It's Official: People Can Hear High-Res

The AES has just published a peer-reviewed paper by Dr. Joshua D. Reiss titled "A Meta-Analysis of High Resolution Audio Perceptual Evaluation" which concludes:
"The overall conclusion is that the perceived fidelity of an audio recording and playback chain can be affected by operating beyond conventional levels."
In other words, yes, people can hear the difference with hi-res. My ears and brain and experience rejoice in this positive reinforcement!

Back in December of 2015, we reported on what John Atkinson of Stereophile reported , which was an AES workshop presentation by Dr. Reiss of Queen Mary University, London, where Dr. Reiss (see his bio ) performed meta-analysis on 80 published papers concerning high-resolution audio wherein he found that people can, in fact, hear a difference between high-res recordings and their lower-res counterparts. But that was just a workshop presentation. The AES has just published his peer-reviewed paper which makes things more...official.

One interesting thing to note about the 80 papers used in this meta-analysis is that Dr. Reiss limited the use of the results from the seminal paper by E.B. Meyer and D.R. Moran, "Audibility of a CD-Standard A/D/A Loop Inserted into High-Resolution Audio Playback" which found that people could not hear a difference between CD-quality and high-res (SACD & DVD-A):

"Although full details of their experiment, methodology, and data are not available, some interesting secondary analysis is possible. D. Dranove noted that 'the percentage of subjects who correctly identified SACD at least 70% of the time appears to be implausibly low.' In trials with at least 55 subjects, only one subject had 8 out of 10 correct and 2 subjects achieved 7 out of 10 correct. The probability of no more than 3 people getting at least 7 out of 10 correct by chance, is 0.97%. This suggests that the results were far from the binomial distribution that one would expect if the results were truly random."
Meyer and Moran have themselves stated, "... there are issues with their statistical independence, as well as other problems with the data. We did not set out to do a rigorous statistical study, nor did we claim to have done so. ..." There have also been serious questions raised regarding the music used, the test environment, and the fact that the test was not a controlled test. In other words, going forward (or backward), if you see a reference to the Meyer and Moran paper citing it as support of the idea that people cannot hear a difference with high-res, you'll know the author doesn't know enough to be making his or her claims with any confidence.

Dr. Reiss' findings dovetail nicely with the recent paper from Tim Wescott "Sampling: What Nyquist Didn't Say, and What to Do About It" which calls into question many people's [mis]reading of the Nyquist-Shannon sampling theorem thereby shaking the very foundation of the digital world that many people have chosen to build upon.

The only question remaining is—are we done? Are we done with the whole "good-enough" approach to digital music reproduction? While we should be, I'm not holding my breath just yet.

Download "A Meta-Analysis of High Resolution Audio Perceptual Evaluation"

DH's picture

Could differentiate hi-res 60% of the time. Just what audiophiles have been claiming for years.

monetschemist's picture

Now all we need to do is remind listeners that just because it's high res doesn't mean it's better (c.f. that terrible new Elvis Costello /My Aim is True/).

By the way that is a most hilarious photo! I'm thinking the grinning young man with fingers in ears is young Werner Von Braun thinking "I can do better than that..."

Michael Lavorgna's picture
monetschemist's picture

... what I'm complaining about is when someone releases (or re-releases) an album in high-res format but the release quality itself is terrible.

One example I mentioned is the Elvis Costello. Compression / loudness to the max, shrill and teeth-grating. What were they thinking? Pretty sure the CD version I have from the late 80s sounds better.

Another boo-boo is Ronn McFarlane's /Indigo Road/ which, at least in the version I received, has 16 bit data in its 96/24 envelope. Its real bit rate reports very low. Hmmm.

Having said that, I also have lots of lovely high resolution music, both new and re-releases, and so I remain optimistic - especially with the publishing of this paper - that high resolution formats will continue to enable and encourage better sounding music.

Michael Lavorgna's picture
But what you are talking about isn't really germane to the topic at hand. This paper is concerned with audibility, not sound quality variation among high-res releases.
monetschemist's picture

enough said.

j. phelan's picture

..most people hear very-small improvements in high-bit (over CD). But if we isolate gear with cutting-edge focus on RBCD, they hear no difference at all !

