Shouldn't we upgrade the 20-20 audible range ?!

[krabapple]

Glad you seem to enjoy mr Reiss and his work a bit more than this-japanese-fellow.
Would you care to cite some of those "experts who do not"? Particlarly the one study which already debunked mr Reiss' meta .. and why not, everything else beyond-20-20.

You'd have found some of them in comments on those AES articles, but AES alas no longer hosts them consistently...though they do still crop up .
You could look up JJs thoughts on this topic. You could look up *Amir's* thoughts on this topic. .

I can't expect you to be aware , but there was something of a political war within AES for years over whether or not to embrace 'hi rez audio'. Bob Stuart's 2004 JAES paper promoted hi rez (on the basis of not very much at all, as regard sample rates) prompted a critical letter to the editor whose cosignees included heavyweights lik Stanely Lipshitz , and ultimate stimulated the famous Meyer & Moran study of audibility of commercial hi rez releases downsampled to CD rate, also published in JAES. Stuart, of course, founded Meridian which sold pricey DVDA/SACD players , developed a lossless package for surround audio (MLP, though DVD-A was hacked in short order), and much later, the hilariously useless and now all but dead MQA. All the while he kept referencing the same dodgy evidence you do along with appeals to 'what recording engineers report' from sighted listening.

Vicki Melchior, who is on the AES Hi Rez committee, summarized the 2019 state of ultrasound evidence in his JAES review thus:

ULTRASONIC FREQUENCIES
A frequent misconception is that high data sampling rates assume the audibility of frequencies above 20 kHz. Scientific study of ultrasonic frequencies continues, for example their role in bone conductive pathways and in the effectsof airborne sound on brain waves measured by EEG. However, a role in normal audio listening has been rejected sinceabout the late 1990s (due to lack of evidence) in favor of the ideas discussed in Secs. 4.4 and 4.3. Individuals ableto hear above 20 kHz might experience subjective differences compared to those who don’t, however an ability to differentiate high resolution and CD data is reported, informally, by individuals whose measured limits are well below 20 kHz.

The 20 kHz limits of human auditory perception for pure tones transmitted by an airborne path were established fromwork on equal loudness contours [25]. The physiological basis for these limits, including the role of the outer andmiddle ear and cochlear processing, are summarized in [47,68, 69]. Studies of bone-conducted ultrasound and EEGmeasurements are discussed in an audio context in [47,65]. There also have been formal listening tests specifically addressing whether test tones and their harmonics in theultrasound region are audible. Except in cases with veryhigh amplitude stimuli [70, 71] where thresholds can bemeasured for some listeners, literature studies have shownnegative results [72]. A recent study aimed primarily at theaudibility of intermodulation distortion (IMD) also con-firmed the non-audibility of the ultrasonic tone pairs usedto generate the IMD [73]

Does that read like a ringing call to arms to you? And this is from five years ago.

NB that her Conclusion section simply states that people can hear 'high resolution' versus CD audio. But this is a bit of a feint. She writes in her earlier 'High Resolution: Why? section that "This section considers four proposals for sonic differences, the third and especially fourth of which are the ones broadly accepted as likely." The four proposals are 4.1 Ultrasonic frequencies (reprinted above), 4.2 Hardware (no solid evidence), 4.3 Dynamic Range (sure; it's not controversial that 16 vs 24bits can be audible, esp in headphone listening, if correct dither isn't used for downconverting), and 4.4 Filtering and the Time Domain (antialiasing/imaging filters can affect time domain performance enough to be audible -- though no evidence is offered). IOW, ultrasound is NOT his likely culprit.

 

[krabapple]

Quibble- with dither (which is universal), bit depth sets the noise floor, not "the smallest recordable difference in level." I did a demo here a couple years ago with a cheap Behringer interface showing tones well below the LSB.

Perhaps I should have written the smallest audible signal above noise floor? Coming from the idea that bit depth sets the number of discrete amplitude values available for a sample.

But for sure, and it's been sad to see over the years the number of times hi rez proponents talk about audible difference betweem 16 and 24bits, when it turns out they are really referring to *undithered* audio.

 

[SIY]

Perhaps I should have written the smallest audible signal above noise floor?

Depending on the signal, you can often hear it below the noise floor. So extra bit depth for consumer playback is even more useless.

