy, I think that every opinion is currently quite questionable.
It should be possible to consult data and statistics on the income of the individual channels and sales formats which, beyond some journalistic rumors, are probably jealously guarded.

Certainly the spread of the so-called “liquid music” at quality levels equivalent (or even higher) to those of any physical medium (and in particular the CD, which required massive investments and guarantees many more jobs) would determine for the industry production of the “discs” a further decisive collapse compared to what has been since its inception.
And just as surely, as Bitrade rightly points out, the ideal solution for record companies would be to keep, on the one hand, users of modest needs (“disposable” music) satisfied through music of the lowest quality level such as mp3, spread through low-cost mass streaming services. And on the other hand to supply the users of music that needs higher quality (classical, jazz, rock) through CDs (or other physical media) continuously updated (or supposedly such).
But is this optimal solution (for traditional and consolidated industry) really feasible?

It is known that stopping technological evolution is practically impossible, a consortium of industries can only delay it as much as possible.

High-definition (more or less) liquid music is here among us and is destined to remain there.
The technological possibility of high resolution streaming has remixed the cards
If the powerful record companies have started to yield to Tidal and Qobuz, it is likely that they realized that, for various reasons (including digital music “piracy”), closing them in front of emerging technology is no longer convenient.
It should also be noted that the hardware side of the music industry has already fully adhered to new technologies by offering equipment of remarkable quality at all price levels.
However, it is also likely that Tidal and Qobuz are still considered by the major as tests for now to clarify market trends (mainly about the sale of really hi-res files) and the most profitable way.
It shouldn’t take much longer to understand their intentions.

1 Like

I received a courteous, precise and honest answer from Qobuz to the question I asked.

"Qobuz doesn’t use files with 32 Bit sample rate, but they are up to 24-Bit.
All files up to 192Khz can be explored on the site, not directly on Audirvana. "
With a further very useful clarification: “Attention, the sample rate is a fixed datum, while the sound frequency (measured in Khz) is a variable, for this reason we talk about 24-Bit up to 192Khz.”
Although I must confess that, due to my (very) poor technical skills, this latter explanation is difficult to understand. But surely it will be very clear for you Damien (our guru), and perhaps for some technically more prepared user than me.

I must also certainly agree with Steve that, for example, in Coltrane’s discography available on Qobuz, “Live AtThe Village Vanguard” in 24/192, it is of such a superior quality that it is distinguishable even to imperfect (but attentive) hearing.

Best wishes to all

Are you really sure that you are comparing the same master/remaster? According to Discogs this album has not less than 21 editions on CD spanning from 1987 to 2016 …