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

[Hayabusa]

I think his points have been mostly refuted through the thread, so what do you suggest to be still valid points (in the context of music reproduction in your home)?

Simple: we hear with our ears, so trying to produce what we can hear seems to be a good starting point for music reproduction.
In that way its good to have good statistics of what these limits actually are.
Me being 61 now I can not hear above 12k but I can hear (not feel) 16Hz quite well.
And is this subject limited to music? what about HT?

 

[Killingbeans]

The upper-limit of the fundamental tone... Ever heard of harmonics?

Yes, but look at the graph in post #2 (part of OP's evidence) and compare it to an FFT of any musical instrument.

The SPL needed to make those harmonics audible seems pretty ridiculous, right?

 

[Hayabusa]

Yes, but look at the graph in post #2 (part of OP's evidence) and compare it to an FFT of any musical instrument.

The SPL needed to make those harmonics audible seems pretty ridiculous, right?

An fft will reduce the level of short lived sounds. So its not a good way to measure in this case.
I am quite sure the levels of some percussion instruments will be much higher than shown in such FFT.

 

[Multicore]

Does anyone really care about "hearing" the high frequency noise from castanets and jangling keys? Sounds like a waste of time to even consider it.

Evidently someone does, hence this thread. But we have to upgrade for someone to scratch this itch.

That's why I think the onus is in their part to present evidence that convinces us.

 

[sigbergaudio]

Simple: we hear with our ears, so trying to produce what we can hear seems to be a good starting point for music reproduction.
In that way its good to have good statistics of what these limits actually are.
Me being 61 now I can not hear above 12k but I can hear (not feel) 16Hz quite well.
And is this subject limited to music? what about HT?

I agree that we should try to reproduce what we can hear. The debate seems to be about whether we can hear beyond 20-20,000hz.

I'm reasonably confident you can't hear 16hz, unless you play REALLY loud - but perhaps something in your room (or your sub) creates resonances that you can hear. That being said, most components aren't limited in the low end of the spectrum, and neither are the music formats. So the limiting factors are your speakers or subwoofers. So if you would like to reproduce 5-20,000hz, you can.

 

[JeremyFife]

Interesting thread, possibly also a bit theoretical and ultimately pointless ...

A call to adopt a higher standard (not the right term) across the entire audio industry - recording to playback - to increase the available frequency bandwidth. Suggestion is to go wider than 20Hz-20kHz and perhaps for that to be 10-25k (higher, 50k?).
I get that from an engineering excellence point of view.

Consensus that we cant actually hear anything above 20kHz with the exception of (my own words) some freaky teenagers. Consensus that the SPL of very high frequencies is too low to be audible anyway. I'm 60 ... I can't hear anything vaguely near that

Statement that real instruments (castanets) can produce sounds well above 20kHz, obviously electronic instruments can create any frequency. That sound is available to us in a live setting, so it should be available when we play back a recording.
But we can't hear that.

Comment from @restorer-john that audio equipment should be capable of reproducing a wider frequency range just as a matter of decent engineering, and that measurement standards should reflect that.
Can't really comment on that, although it does seem as if there are no real commonly agreed standards for testing amplifiers and there should be (may have misunderstood that part)

No issue with the idea of pursuing engineering excellence.
No interest in changing anything just to play a sound I can't hear.

I'll continue to follow this thread ... but I can't seem to care about it

If I want to experience high frequencies (that I can't hear) I'll go to my local Scottish flamenco bar (there isn't one, unfortunately - I'd love that)

 

[Killingbeans]

An fft will reduce the level of short lived sounds. So its not a good way to measure in this case.
I am quite sure the levels of some percussion instruments will be much higher than shown in such FFT.

Sure, but we can still assume that the ultrasonic harmonics will be 50dB or more below the fundamental?

Those percussion instruments need to be really insanly loud. Hopefully the tensor tympani muscle will do its job and stop you from "enjoying" the harmonics.

 

[ElJaimito]

I have not read all the posts. I was a "golden-eared" (not my puffery - as described by others) studio engineer in my youth, and I could @ 24 years detect 22 kHz easily and somewhat above with effort, at normal studio listening levels (around 89 dbC at the listening position). You can't hear even 20 kHz, it is a presence not a sound, IME. And it is not in the least musical. Note also that at these wavelengths, it is utterly inconsistent as a presence - a movement of the ear by a mere mm or two will cause cancellation and addition making it wobble and wave as if in a comb filter (which effectively it is).

I am now 73 and my hearing limit is around 12 kHz. I'm not complaining - I was almost deafened in one ear as a result of an imbecile DJ playing loud music through the stack as I was changing a bass driver inside the box while on tour with a major in Dallas in 1983, but after ten years or so it got better (or the other ear deteriorated to match!). And my wife @ 78 is almost stone deaf, so I am NOT complaining.