 

[ElJaimito]

I had used to be an (associate) AES member, when there was still a British section. I went to a lot of lectures, and read a load of papers, many of which were very useful, and I still have the imprints. However, as with everything else in audio, the UK AES drifted a long way from being a wholly scientifically-based organisation over the early 2000's and post the demise of the British section, I could no longer think of any justification for maintaining membership while no longer working in audio, a position emphasised by a hugely increased membership fee. It's easy to overlook the fact that the AES, while still somewhat reputable, is a trade organisation, whose task is to foster the commercial interests of its supporting members. I used to do the APRS too, back in the day, then didn't bother (it now appears moribund). Magazines that used to host reliable information turned into sounding boards for manufacturer press releases. Come back Studio Sound! O tempora, o mores! (Below 20K, that is).

 

[lashto]

You'd have found some of them in comments on those AES articles, but AES alas no longer hosts them consistently...though they do still crop up .
You could look up JJs thoughts on this topic. You could look up *Amir's* thoughts on this topic. .

or instead of dropping names you can actually post some links and/or quotes ...

Victor Melchior, who is on the AES Hi Rez committee, summarized the 2019 state of ultrasound evidence in his JAES review thus:
...
Does that read like a ringing call to arms to you?

This AES meta-study quote .. wow, quite some progress (from the initial nay-nay arguments like "there is no proof" and "all science is rotten").
Same as this thread, your meta-study actually makes a case for CD/20-20-is-not-enough and recommends high-resolution audio. It even cites & acknowledges as relevant science ~everything in this thread: the hypersonic effect, the non-audible effects through bone/tissue/etc conduction and so on..
A highly welcome and quite unexpected amount of nay-nay-progress in just a few weeks. At this pace, by Christmas we will all sing kumbaya at 50kHz

.. his Conclusion section simply states that people can hear 'high resolution' versus CD audio.

so, it is not only the effects of inaudible that I was mentioning, high-resolution is even audible. Lots of progress indeed ...

But this is a bit of a feint. He writes in his earlier 'High Resolution: Why? section that "This section considers four proposals for sonic differences, the third and especially fourth of which are the ones broadly accepted as likely." The four proposals are 4.1 Ultrasonic frequencies (reprinted above), 4.2 Hardware (no solid evidence), 4.3 Dynamic Range (sure; it's not controversial that 16 vs 24bits can be audible, esp in headphone listening, if correct dither isn't used for downconverting), and 4.4 Filtering and the Time Domain (antialiasing/imaging filters can affect time domain performance enough to be audible -- though no evidence is offered). IOW, ultrasound is NOT his likely culprit.

I did not want the filter arguments in this thread: it is debated somewhere else on ASR by others who think that the 20-20 world is just fine. Maybe you can link the study in there.
Although it looks like your meta-study offers "no evidence" (about the filters and its main conclusion) and even less evidence about ultrasonics or the other themes. And it ends-up with a fairly long "conclusion" section which sounds a lot like a non-conclusion.
So, we have a researcher with med-science background and a single audio-related paper, who investigated quite little and found quite nothing. A bit dissapointing I would say... but still a big step from nay-nay-for-the-sake-of-nay-nay

 

[lashto]

What about priorities and the order of things.

Not so easy to comment on personal choices like priorities.

And the practicall limitations of ones personal Hifi ? I would put fixing perfect fidelity to 40 kHz between place 1023 and place 1040 of things to fix for my personal audio nirvana ? There so much else not perfect or basically wrong to fix before this becomes important that you never get to it .

And I would put revolutionizing the entire audio chain at no1.
I want the 20-20-upgrade because it is a foundation-change, everything in audio (i.e. everything between mics and speakers, including software) is designed based on that range. Even a small tweak can change the whole industry: including your wishes and everything that anyone else cares about.

Generally, I am not a fan of small steps and would prefer going for a bigbang like this :

.. an ideal high-performance [audio] chain would be one whose “errors” .. are equivalent to those introduced by sound traveling a short distance through air.

(only like the spirit of that change, not MQA itself which seems to be just an expensive non-solution)

Otherwise, quite happy with the current system. I have both the funds and the wish for a big upgrade/change .. but why would I buy more of the same 70's tech in a different box?!

And are you 17 years old and female ? If not join the rest of us with limited hf hearing .

