Yes, anyone, even old people can hear 21 kHz (test attached)

[chervokas]

It happens because our ears and our brain are not doing the Fourier transform, even the fast one. We hear the shape of a wave in its entirety, not as a spectrum of harmonics.

This is actually not exactly an accurate premise. In fact our inner ears ARE doing something very much like a Fourier transform. When a soundwave reaches the fluid in the cochlea through the ossicular bones rapping on the cochlea's oval window, the basilar membrane of the cochlea responds according to place. A complex wave has multiple frequency components and each of those components most excites a particular point on the basilar membrane where the frequency displacement is greatest and different neurons are fired at different locations along the basilar membrane corresponding to these separate peaks. It's modelled as a series of auditory filters. In fact it's very similar to a Fourier transform. Our hearing does in fact work by breaking down a complex sound into component frequency elements that each fire different sets of neurons.

I don't know what's happening with the square wave example in the video, I haven't watched it. But any explanation of the effect based on the assumption that our ears and brain are not doing anything like a Fourier transform or are not separating the wave form into component frequencies and responding with different nerve impulses corresponding individually to the separate frequency components, is incorrect. A sine tone at a single frequency activates one place on the basilar membrane but a complex tone activates multiple places that drive separate neurons to fire in correspondence to each separate frequency component.

 

[Chrispy]

Just saw the title and skipped everything to just LMAO

 

[krabapple]

Perhaps the mods might warn the OP to stop littering ASR with conspiracist garbage posts?

 

[Keith_W]

Perhaps the mods might warn the OP to stop littering ASR with conspiracist garbage posts?

Nah, the whole "we can't hear above 20kHz thing" is a conspiracy of the Illuminati Lizard Men who don't want us to eavesdrop on their secret communications.

 

[theREALdotnet]

Any steady / periodic wave, yes, according to Fourier.

Well, a periodic signal can be constructed from (or deconstructed into) a discrete set of sine waves: the fundamental frequency and its harmonics (multiples), potentially all the way to infinity.

Any non-periodic signal can also be constructed from sine waves, but this requires frequencies in between the harmonics. In the extreme case, the Dirac impulse, you need all frequencies from zero to infinity.

 

[Barry_Sound]

Yeah the dying CD industry overlords will summon all their marketing firetruck teams to defend their silly 16/44 myths carefully planted over decades

Who are CD industry overlords? Can you name some?

CD is becoming a niche much like vinyl. I prefer CD for a couple reasons and they sound good enough to me. In fact my Philips CD player doesnt even support full 16 bit but rather 2x14 bit in dual mono.

 

[Hayabusa]

Who are CD industry overlords? Can you name some?

CD is becoming a niche much like vinyl. I prefer CD for a couple reasons and they sound good enough to me. In fact my Philips CD player doesnt even support full 16 bit but rather 2x14 bit in dual mono.

Its functionally a 16 bit dac made by 14 bits dacs and oversampling.

So many dacs nowadays create n-bit dacs with m-bit dacs were n>m

 

[Barry_Sound]

Its functionally a 16 bit dac made by 14 bits dacs and oversampling.

So many dacs nowadays create n-bit dacs with m-bit dacs were n>m

You are right for the stock model, mine is non oversampling.

 

[chervokas]

No, you cannot hear above 18 kHz, I cannot hear above even less than that. But we can still hear the shape of the waves at 7 kHz or even higher. And in order to reproduce anything other than a sine waveform at 7 kHz, the audio equipment requires 21 kHz. So we can hear if the audio equipment is cutting out above 20 kHz. So for the old people 48 kHz is absolutely necessary, for the young people a lot more than that.

Have you ever looked at the frequency response of the classic mics commonly uses in making our beloved music recordings? A Neuman U87? A Shure SM7B? An AKG C414? A Sennheiser 421? Most of them have frequency responses that fall off the table after 10kHz. The AKG will get you out to 20kHz before it starts sharply rolling off. The Sennheiser out to 15kHz before it starts sharply rolling off. Unless you're listening to recordings of music make with specialty broadband mics, you're likely not listening to recordings that have much if anything recorded above 10kHz, never mind above 20kHz, even if they're analog or high sample rate digital, at least not if they're being made with these classic and commonly preferred mics. If the recordings is being made with something like the broader band Earthworks mics that are flat out to 30k and even, depending on the model, 40kHz, then you'll have some of the HF info on a recording. But even something like a Brüel & Kjaer 4003 rolls of quickly after 20kHz. Is a recording medium that can reproduce frequencies at 40kHz necessary when the signal is recorded with mic that are rolling off steeply above 10kHz?

 

[chervokas]

This is actually not exactly an accurate premise. In fact our inner ears ARE doing something very much like a Fourier transform. When a soundwave reaches the fluid in the cochlea through the ossicular bones rapping on the cochlea's oval window, the basilar membrane of the cochlea responds according to place. A complex wave has multiple frequency components and each of those components most excites a particular point on the basilar membrane where the frequency displacement is greatest and different neurons are fired at different locations along the basilar membrane corresponding to these separate peaks. It's modelled as a series of auditory filters. In fact it's very similar to a Fourier transform. Our hearing does in fact work by breaking down a complex sound into component frequency elements that each fire different sets of neurons.

I don't know what's happening with the square wave example in the video, I haven't watched it. But any explanation of the effect based on the assumption that our ears and brain are not doing anything like a Fourier transform or are not separating the wave form into component frequencies and responding with different nerve impulses corresponding individually to the separate frequency components, is incorrect. A sine tone at a single frequency activates one place on the basilar membrane but a complex tone activates multiple places that drive separate neurons to fire in correspondence to each separate frequency component.