The interesting thing is I can still detect the addition of "air" in a mix at frequencies I know I cannot hear (say around 6dB at 18 kHz) and for a long while I wondered why. AFAICT it comes down to the ear still being able to assess the attack envelope of a transient. Discovering this cheered me up and enabled me to get back to recording and mixing without overly worrying about deficient HF balance (though I also use tools to assess the frequencies I can no longer hear directly). I narrate all this in case it's of any comfort to the high percentage of AS readers for whom 20kHz is a distant memory, as with me!

But as for the idea of widening the bandwidth limits, I regard this as extremely unwise. GA Briggs (Wharfedale founder) had a motto (imagine it posited in a broad Yorkshire accent), "the wider you open the door, the more the muck blows in". I heartily agree. In any studio in any city worldwide, unless you have a fully isolated base structure (and even then), a look at the output of a microphone with decent response down to 10 Hz or so will show massive underlying muck from cars, trucks, trains and planes, especially helicopters, all creating a groundswell of nasty mess that has no relation to anything musical, that is best dealt with by a sharp low cut filter below the lowest frequency of interest in the music (it's a problem if recording an organ with long pipes!).

At the top end, percussion and piano (also percussion!) have significant harmonics up to and above 20kHz, but I challenge anyone in a true A/B test to detect the difference between a 44.1 kHz and a 96 kHz recording (with the commensurate sharp anti-aliasing filters limiting the harmonics). Adding "air" in either case still works even with the limitation to 20-odd kHz in the 44.1 case.

If you doubt my words check with other reliable sources (Dan Worrall is a good start for this).

And let's recall that FM radio with its around 15 kHz or sometimes 13 kHz top limit was regarded as a good benchmark for best possible reproduction for many years, certainly in the UK, before CD quality came along. And don't get me started on vinyl, where almost all perceived high frequencies are actually saturation-style distortion in place of the long missing finer grooves.

I often use a low pass filter on instruments that don't really have much high harmonic content - as well as a high-pass if called for - just to clean everything up before eq'ing and fx'ing.

YMMV but I really very much doubt it. 20 to 20 is a good choice IMMHO.

 

[RayDunzl]

Me, I like a DC-Daylight design or at least something flat out to 100kHz in preamps, power amps and integrateds.

I got it here...

Preamp

Amps

They're at least 20 years old now.

Wouldn't want to figure out what's wrong if they broke, too complicated inside.

Of course, my "hearing" spec is way way below any of that foolishness above.

 

[lashto]

You're not supposed to be happy with that.

feels good to know I'm not the only one

Compact disc was always specified from the earliest working groups from DC to 20kHz. Most high quality players were flat from a few Hz and up from the 1980s on. A large value DC blocking/coupling capacitor was the only reason they didn't go to DC. (and the problems that would have brought)

A 1989 TOTL Sony:
View attachment 376252

I did read a few interesting stories from the CD era. With the 70s tech they had, those guys actually did an amazing job. 20-20 was a big wow-achievement for their time. But it's 50 years later, should be fine to ask for a bit more.

Amplifiers, preamplifiers, vinyl cartridges and phono stages were often 50kHz out to hundreds of kilohertz in many cases. Some power amplifiers were specified out to several megahertz...

But measurements need to have a specified bandwidth. The greater the bandwidth, the more noise and artifacts. 20Hz-20kHz is about as tight as you can specify and still be classified as somewhat HiFi.

Back in the day, amplifiers with low power outputs and inexpensive builds, would specify 40 or 50Hz to 15kHz responses.

Me, I like a DC-Daylight design or at least something flat out to 100kHz in preamps, power amps and integrateds. They just sound better than bandwidth challenged designs- to me.

The 2024 recording & playback chain looks very "last-century" to me:

• most mics cannot even cover 20-20. By the way, does anyone know a mic which can record the full range of 24/192? Or at least 1Hz-60kHz? And if not, is anyone trying? (and not even talking about the pretty awful mic linearity & distortion.)
• ADC/DACs seem fine 20-20 (finally and thanks ASR for that!). But distortion does not look very good at 20Hz and ~noone specs/measures below. Plus the deltasigma stuff above 20kHz: i.e. copious amounts of "sound I did not pay for". Some filter it, some couldn't care less. Cause they know: "it's inaudible", "don't worry", "it's fine" .. is it really?! Or the only "fine" is in your cost-cutting reports?
• And more generally, this delta-sigma stuff is not exactly perfect and also 50+ years old .. how about something newer/better? E.g.: clean 1-100 or more, does not generate crappy extra 'sounds' and without the modulated noise-floor.
• Amps used to look pretty good and I highly welcomed the push for "clean to mHz". But nowadays we have the classD groupies saying that "clean to only 15kHz" is just fine. Not sure who asked for that kind of 'progress'. Not me and no thanks. And btw, my cat would like to have a word with some of you, too.
• Speakers rarely can cover even 20-20. And those who do, cost kidneys. Also have serious distortion below 100Hz. But again we are told "that's fine". At least in this case there are studies to support that view (i.e. ears are less sensitive in that area). But looks like everyone took those studies as code for "let's stop trying for anything better". Please don't. And how about tweeters linear to 50-100 kHz, does anyone have one? And I do not mean extra supertweeters that cost another kidney, just 'normal' ones.
• More generally, almost all audio equipment is only specified for 20-20 and very few check outside that. It's quite hard to keep any ultra/infra sound in the recordings, even if you want to. So studios mostly just get rid of that. And then they sell us 24/192 music. And we are "very happy" with that.

Oh short: please give us clean recordings 1Hz-60kHz. Or 1-100 or even more ... why not ?!
And stop putting "sound I did not pay for" in that 1-100 area and telling us that "it's fine". Or even better, just stop generating that kind of grrr anywhere.
And give us gear that is 100% clean 1-100. Or even 1Hz-1GHz ... why not?!

/rant...

P.S.
IANAEE and anyone is welcome to correct any technical mistakes I may have made

 

[sigbergaudio]

• Amps used to look pretty good and I highly welcomed the push for "clean to mHz". But nowadays we have the classD groupies saying that "clean to only 15kHz" is just fine. Not sure who asked for that kind of 'progress'. No me and no thanks. And btw, my cat would like to have a word with some of you, too.

Not sure I understand your references to Class D amplification. To not cherry pick the newest of the new, here is NCore, the review is from 2018, I believe this design went into production in 2012 or 2013. Is this not clean enough or linear enough?

 

[MRC01]

... The interesting thing is I can still detect the addition of "air" in a mix at frequencies I know I cannot hear (say around 6dB at 18 kHz) and for a long while I wondered why. AFAICT it comes down to the ear still being able to assess the attack envelope of a transient. ...

FWIW, I discovered the same while ABX testing.

... But as for the idea of widening the bandwidth limits, I regard this as extremely unwise. GA Briggs (Wharfedale founder) had a motto (imagine it posited in a broad Yorkshire accent), "the wider you open the door, the more the muck blows in". I heartily agree.

Yep, I've lost count of how many high-def 96 and 192k recordings I've seen with high levels of ultrasonic noise - clearly not anything in the music, but artifacts introduced during recording, editing & mastering.

 

[popej]

The interesting thing is I can still detect the addition of "air" in a mix at frequencies I know I cannot hear (say around 6dB at 18 kHz) and for a long while I wondered why.

What about intermodulation distortion of the speaker? Are they in audible range?

 

[lashto]

You haven't looked around enough. There are such recordings. They are empty experiments. There's thousands of them. Try https://www.audiocheck.net/blindtests_frequency.php?frq=32

thanks for those. I actually asked for Music. Say, full orchestra recorded at least 5-50

As to gear, my current headphones have these specs: HEADPHONE FREQUENCY RESPONSE 5-40,000 Hz. The DAC resolves 32/384. That should cover the range you claim doesn't exist in audio by a factor of 4 at the least.

And did you or anyone test that those HPs actually work to 40kHz?
DACs are ~fine .. or better said, have the smallest issues.

I have not read all the posts. I was a "golden-eared" (not my puffery - as described by others) studio engineer in my youth, and I could @ 24 years detect 22 kHz easily and somewhat above with effort, at normal studio listening levels (around 89 dbC at the listening position). You can't hear even 20 kHz, it is a presence not a sound, IME. And is is not in the least musical. Note also that at these wavelengths, it is utterly inconsistent as a presence - a movement of the ear by a mere mm or two will cause cancellation and addition making it wobble and wave as if in a comb filter (which effectively it is).

I am now 73 and my hearing limit is around 12 kHz. I'm not complaining - I was almost deafened in one ear as a result of an imbecile DJ playing loud music through the stack as I was changing a bass driver inside the box while on tour with a major in Dallas in 1983, but after ten years or so it got better (or the other ear deteriorated to match!). And my wife @ 78 is almost stone deaf, so I am NOT complaining.

12kHz at 73 is very good and above average. Especially for someone who worked in that field. I'd be happy with that too. We'll see (currently still at 15-16kHz). There is also some interesting research into "rebuilding" ear-hairs/etc...