I am already there with the 'limited'. But being a 17 years female sounds interesting ... is there a thread on reincarnation?

 

[krabapple]

or instead of dropping names you can actually post some links and/or quotes ...

For fairly obvious reasons, these people don't spend their time running highly contrived experiments to show you can make someone hear 30kHz if you blast it into their faces.

Every reference you cite is something I've seen (and probably compiled myself) in years past. There's a place called Hydrogenaudio that hashed this stuff out in real time. My job isn't to educate you when all you want to see is what you already believe.

Here's a thought (hardly novel): If hearing 'hi rez' sample rates was something that had much bearing on listening, you wouldn't need a meta-analysis of 20 years of papers to detect it, would you?.

This AES meta-study quote ..

It's not a 'meta study'. It's a literature review. They are different things.

wow, quite some progress (from the initial nay-nay arguments like "there is no proof" and "all science is rotten").

You don't do nuance, do you?

Same as this thread, your meta-study actually makes a case for CD/20-20-is-not-enough and recommends high-resolution audio. It even cites & acknowledges as relevant science ~everything in this thread: the hypersonic effect, the non-audible effects through bone/tissue/etc conduction and so on..
A highly welcome and quite unexpected amount of nay-nay-progress in just a few weeks. At this pace, by Christmas we will all sing kumbaya at 50kHz

You can't hear ultrasound, but your refusal to read seems a more dire issue. Again:

Melchior, reviewing the evidence, does NOT credit 'ultrasound' as something heard in normal listening:
"However, a role in normal audio listening has been rejected sinceabout the late 1990s (due to lack of evidence) in favor of the ideas discussed in Secs. 4.4 and 4.3."
The ideas discussed in Sec 4.3 and 4.4 are NOT about hearing ultrasound. 4.3 is about bit depths. 4.4 is about artifacts from poor filtering.


Your whole schtick has been about how we 'need' ultarsound.


Melchior simply reiterates what we here have been saying: there's no significant 'there', there, as regards ultrasound. If you call reiterating that 'highly welcome', that's progress, I guess.

so, it is not only the effects of inaudible that I was mentioning, high-resolution is even audible. Lots of progress indeed ...

Again, not ultrasound. Not bandwidth. But rather, the uncontroversial idea that 24 bits *can* be audible versus 16....though in the normal run of proper production, and normal listening to playback, it shouldn't be. Melchior should have mentioned that part too.

No progress whatsoever for *your* hobby horse.

I did not want the filter arguments in this thread: it is debated somewhere else on ASR by others who think it's all just fine in the 20-20 world. Maybe you can link the study in there.
Although it looks like your metastudy meta-study offers "no evidence" (about the filters and its main conclusion) and even less evidence about ultrasonics or the other themes. And ends-up with a fairly long "conclusion" section which sounds a lot like a non-conclusion.
So, we have a researcher with med-science background and a single audio-related paper, who investigated quite little and found quite nothing. A bit dissapointing I would say...

Too bad about what you ''want' in this thread. I don't want warmed-over noise like your threads cluttering up a forum about audio science, but I'm not a moderator and we don't have a strict TOS.

Melchior is an AES colleague of Dr. Reiss; they both serve on AES's Technical Committee on High Resolution (she's currently its Chair). Too bad her conclusions from a 2019 review of the 'thousands' of papers, came up short of your dreams about ultrasound. The issues I raised about Reiss seem far more pertinent to this than the weak sauce you're ladling here.

And please add to your already long list of education materials about how science is done: understanding the difference between a meta-analysis and a literature review.

 

[lashto]

You don't do nuance, do you?

I wonder which nuance would that be?! Here's a few 'nuances' that I chose to ignore up to now...

The most nay-nay posts which present ~zero evidence but have a very annoyed-agressive tone. Like someone stole something from them. Wonder what that might be.
The 'nuance' of your very 'subtle' suggestion that threads like this should be closed/censored/etc .. because you do not agree with them.
The 'nuance' of your paper's author, which you initially quoted as the non-existing Victor Melchior and 'miraculously' morphed into the opposite sex Vicky (after I linked the right reference).
The 'nuance' of you & others tirelessly bad-mouthing the entire global science landscape and the AES science in particular .. and then using an AES paper as some indubitable evidence.
The 'nuance' of you criticizing your own quoted paper for lack of evidence and calling it "just a literature review" ... but still using it as your best proof.
The 'nuance' of chosing a paper by the AES member with the least audio experience/background as your proof.
The 'nuance' of using a paper that concludes "we need to go higres" to somehow disprove a thread that just asks "how about going highres"...