I should add to this that not only is the functioning of the cochlea tonotopic -- where different points along the length of the cochlea respond to different frequencies in a complex waveform, but the higher brain functioning level of our auditory cortex is similarly tonotopic. Susan Rogers, who is a former top recording engineer (famously as Prince's staff engineer) who went on to get a PhD in psychoacoustics and music cognition, and now is the director of the Berklee Music Perception and Cognition Laboratory, describes the auditory cortex as being laid out like a piano keyboard with different physical locations responding to different frequencies. So, it seems like our hearing at both lower and higher orders of processing, very much works by breaking down a complex wave into component frequency parts.

 

[dedobot]

In that region somewhere my ears are crossed with my inner voice so I'm not sure exactly what I hear !

 

[FrenchFan]

I am convinced that I can hear at least 25KHz

You only need a 2K watt tweeter 103db 1W/1m, set to max
and put your head in the tweeter.

And I am not young.

It is only a question of the attenuation curve of
your ear.

But then I have a headache

 

[pablolie]

My interest in these videos starts fading after 90 seconds, takes waaaaaay too long to establish the obvious.

 

[pablolie]

I am convinced that I can hear at least 25KHz

You only need a 2K watt tweeter 103db 1W/1m, set to max
and put your head in the tweeter.

And I am not young.

It is only a question of the attenuation curve of
your ear.

But then I have a headache

View attachment 431278

1. How much real content above 20kHz do you think is really in music, no matter if they record it at 192/24?
2. How often do you listen to music with your head to one tweeter?

 

[FrenchFan]

Is there any practical difference between "wave is made up of sines" and "wave can be reconstructed from sines"?


Can a wave with zero rise time physically exist? To my knowledge, no.

A wave may not be but a
physical phenomenon may be.

The transition between 2 quantum states is
not parameterized in time.

The passage from one to the other is instantaneous.

Either we are in one state or another,
but especially not between the 2.
where is the time in there?????

 

[pablolie]

...
Either we are in one state or another,
but especially not between the 2.
where is the time in there?????

Poor Schroedinger cat, whether alive or dead or both at the same time - I am sure it's not happy in a box... :-D

 

[FrenchFan]

1. How much real content above 20kHz do you think is really in music, no matter if they record it at 192/24?
2. How often do you listen to music with your head to one tweeter?

I didn't say I listened to music.

I said, with humor, what we could
hope to hear by extrapolating the
isothonic curves and putting Watts.

With a little power I hardly reach
14KHz, with my great age ....

In audio we can consider ourselves
old at 20 years old, and
that has been long since passed.

 

[pkane]

Poor Schroedinger cat, whether alive or dead or both at the same time - I am sure it's not happy in a box... :-D

A cat? Unhappy in a box??? Never heard of such a thing

 

[earlevel]

One is a 7 kHz sin wave and another is "artificially constructed" square wave composed of 7 kHz base and 21 kHz 3rd harmonic.

First, I want to clear this up: It's not an artificially constructed square wave, just 7k and 7k+21k. The 7k and 21k are the same amplitude, which is not true of a square waves, for which that third harmonic would be a third of the amplitude (-9.55 dB). In other words, that "third harmonic" is boosted by around 10 dB from what it would be if it were from a square wave.

Short explanation. No, almost nobody can hear a sin wave beyond 18 kHz, however anyone can tell a sin wave from a square wave at 7 kHz, even though the square wave is made of the 3rd 21 kHz harmonic.

I don't hear a difference. Maybe you hear a difference due to distortion, the second signal is 4 dB louder than the first.

It happens because our ears and our brain are not doing the Fourier transform, even the fast one. We hear the shape of a wave in its entirety, not as a spectrum of harmonics. It doesn't matter that the audio equipment requires 21 kHz to store the 7 kHz square wave - our brain is unrelated to the electrical circuits, it works differently.

Ludicrous, and would mean that even analog systems would sound different depending on bandwidth. There is no mechanism in the ear that makes the shape important.

And btw the simple proof that 16 bits are not enough is that dithered music sounds better than un-dithered. Dithering is destroying the music information. If we did not hear the step between two adjacent digits in 16 bits the dithering would not be necessary, as it's not necessary for 24 bit music.

This is completely wrong. Truncating measurements to any bit level means you have the potentially to be off by an average of half a least significant bit. But that half a bit is not a randomly distributed error, so it can make obvious patterns in simply cases (it still happens in complex audio cases, it's just gets harder to hear the patterns). So we dither to distribute the error to be similar to the error of analog, uncorrelated with the signal.

If we did not hear the step between two adjacent digits in 16 bits the dithering would not be necessary, as it's not necessary for 24 bit music.

That half LSB is much larger in 16-bit than 24-bit. That's the only difference. The error is so tiny in 24-bit that it's down nearly -140 dB, which is below even the quietest possible analog noise floor, and far below hearing. At 16-bit it's around -90 dB, which can be heard easily if your system is cranked. You're right that dither is not necessary for 24-bit music, I think I'm hated by a few over on Gearspace by saying it.

 

[AdamG]

You remind me of that guy Sunjam wreaking havoc here lol. I'm starting to think you're him in disguise using your backup account

View attachment 430699

Upsampling 16/44.1 collection a good idea?

Playing around with NOS DACs and DSD upsampling instead of speakers and room correction is like trying to move a mountain with a fork when you have an excavator right in front of you. Don’t think playing around with external upsampling and speakers and room correction have to be mutually...

www.audiosciencereview.com

Excellent catch. You are correct about who this is/was. Another Troll attempt but you guys are just too good. We the “Moderator” Team are humbled by your ability to catch out the Trolls and Puppet accounts so quickly. This account has been banned and we are locking the Troll Feeding Troft.

Your assistance is very much appreciated Sir.