The interesting thing is I can still detect the addition of "air" in a mix at frequencies I know I cannot hear (say around 6dB at 18 kHz) and for a long while I wondered why. AFAICT it comes down to the ear still being able to assess the attack envelope of a transient. Discovering this cheered me up and enabled me to get back to recording and mixing without overly worrying about deficient HF balance (though I also use tools to assess the frequencies I can no longer hear directly). I narrate all this in case it's of any comfort to the high percentage of AS readers for whom 20kHz is a distant memory, as with me!

Ears are not the only sound receptors. The info may also be coming some other way (e.g. bone conduction is good in that range).

But as for the idea of widening the bandwidth limits, I regard this as extremely unwise. GA Briggs (Wharfedale founder) had a motto (imagine it posited in a broad Yorkshire accent), "the wider you open the door, the more the muck blows in". I heartily agree. In any studio in any city worldwide, unless you have a fully isolated base structure (and even then), a look at the output of a microphone with decent response down to 10 Hz or so will show massive underlying muck from cars, trucks, trains and planes, especially helicopters, all creating a groundswell of nasty mess that has no relation to anything musical, that is best dealt with by a sharp low cut filter below the lowest frequency of interest in the music (it's a problem if recording an organ with long pipes!).

good point about the outside noise. But I guess not many would mind a studio in the woods/mountains/etc.
And of course bigger bandwidth=bigger problems. How about an attitude like "bring em on" ?!

At the top end, percussion and piano (also percussion!) have significant harmonics up to and above 20kHz, but I challenge anyone in a true A/B test to detect the difference between a 44.1 kHz and a 96 kHz recording (with the commensurate sharp anti-aliasing filters limiting the harmonics). Adding "air" in either case still works even with the limitation to 20-odd kHz in the 44.1 case.

ABX is not the point. At least not mine. I want those infra/ultra sonics even if they are not 'audible'.

YMMV but I really very much doubt it. 20 to 20 is a good choice IMMHO.

It's perfectly fine for anyone to say "20-20 is more than enough for me".
OTOH, if someone says "20-20 ougtta be enough for everything", they should show some proof. Still waiting for someone to post some ...

 

[sigbergaudio]

OTOH, if someone says "20-20 ougtta be enough for everything", they should show some proof. Still waiting for someone to post some ...

I would rather say you've seen lots of proof in this thread. What form of proof would you prefer / would you find satisfactory? I am asking so neither of us have to waste time on evidence you would dismiss off hand.

 

[SIY]

should be fine to ask for a bit more.

I still see no actual justification for why. I mean, if it's important to you for some esthetic reason, feel free to do it, but wondering why others aren't wasting their time chasing something that has no sonic reward is a bit odd.

 

[solderdude]

12kHz at 73 is very good and above average. Especially for someone who worked in that field. I'd be happy with that too. We'll see (currently still at 15-16kHz). There is also some interesting research into "rebuilding" ear-hairs/etc...

Then why do you think CD quality 0Hz - 20kHz is not enough ?

Ears are not the only sound receptors. The info may also be coming some other way (e.g. bone conduction is good in that range).

How well do you think bone conduction works when listening to speakers/headphones ?
Besides also bone conduction still uses the ears, it just bypasses the earcanal + drum + ear-bones.

And did you or anyone test that those HPs actually work to 40kHz?

I have seen plenty of headphones reaching at least 30kHz (-10dB)

Usually headphone FR response is given at either -10dB or -20dB unless specifically mentioned.

 

[JustJones]

 

[DonR]

I got it here...

Preamp
View attachment 376314

Amps
View attachment 376315

They're at least 20 years old now.

Wouldn't want to figure out what's wrong if they broke, too complicated inside.

Of course, my "hearing" spec is way way below any of that foolishness above.

Covers it all... from whales to bats.

 

[lashto]

Not sure I understand your references to Class D amplification. To not cherry pick the newest of the new, here is NCore, the review is from 2018, I believe this design went into production in 2012 or 2013. Is this not clean enough or linear enough?

View attachment 376318

FR barely covers 20kHz so already no dice.
And afterwards, there are copious amounts of "sound I did not pay for". I do not want to be 'reasurred' that it does not matter (particularly not by people who show zero proof). I just want clean tech, something that does not generate any crap.

Also this: distortion goes up with frequency. This is actually one of the latest/best Ds .. and even here, it is a bit of a stretch to say "good enough to 15kHz". Many others start going up at 1kHz already.

Also phase: not measured anymore, but last time I saw a graph it had pretty serious phase deviations above ~10kHz

And about all that, we get the same good old tune: don't worry, be happy, it's not audible, it's fine, it's tralala... some even say "it's perfect"!
Me not happy.