Or maybe the 'nuance' on how some us senator who had allergies last century is 'proof' that contemporary japanese science is wrong.
I must admit, that 'nuance' of political FUD was very funny and quite priceless!

 

[lashto]

Much better to forget those 'nuances' and actually talk about your best proof: the AES "literature review".
We skip the chapters you disagree with (i.e. most of them) and only look at the one you ~agree with. That best-ever chapter on Ultrasonics.

It concludes that ultrasonics don't matter based on the main argument "lack of evidence during the 90's". An 'excellent' argument in 2019 when the paper was written. Good thing they did not use "lack of science during middle ages" .
But indeed there was "lack of evidence" during 90's: as in no evidence of any kind, neither pro nor against ultrasonics. Even the very few ultrasonic studies of the 90's were 'abandoned' (like the celebrity Meyer&Moran that noone cites anymore) because they used improper music/signals with little to no ultrasonics and/or improper playback equipment and/or improper test subjects and/or...

That Ultrasonics chapter also states with utmost confidence that:

20 kHz limits of human auditory perception for pure tones transmitted by an airborne path were established from work on equal loudness contours [25]

The main evidence, the cite25, is a paper from 1956! Which did not even investigate audible limits but the loudness contours. And it clearly states that its results only "cover a range of frequency of from 25 to 15 000 c/s". A bit mindboggling this 15kHz 'evidence' on 20kHz.

Another evidence cited for "20 kHz is the limit": cite70, an ultrasonics/audibility study which found that 6 out of 15 subjects did hear ultrasonics:

• 6 heard 22 kHz
• 4 heard 24 kHz
• 20-22kHz were heard at a level just below 80dB, and 24kHz just below 90dB ("figure3" in the paper).

And from an ~unrelated 1956 paper and a study that says 20-24kHz is clearly audible, the "literature review" concludes that "20 kHz is the limit".
/s For sure it is, the indubitable "literature" proved it


P.S.
if someone still thinks that this "literature review" is any good, please post it in the other ASR thread where many argue that there is nothing wrong with the CD/20-20 filters. Tell everyone in that thread that you found the best paper that proves how those filters are audible/bad/wrong. I'll bring popcorn!
P.S.'2
the initial post included a confusion between a dustbin 'super' tweeter of the 70's pioneer pt-100 and "pioneer pt-r100" which suposedly goes to 120kHz. The perils of one small letter...

 

[krabapple]

I wonder which nuance would that be?! Here's a few 'nuances' that I chose to ignore up to now...

The most nay-nay posts which present ~zero evidence but have a very annoyed-agressive tone. Like someone stole something from them. Wonder what that might be.

Fact is, like the people who say there's a 'side' saying 'ALL AMPS SOUND THE SAME', you resorted to lazy imprecision. Ignoring the important details.

The 'nuance' of your very 'subtle' suggestion that threads like this should be closed/censored/etc .. because you do not agree with them.

Where on earth did I claim there was anything 'nuanced' about my disdain for your noise? I hope it's loud and clear.

The 'nuance' of your paper's author, which you initially quoted as the non-existing Victor Melchior and 'miraculously' morphed into the opposite sex Vicky (after I linked the right reference).

Um, no, don't flatter yourself. I posted the link to Melchior (2019) in the first place. And her public name is Vicki, with an 'i', not a 'y' btw.

The 'nuance' of you & others tirelessly bad-mouthing the entire global science landscape and the AES science in particular .. and then using an AES paper as some indubitable evidence.

You keep using 'you and others' as a hand-waving convenience here. That's the opposite of nuance.

I didn't question 'AES science in particular'. More properly you would say I question particular work that has been published by the AES.

I gave historical background on the AES and hi rez that you were completely ignorant of (like much else in this area ) and in the course of that referenced objections to JAES's promotion of it, and work like Meyer and Moran, which were published in --- JAES.

