Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Two Ultrasonic Frequencies Create an Audible Sound (Read 96204 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Do Frequencies above 20khz really matter?

Reply #50 –
Quote from: Ed Seedhouse on
It is certainly possible to produce audible frequencies by heterodyning two ultrasonic ones, and this was discussed in at least one popular science or audio magazine a couple of years back.  New Scientist was one if my memory serves me.  There are advantages, but also disadvantages, such as high distortion and poor LF.  It was, I believe, considered as a technique for steering audio to individuals in a crowd.

A quick google search gives this url: http://bcwildfire.ca/hprScripts/WildfireNe...Fire.asp?ID=339

And there is a web site selling such a product at: http://www.atcsd.com/site/content/view/13/104/

However what this has to do with high end audio escapes me...


Awesome, glad you found them, I heard about Trent Reznor using them to direct sounds at one person in a crowd to see what they did. Has anyone realised how these work? I'm pretty sure it's to do with the subject of this thread:

1) Regular audio recorded (speech, music, whatever, but for this example we'll say a sine at 1KHz
2) some sort of smart conversion where present frequencies are converted into ultrasonic, highly directional freqs (this would be complex converting speech or music, but in this instance, to get the 1KHz to beat from 2 other sines, the diff has to be just 1k, so it could be 21 and 22k, or it could be 56 and 57k like the examples on this page. Any ultrasonic freqs with a 1k diff.
3) The two sines would have to come from separate sources on the speaker, or a LPF at 20KHz or something so you don't hear the 'distortion' (ironic because the distortion eventually becomes the desired 1KHz sine) from the source, I think when I saw it on a TED talk it was like a square with loads of tiny tweeters, which makes sense for the complex sounds they will be trying to make directional
4) The sound is played and aimed at a person, say, 10-20 metres away, maybe loud

RESULT: No sound is audible to anyone off axis from the speaker, but the 'target' person who is being aimed at hears a 1KHz sine as they get the beating effect on their basilar membrane. It is audible only to them. Complex sounds could be transmitted with the same theory. Proof of concept!


And it doesn't matter that the original sines are ultrasonic/inaudible, there is no reason their energy can not create a beating effect just like any other waveform.

This is essentially how radio works, except going into more extreme frequency ranges, which means we need electrical equipment to decode it.

Quote from: boomtopper on
And one other problem I found with this and of course the most obvious. Is that those frequencies mentioned are way out of our hearing range . So that puts out the idea that beats being caused by our perception of the 2 sounds.


The above examaple might finally clear it up for you - I almost wondered if you were joking then, the ENTIRE thread has been about exactly this

Quote from: 2tec on
Quote from: PlayMusic on
No, I was saying that the physics of sound ...

Indeed, I was referring specifically to the physics of sound, rather than the psycho-acoustics of human hearing. Psyche, of course, in this case refers to mind, of which the ear is not generally considered to be a part of. Place theory, from what I understand, relates to how pitch is sensed by the inner ear. Psychoacoustics, of course, is the study of the psychological correlates of the physical parameters of acoustics.


Fair point, I see your angle. Place theory would come equally under psychoacoustics as it regard the perception of pitch,  but really it belongs to biology. You're right about the word psyche but generally psychoacoustic studies refer more to how the biology of the ear reacts than what happens in the mind (yet another area of psychoacoustic we know little about).


Quote from: Ron Jones on
Quote from: PlayMusic on
Oh, please say whether you read all that or not.

I read it until I started seeing namedrops, then I stopped reading. It was interesting up to that point, however.


Namedrops as in Rupert Neve? Purely because the thread was in response to a story recalled by Neve about Geoff Emmerick. That's what the original poster had linked to, it's a good little read if you haven't seen it.

Quote from: boomtopper on
What it bottles down to is that the harmonic that is generated is not a product of psychoacoustics but rather from the systems that the signal has passed through. This then leads back to the point mde by rpp3po made saying that this is an Unwanted Artifact


I wonder if you really read my posts, you're entitled not to but I've explained it carefully. The point in the system that the beating can happen varies - it could be a speaker, or it could be your basilar membrane in your ear. Where it happens depends on the conditions of excitation. So at the start of this post I describe how I think the ultra directional speakers work - in this case the beating occurs in the ear - but speakers or guitar strings do it as well. Just hit two keys next to each other on your keyboard and the speakers with make an LFO style beating.

Or try this test

1) Get two sine generators
2) Set one to, say, 15KHz and the other to 16KHz
3) Play them out of the same or both speakers, you will hear a third sound which is the beating is at 1KHz - because that is the difference between the tones. This is when the beating occurs ON THE SPEAKERS, but now try this...

1) Pan the 15KHz to the left speaker
2) Pan the 16KHz to the right speaker
3) If your speakers are in an exact triangle, and you place your ear exactly in line with the point where the on-axis direction of the tweeters cross, you should hear the 1KHz tone again - but this time, [i/it is happening in your head[/i]. The reaction between the two tones, the beating effect at 1KHz, is hapening on your basilar membrane (or, to be fair, it could be the air at the point where the two 15 and 16k sines meet right next to your ear. I dont know where it happens first, but it isn't the speakers this time, because the two sines are panned).