Michael Lavorgna's picture won't move any faster ;-)

So you are using anecdotal information to refute the claims in a peer-reviewed AES paper that included 18 published experiments, over 400 participants, in over 12,500 trials?

j. phelan's picture's what I hear everyday. And reviews of (top-end) CD playback indicate very slight differences. Some were published in Audiostream...

Take your pick on studies. It seems the vast majority of (Internet) forum users are siding with Meyer and Moran. There is very little excitement in the 'advance' of hi-rez.

Even Robert Harley says (in the July issue of TAS) that Meridian's MQA is "far greater of an advance" than any hi-rez system was.

I think the days of touting high-bit formats are coming to an end...

Michael Lavorgna's picture
> And reviews of (top-end) CD playback indicate very slight differences. Some were published in Audiostream...

CD playback reviewed in AudioStream? Please share those links as they don't exist ;-)

As I said, your references are all anecdotal. What you hear every day, Internet forum users, Robert Harley's embarrassingly ridiculous MQA love fest, refute a scientific study? Perhaps in some alternate universe where unicorns roam free.

Do you understand that "high-bit" formats is a nonsensical term? It's either bitrate, bit depth, or sample rate.

I find it's always best to know what you are talking about before talking or typing.

j. phelan's picture

...on the Meitner MA-2, you said "it all but zips up the remaining loose ends (between CD and hi-rez). Said it loud and clear !

'Computer Audiophile' couldn't hear a difference between hi-rez and CD - with the Berkeley Alpha Ref DAC. But there are many others -

Then, you're cherry-picking your studies - not all say hi-rez is better.

Finally, there's nothing 'nonsensical' about formats. 24-bit vs. CD - that's the debate right now.

While not alone, your love fest for hi-rez is unique in high end audio...

Michael Lavorgna's picture
You said we reviewed "(top end) CD playback" as compared to high-res. We did not.

I am not cherry picking my studies, I'm reporting on the most recent study.

You misunderstood. I said "Do you understand that 'high-bit' formats is a nonsensical term? It's either bitrate, bit depth, or sample rate."

There is no debate right now in terms of this study.

> your love fest for hi-rez is unique in high end audio

Nice try but well off the reality mark. Again, try to keep your comments here to things you know and understand.

Michael Lavorgna's picture
Here's a sentence that follows the one you partially quoted:
That said, I would not want to give up my higher resolution music for its CD counterpart and something like the truly stunning 24/88.2 A Calm In the Fire of Dances by Deep Rumba from Kip Hanrahan's American Clavé label from HDtracks is downright dripping with flavor when played through the MA-2.
If you want to continue this conversation, I'd suggest you sharpen your game.
j. phelan's picture

...but studies never meant a thing to audiophiles. If they did, they would have adopted solid-state gear in the 1960s and digital-CD in 1983.

You "wouldn't want to give up hi-rez". No, you wouldn't - but you said the "differences are zipped up" !!

Michael Lavorgna's picture
...I wonder why I bother.

This is one of those times.


Doak's picture

Get over it. All (other) things being equal - recording, mastering, transfer, etc - high res makes a very REAL improvement in SQ.
That's it. Done.

nick's picture

i agree with this comment entirely, and i have tried almost every hi-res format ever available. a properly designed dac playing rebook can come so close to hi-res that the difference is negligible. in fact, i think redbook through my current pcm 1704 based dac sounds better than any hi-res recording i've owned. just my opinion though.

Michael Lavorgna's picture
...when it comes to AES papers. For example, if people who are involved in recording music and designing ADCs and DACs were to say, "Ah who cares, some people think CD-quality is good enough so why should we bother recording in and making anything but CD-quality stuff."

My point being, there are very good reasons why recording studios went high-res years ago. As I argue here, there is absolutely no reason why a high-res original should be delivered to us in a CD-quality container when it's a download. I want the original recorded quality.

The point about "a properly designed DAC" and how Redbook can come close to high-res is certainly a valid one * but * it does not refute the outcome of this study which is that people can hear a difference with high-res playback.

Or, to put it another way, why should we be against the advancement of technology? The CD-standard was developed over 25 years ago and was based on the technology of the day which is, to be kind, hopelessly outdated.

nick's picture

you're right, in most circumstances/systems hi-rez will sound better. progress is a good thing. just for me personally, it's not worth the extra $ that is usually required. to each his own.