I linked to the minutes of the AES High Rez committee and anyone with eyes can read the drive to promote 'hi rez' there. Fine, that's it's remit, the AES isn't an academic research department after all, it is an organ of a profession. Melchior , who's been a member of that committee for years and now chairs it, promotes high rez in her literature review (not 'meta-study') -- I repeat , she comes out for high rez, not surprisingly ....
....but she does not attribute its benefits to 'ultrasound'.
That's the key point here.
I'm keeping laser focused on your dogged 'ultrasound' cheerleading because you seem to be trying to shift focus away. from it. It is the crux of the matter.

And I pointed out that a reading of the minutes shows it setting a new course that appears to downplay claims about needing wider bandwidth at home, in favor of defining 'high resolution' in less highly debatable ways.

I aslo asked why Reiss doesn't seem to quote his own meta-analysis (published in JAES), which came sorta out of nowhere, but gosh, could being on the High Rez Committee possibly have been an impetus? I question whether his include/exclude decisions for his meta-analysis were correct ; that's a legit and not uncommon question about those sorts of analyses. It's a hazard any author of such papers faces because there is always a level of arbitrariness involved: where to make the cutoffs? I can tell you haven't a clue about any of that.

None of this constitutes me criticizing 'AES science' generally. Obviously.

Others can and have expressed the view that the AES and its publication JAES have become less rigorously scientific over the years. Perhaps a literature review, or a meta-analysis , is in order.

The 'nuance' of you criticizing your own quoted paper for lack of evidence and calling it "just a literature review" ... but still using it as your best proof.

I certainly didn't write 'just a literature review'. You're fantasizing. And I certainly wasn't 'criticizing' it by identifying its type correctly as a literature review. Floyd Toole's masterful book is largely a literature review itself.

So this looks like just a biased, mistaken reading on your part.

The 'nuance' of chosing a paper by the AES member with the least audio experience/background as your proof.

Ah, now who is questioning the messenger? You know nothing of Melchior's 'audio experience/background' except the precis you read. And you wouldn't be qualified to judge it in any case. You've demonstrated that.

The 'nuance' of using a paper that concludes "we need to go higres" to somehow disprove a thread that just asks "how about going highres"...

You keep ignoring the important nuance that I already pointed out -- Melchior's not advocating a need for more bandwidth. Which is what YOU have been blathering about.

Or maybe the 'nuance' on how some us senator who had allergies last century is 'proof' that contemporary japanese science is wrong.

'Some US Senator' was a rather important guy in his day, but of course you're ignorant of that history too . He basically got an entire new Institute of the National Institutes of Health created. Which was quite a thing. And it caused quite a stir among biomedical researchers. And it still has many critics, even though its changed its name to be more palatable to the reality-based scientific community. But you were ignorant of all of this. Until I told you about it.

It was offered as an example of pseudoscience that can self-perpetuate even in otherwise respectable settings. Historical examples include parapsychology labs in respected universities.

IOW, the mere existence of an official-sounding 'Center' or 'Institute' for XXXX Science is not sufficient to conclude it's doing good Science. You're the one who seems to think that the existence of something like that in Japan, is.

Your 'contemporary Japanese science' is a few decades old now. And it's all come from three labs there, basically. It appears not to have set the audio world on fire. Your case for a 'need' for 'ultrasound' in home playback hasn't been made by it, which might be the reason.

Rather than braying at your own jokes, you should be thanking me for educating you.

 

[krabapple]

Much better to forget those 'nuances' and actually talk about your best proof: the AES "literature review".
We skip the chapters you disagree with (i.e. most of them) and only look at the one you ~agree with. That best-ever chapter on Ultrasonics.

It concludes that ultrasonics don't matter based on the main argument "lack of evidence during the 90's". An 'excellent' argument in 2019 when the paper was written. Good thing they did not use "lack of science during middle ages" .

So much nonsense you're spinning from this one sentence:

However, a role in normal audio listening has been rejected since about the late 1990s (due to lack of evidence) in favor of the ideas discussed in Secs. 4.4 and 4.3.

But indeed there was "lack of evidence" during 90's: as in no evidence of any kind, neither pro nor against ultrasonics. Even the very few ultrasonic studies of the 90's were 'abandoned' (like the celebrity Meyer&Moran that noone cites anymore) because they used improper music/signals with little to no ultrasonics and/or improper playback equipment and/or improper test subjects and/or...

This canard again.

Meyer & Moran were responding to the widespread mythology at the time that DVDAs and SACDs sounded better than CDs simply by virtue of them being 'hi rez'. Such claims were made indiscriminately, without regard to whether there was actual >22kHz content on the discs. They used commercial releases that had been lauded for exactly those reasons. Some of them can be shown to have little or no 'ultrasound' content. Others did have such content. Downsampled to CD rate, results did not support audible difference.

It was entirely disingenuous for high rez cheerleaders to have objected to M&M's samples, when they themeselves hardly bothered to differentiate between 'true' HD releases and CD-rate released as HD, when they raved on and on about the virtues of HD. (there's also the peculiar idea that 48kHz SR isn't really 'high rez')

Of course, the real reason DVDA or SACD sounded different from (and thus potentially better than) a CD release is when the mastering is demonstrably different. True then, true now.

That Ultrasonics chapter also states with utmost confidence that:

The main evidence, the cite25, is a paper from 1956! Which did not even investigate audible limits but the loudness contours. And it clearly states that its results only "cover a range of frequency of from 25 to 15 000 c/s". A bit mindboggling this 15kHz 'evidence' on 20kHz.

It is a historical review. Citing old work first doesn't make it the 'main evidence'.

Another evidence cited for "20 kHz is the limit": cite70, an ultrasonics/audibility study which found that 6 out of 15 subjects did hear ultrasonics:

If you think you are catching the author out here, you've read lazily yet again . See below

Lets' dispose of your biased readings here. Here's the complete section on 'Ultrasound', and a bit from the section before. I'm highlting parts that refute your implications that Melchoir ignored work past 1999. My notes in bracketed italics.

From the Intro to 'High Resolution: Why?' (section 4)

High resolution formatting does not guarantee a perception of transparency, but music professionals with access to first generation data have widely reported subjectively better sound

[the usual sighted listening stuff]

and a meta-analysis of previously published listening tests comparing high resolution to CD found a clear, though small, audible difference that significantly increased when the listening tests included standard training [67]. This section considers four proposals for sonic differences, the thirdand especially fourth of which are the ones broadly acceptedas likely

[This cite is the Reiss (2016) meta-analysis. Melchior is relying on it to assert there is something to HD, but what is it? She reviews 1. Ultrasound, 2. Dynamic Range (bit depth), 3. Hardware , and 4. Filtering/Time Domain as factorss in turn]

4.1. Ultrasonic Frequencies. A frequent misconception is that high data sampling rates assume the audibility of frequencies above 20 kHz. Scientific study of ultrasonic frequencies continues, for example their role in bone conductive pathways and in the effectsof airborne sound on brain waves measured by EEG. However, a role in normal audio listening has been rejected since about the late 1990s (due to lack of evidence) in favor ofthe ideas discussed in Secs. 4.4 and 4.3. Individuals able to hear above 20 kHz might experience subjective differences compared to those who don’t, however an ability to differentiate high resolution and CD data is reported, informally, by individuals whose measured limits are well below20 kHz.

[The implication here is that, whatever those individuals who literally cannot hear above 20kHz are using to 'differentiate' CDs from HD-- in sighted listening -- it's not likely due to actually hearing frequencies above 20Hz. The rare few who can, 'might experience subjective differences', but see the next paragraph about experimental conditions under which this can happen]

The 20 kHz limits of human auditory perception for puretones transmitted by an airborne path were established fromwork on equal loudness contours [25]. The physiologicalbasis for these limits, including the role of the outer andmiddle ear and cochlear processing, are summarized in [47,68, 69]. Studies of bone-conducted ultrasound and EEG measurements are discussed in an audio context in [47,65].

[These are reviews/advocacy papers by fervent hi-rez cheerleading AES member Bob Stuart, from 2004 and 2019, which do cite Japanese work from the 21st century. Citing other reviews is a common space-saver in literature reviews]

There also have been formal listening tests specificallyaddressing whether test tones and their harmonics in theultrasound region are audible. Except in cases with very high amplitude stimuli [70, 71] where thresholds can bemeasured for some listeners, literature studies have shown negative results [72]. A recent study aimed primarily at the audibility of intermodulation distortion (IMD) also confirmed the non-audibility of the ultrasonic tone pairs used to generate the IMD [73] [Cite 70, Ashihara's paper, is cited for the very reason you claim she ignores: that it shows ultrasound can be heard---but only under highly contrived circumstances. She then cites a 2019 paper , with Bob Stuart as co-author, that *failed* to support audibility of ultrasound, if its effects are due to IMD]

 

[kemmler3D]

And I would put revolutionizing the entire audio chain at no1.