NeilS's picture

Maybe I missed something, but was the paper's conclusion about higher resolution affecting perceived fidelity necessarily a conclusion that it affects perceived fidelity positively?

"...In summary, these results imply that, though the effect
is perhaps small and difficult to detect, the perceived fidelity
of an audio recording and playback chain is affected
by operating beyond conventional consumer oriented levels.
Furthermore, though the causes are still unknown, this
perceived effect can be confirmed with a variety of statistical
approaches and it can be greatly improved through

Michael Lavorgna's picture improved fidelity. Whether or not people prefer it is another matter.
monetschemist's picture

we all have poorly recorded albums that sound better when we don't play them at all!

That doesn't imply that lower quality / resolution playback is a bad thing.

NeilS's picture

Is it fair to say that's your conclusion, not that of Reiss?

Michael Lavorgna's picture
The definition of high resolution audio is not in dispute. The question that was asked, and answered, was can people hear a difference.

If you'd like to argue that high resolution means a loss in fidelity as compared to CD-quality, you have a mountain of research to overcome.

NeilS's picture

I didn't say anything of the sort. I just asked two questions. I may have misunderstood your closing comment to the article which seemed to me that you were saying CD quality isn't "good enough" based on the conclusions of the Reiss paper.

"The only question remaining is—are we done? Are we done with the whole "good-enough" approach to digital music reproduction? While we should be, I'm not holding my breath just yet".

Michael Lavorgna's picture
I meant no such thing. I've expressed my opinions on this topic here. From that post:
"For me, high resolution music as some logical differentiating factor is dead. For new digital releases being sold as a download, give us the original recorded quality: Whether that means 16/44.1, 24/44.1, 24/48 and so on. And let's just call it what it is—lossless."
monetschemist's picture

A nice concise way of stating the option, thanks Michael.

And it probably does bear repeating, given that there is still a tendency to make lossy and downsampled versions available but not the original quality.

Michael Lavorgna's picture
I'll see if I can work it into other things in the future ;-)


NeilS's picture

for the clarification.

Michael Lavorgna's picture
I often wonder how much I should repeat myself and tend to err on the side of - not very often. But in some cases, such as this, it may have made things clearer.


NeilS's picture


rl1856's picture

I attended a demonstration at Axpona Atlanta that highlighted differences in resolution quality. The same tune was played multiple times, at different resolution levels. It was quite easy to hear the improvement in moving from basic MP3 to 256/320/CD. However, the biggest jump in sonic quality was from CD to 24/96, with a very small but noticeable improvement when moving to 24/192. Everyone in the room nodded their head in agreement, including John Atkinson who was in the room at the time.

On a different but related note, it has been proven that humans are capable of hearing sonic content above 20khz and are aware when ultrasonic content is missing or distorted. Makes sense then that humans can identify differences between RedBook and higher resolution levels.

Brown Sound's picture

Hey Michael, that link is doing the 404 dance. Sorry, sir.

Michael Lavorgna's picture
Fixed. Thanks for the heads up.
jim tavegia's picture

IMHO everything in the recording chain matters, especially in reissues if you want to hear an improvement over the original release. This has been talked about a great deal.

How was the initial tracking done and are those masters in great shape and stored properly? What generation are you using? Who is mixing down from those masters and then who is doing the mastering for the release? All of this matters and sloppiness somewhere will render any improvements null and void. I am convinced that not everyone is capable of properly doing a high resolution release. I own some that are not that much better than the original LP, which was not that great to begin with. $30 wasted on an SACD release.

I have some excellent red book releases and more often than not I can tell that who the mastering engineer was as I now have some favorites that I can count on. Unfortunately there is too much money to be made is just "reissuing" something and unless you read a review it may not be better than the first release of it. Unfortunately at $30 to $50 a pop you must decide if the gamble is worth taking.

I just bought a download of a redbook release from PrimeSonics of Clarinet and Piano music that is excellent. You can tell that the folks, from start to finish, really cared about the performance and finished product. It is only redbook, so great sound can come from CD quality.