OK, a question for you now.

What are you willing to give up in the 20-20 range to get it?

"Just make the tech better" isn't going to happen, not in cars, not in phones, not in audio. No free lunches.

A $1000 speaker can't just gain features and remain $1000. This is simple reality regardless of what can be heard, or can't.

So how much are you willing to pay to add the next 30Khz on top? Or what performance in the 20-20 range would you give up to get it?

 

[Robin L]

OK, a question for you now.

What are you willing to give up in the 20-20 range to get it?

"Just make the tech better" isn't going to happen, not in cars, not in phones, not in audio. No free lunches.

A $1000 speaker can't just gain features and remain $1000. This is simple reality regardless of what can be heard, or can't.

So how much are you willing to pay to add the next 30Khz on top? Or what performance in the 20-20 range would you give up to get it?

I know I must have posted this before - Hi Fi News posts reviews of Hi-Rez downloads, along with measurements of ultrasonic content:

Hi-Res Downloads, April 2024

Carducci Quartet Shostakovich: String Quartets Nos. 9 & 15 (96kHz/24-bit, WAV)

www.hifinews.com

So it's not as if such stuff doesn't exist for those who are interested. The question is: how many people are interested and how does this advance the state of the art? I would suggest that for most adult audiophiles, ultrasonic content is inaudible. By the time someone can afford that sort of kit, most of the top octave is already shaved off. Lastho's hobbyhorse doesn't connect to reality. If ultrasonic content was all that important, there would be a lot more written about it and a lot more content available. But there isn't and there isn't and for good reason: the overwhelming majority of audiophiles couldn't hear it if they wanted to, and the overwhelming majority of audiophiles don't care. Linearity within the known audible range is still difficult to achieve, it's still the primary issue to address.

BTW: There's only one recording here that unambiguously contains ultrasonic content (i.e., not an artifact of analog tape noise) and that recording has that signal -80 db below the average level within the normal audible range.

 

[solderdude]

I want the 20-20-upgrade because it is a foundation-change

Where do you get the 20Hz from ? It is not specified in CD specs,
In red book audio only the upper limit is specified (22.05kHz) and not the lower limit (which is basically DC).
RB CDA is a fixed format. If you want more there are other digital formats that have higher bit depth and sample rates (frequency response)
You can keep wining about 44.1kHz/16 bit as a 'not enough' audio format. That format is fixed. You can't nor to need 'change' that as there already are other standards that far exceed 20kHz.
If you want more there is already plenty of that around. For mortal human beings that 20-20kHz (at what cutoff points ?) is more than enough.

 

[lashto]

Where do you get the 20Hz from ? It is not specified in CD specs,
In red book audio only the upper limit is specified (22.05kHz) and not the lower limit (which is basically DC).
RB CDA is a fixed format. If you want more there are other digital formats that have higher bit depth and sample rates (frequency response)
You can keep wining about 44.1kHz/16 bit as a 'not enough' audio format. That format is fixed. You can't nor to need 'change' that as there already are other standards that far exceed 20kHz.
If you want more there is already plenty of that around. For mortal human beings that 20-20kHz (at what cutoff points ?) is more than enough.

I cannot keep up with all the posts so not sure who/where linked the 20Hz limit to the CD format.
I only use it because 'everyone' calls 20-20 full-range nowadays .. although it is not the full-range of anything. And while there is no 20Hz limit in the CD format, there is very little infrasonic on commercial CDs. No idea if some precise numbers/stats exist, but most recordings seems to 'end' somewhere in the 30-40Hz range.

And pretty much the same about the recording formats, no idea who/where complained about those.
16/44 is what it is. Should be possible to AI-extend those old recordings and nowadays we have 32/768 or more, no problems in the formats area. Guess because it is much easier to push the limits with software .. and because that is done by the IT industry which, in terms of progress/evolution, is the opposite of the stuck-in-the-70s audio industry.
Does not look like the audio industry will ever provide the hardware to record & play even 24/96 or 24/192 at full capacity. 22-23 bits is a pretty hard limit of the current hardware and may be a forever limit (at least according to current state of physics). And mics/speakers to reliably deal with even 50kHz are one in a million and cost 3 kidneys.

anyway, looks like people needed and liked your clarifications so many thanks.

 

[lashto]

OK, a question for you now.

What are you willing to give up in the 20-20 range to get it?

Nothing
And I am actually willing to pay less for a 1-120 speaker. Same as my 2023 4K TV did cost less than my (way worse) 80s PAL TV.

"Just make the tech better" isn't going to happen, not in cars, not in phones, not in audio. No free lunches.

But it does happen. Almost everywhere but audio.
Computer hardware/software is the poster child of "just make tech better". And to stay with the vision/TVs comparison: a 2024 TV can do xx more pixels than a 70s one, each on of them being xx times better. And it does that for xx less money.

A $1000 speaker can't just gain features and remain $1000. This is simple reality regardless of what can be heard, or can't.

So how much are you willing to pay to add the next 30Khz on top? Or what performance in the 20-20 range would you give up to get it?

You are of course right, that is not possible with the current tech. Same as 8K TVs are not exactly feasible with 70s cathode rays.
That is why audio needs a "revolution". Of the same magnitude as moving from cathode-rays to OLEDs. And OLEDs also do PAL resolution much better than a PAL TV of the 70s. So no, there is notthing to give up in the 20-20 range. On the contrary, that range will be much improved too..

Guess we are kinda back to one of the initial questions I posted: what keeps audio from evolving like the other domains?

 

[JustJones]

I haven't read all of this thread but when did human hearing evolve so that 20hz-20khz isn't good enough for our survival. Yes, babies, pre teens, some teenagers and very few outliers into their 20's hear a little beyond, 22-23khz? so can anyone point to some biological studies showing human hearing evolving ? How many more generations before we can hear ultrasonic dog whistles?

 

[lashto]

View attachment 377623


As can be seen the most 'energy' in a cymbal crash (purple line) is in the 3kHz-10kHz range. 20kHz is already 40dB below the 3-6kHz range... and 70dB SPL while the crash itself is a whopping 110dB SPL over a wide bandwidth. Really only experienced by a drummer not wearing IEMS or hearing protection.

Where Lashto thinks the 'magic' might be is not where he thinks it might be.

Is that your own measurement? Or maybe you can link the source !?

There is very little research in that area, almost none except that "life above 20kHz" study. Which is a sort of one-and-only-standard, quoted everywhere. On crash cymbals that one states things like "40% of the sound energy is above 20kHz" and "still going at our 103 kHz measurement limit". Seems like they may have recorded a bit more cymbal ultras than your graph shows..
Anyway, if you know any other sources please link em ...

 

[Hayabusa]

How do people enjoy bass below 20hz? It can't be heard. Just like we can't hear ultrasonics.

I say reproducing the audible range is what matters. I consider the capability to do that to be perfect.

Needlessly reproducing sounds outside our audible range is pointless and wasteful. I do not understand why anyone would think it was important or a worthwhile goal.

I purchase and use my audio equipment to listen to music that I can hear and enjoy. I am not missing anything by leaving out very low frequency rumbles or ultrasonic screeching. Those frequencies are nothing more than garbage noise to be filtered out.

I can hear bass to 16Hz, it takes good subs that have sufficient low distortion to test this, do you ever tested this yourself?

 

[lashto]

I haven't read all of this thread but when did human hearing evolve so that 20hz-20khz isn't good enough for our survival. Yes, babies, pre teens, some teenagers and very few outliers into their 20's hear a little beyond, 22-23khz? so can anyone point to some biological studies showing human hearing evolving ? How many more generations before we can hear ultrasonic dog whistles?

20-20 may be enough for survival, no idea. Although any extra Hz should be useful there, too.

The two infra & ultra sound studies in the OP clearly show an "evolution" of the audible limits since ~1920 when serious measurements started. Most researchers say it's because the recording/playback hardware is better, the tests are better organized etc... sounds like the most plausible explanation to me too. But I did read a few short mentions like "maybe our ears are evolving too". Do not remember the source, will post if I (re)find it.