My own experience now is that if the tracking starts at 24/192 or DSD with no overs one can make a stellar release, but the mixing and mastering must be done by someone who cares more about the sound than just making money. Great tracking of a performance can be ruined by someone up the chain.

And finally, the listener must be more than a casual listener and really care if something sounds better or not. To ask me to be part of a wine tasting trial would not make sense in that I am not a wine drinker and know little about it, other than what might taste better to me, but it would harder for me to be a good judge of what "better" is. You would have to really care about it to make an honest judgement. I am not sure in these audio tests there are not some biases before they are started.

I am amazed at how great some of the old releases are of my favorite artists compared to some of what is new coming out. Often less engineering is more. When I look at all the software plug-ins being offered for sale for use in DAWs, I am not surprised at why many recordings today sound like they do. Just because you bought a plug-in to alter a recording doesn't mean you should use it. Too often they are over-used.

brianls's picture

Jim - You put your finger on something that is so important - What is the providence of the music we can buy online? You're right that it's a complete gamble to buy a 96/24 album because it could simply be a 44.1/16 that was up-sampled. When I read that people are using spectral analysis on downloaded tracks to see if they got ripped off, I'm less likely to take that risk. That's why I appreciate sites like They actually have rules about what is meant by 'native DSD' on their site, and they certify that all of the music they sell conforms to those rules. That doesn't guarantee that the SQ will be awesome in every case, or that what was recorded is worth listening to in the first place. But at least the providence is not in question. They even describe how each recording was done - what equipment was used often including mics, cables, power conditioning, and room treatments, and who the recording engineers were. Wonderful. (I don't work for Native DSD, but I do buy a lot of music from them :) )


jim tavegia's picture

I have not done much homework on spectral analysis, but I should as I do much recording from home and local schools and universities. Sony offers some nice software and I really should get into it and, like you say, know what I am buying. I do trust my ears, even though they are old ones.

I can say that in my own work, doing less to a sound file is a very good thing. I would rather do take 10 then tell someone I can fix it later, when often it can't be fixed. I never track at anything less than 2496, but have my eye on a Tascam DA-3000 that does up to 24192 pcm and DSD. That might be nice.

I was listening to a sound file off a plug-in company's website and they showed what they could do to a vocalist to make her sound very good. The problem was that when you heard her bare track, she was awful and missed so many notes, which they fixed all right and then drenched her voice in reverb to make it passable. It was very sad, but I guess did prove their point to a degree. In that case, it was about the money as someone paid to get her into a studio when she really needed more practice time and a vocal coach.

Thanks for the push to look at spectral editing.


mskaye's picture

I bought an Ayre Codex recently (based on lots of reviews including Mr. Plaskin's) with the dream of finally enjoying the slew of hi-resolution downloads now available. For the most part it's been a worthwhile venture - but the differences are subtle. For example, the standard download of Bowie's BLACKSTAR sounds great. The hi rez sounds better - better definition in low level details and separation of instruments and layering. My ears are not amazing amazing so I can imagine that some of our listeners will hear more improvements than I. However, my Lector tube output cd player (redbook) smokes both. Just more body and a more analogue-ish warmth and truth. It might be an artifact of sorts but to my ears, it's more enjoyable. What it makes me want to do - and Michael L - can weigh in here - is invest in a Total DAC.

jim tavegia's picture

Many tube products do take away the sonic edge that often digital brings to the table. It is in its own way an EQ of sorts, which is not a bad thing. It is about playing music the way YOU like it to be played.

Now you just have to find a USB dac that does that same thing for you.

AJ Soundfield's picture

Michael, for affirming the validity of audio double blind testing, the defacto standard and entire basis of this AES study.
Many audiophiles dismiss controlled (blind) testing in favor of uncontrolled "long term viewing" and other such daydreaming. Good to see an online publication supporting rigorous short term testing like all the papers compiled for the meta analysis, free of self deception and biases. Bravo.

krabapple's picture

LOL. Nicely done. yes, I'm guessing everyone at Stereophile and TAS will be just fine with double blind tests, now that a 'small but important*' difference has been gleaned from data-mining and statistical meta-analysis of experimental DBT results?

(Dr, Reiss makes this claim for 'importantce', but that;s his subjective call, *NOT* something he could glean from the meta-analyslis)