Okay, I have explained it to death, I'm sure some of you get it but if you don't the just buy this book by Perry R. Cook on psychoacoustics because you need to start from the basics.

Do Frequencies above 20khz really matter?

Reply #51 –
Quote
this is an Unwanted Artifact


BTW, the point is that people could use the phenomenon to their advantage. Things are only even unwanted if you don't want them, but the speaker company referred to on this page obviously found a way to make it useful. A lot of pioneers think like that.

Do Frequencies above 20khz really matter?

Reply #52 –



Each component of a sound gives rise to a displacement of the basilar membrane that is pictured. The displacement due to each individual compnent is spread to some extent to either side of the peak. Whether or not two components that are of similar amplitude and close together in frequency can be discriminated depends on the extent to which the basilar membrane displacments due to each of the 2 components are clearly separated or not. So as the picture shows the basilar membrane will only give rise to frequencies within the audible range. So how would frequencies in the 50khz do this?

Another i've found from this google book. it says:-

"it is generaly agreed that the beats are exceedingly difficult to hear for frequencies above 1000hz"

So frequencies beyoned the range of the basilar membrane how could they cause displacments?

Do Frequencies above 20khz really matter?

Reply #53 –
Quote
So how would frequencies in the 50khz do this?


First, we don't actually know for certain how useful the 50KHz range would be. It is just theoretical. If it makes things easier for you to believe, let's talk about the 20-25KHz range.

The point is, the beating effects are theoretically still audible, but only if the beating frequency is in the audible range of 20Hz-20KHz.

Quote
it is generaly agreed that the beats are exceedingly difficult to hear for frequencies above 1000hz


This is an interesting point, and I don't actually doubt it, but it doesn't mean the beating doesn't happen, because it still does. I do wonder what it then takes to make the beating audible, when the two sines causing the beating are over 1KHz.

BUT...

That speaker company has done it. To be sure we would need to see their white papers, but there is NO WAY of making any audible frequency ultra directional UNLESS it used ultra high treble or ultrasound. THe proof that it can happen is that the company has released a product that does it.

EDIT: Actually, after reading that book link, it does actually say:

Quote
Monaural beats can be heard over the entire frequency range


And it would seem that you keep confusing binaural beats with monaural beats.

I think binaural beats are when the two sine sources are coming from separate speakers, and monaural when they come from the same source.

I wonder...

If in their testing in that book link, they didn't consider the higher directionality of freqs over 1KHz. If the test subjects weren't precisely in line with both tweeters, then they wouldn't hear the beating on the basilar membrane, as described in my example test in an above post. Just a thought.

Do Frequencies above 20khz really matter?

Reply #54 –
Quote from: PlayMusic on
A lot of pioneers think like that.


I'd not necessarily call chasing 50 year old wind mills a pioneer enterprise.

You basically have not a single bit of data or indication that the continuously reiterated findings about the inaudibility of ultrasound could be wrong.

There is just this notion of 'psychoacoustics' - the new audiophile buzz word, always there with a warm bed for those, whose esoteric beliefs have been cornered by sober physical argument.

Don't get me wrong, there is a lot of interesting stuff going on the field of 'psychoacoustics'. But I wouldn't count using the term as a kind of super power to extend the properties of well known physical constraints to it. Fact is adults can't even tell apart incredibly loud amounts of 19 kHz signals from silence. But you deem it important to assume that there could be a kind of psycho-induced super hearing ability. Yes, there could also be little green men living at the back of the moon. Can't rule it out.  But there's only ever a finite set of questions that one can follow up upon in a lifetime, and just a "could" should not be enough to qualify.

All examples presented, about how ultrasound waves can form byproducts within the audible band, are perfectly explainable with well known properties of waves. No mysterious psycho antennas needed.

Do Frequencies above 20khz really matter?

Reply #55 –
Okay, lto clarify, every time I refer to psychoacoustics I am basically saying:

The act of sound being perceived by a human

My definition could be wrong, apologies if so. I'm not making out that it is a super power, but I generally refer to it as everything beyond the mathematical physics of air moving. Generally the books on psychoacoustics I have read appear to consider the subject like this.

For example, the original psychoacoustic effect, a shepard scale, is something that is reproduced on a speaker system, it is not something that exists in your head, but the side effect - the psychoacoustic effect - is that we percieve a constant rise or fall in pitch.

It's not a magic trick, it is something of an illusion. This is how I have seen the debate - that two inaudible frequencies above our range of hearing may (under certain conditions) give us the illusion of a new pitch. Yes, this new frequency/beating effect could originate from a speaker, but under certain conditions I have described in other posts, the beating effect could originate for the first time in the ear. So this would be considered a psychoacoustic effect, though you could also say it is biology or physics.

The reason the field of psychoacoustics currently lives in the biology of the ear and the physics of the air more than it does past the neurons the connect the ear to the brain, and the processes within the brain, is because we don't know much about it at all. Binaural beats is one area that does involve effect on audio perception actually within the brain, but that is a whole separate debate.

So whether you agree on my personal definition of psychoacoustics or not is irrelevant. You are welcome to correct me on the 'industry definition' of psychoacoustics, but it is not the subject of the debate here.

Maybe we should try and re-define the question and the point of debate, which I feel I have covered but this is spidering into all sorts of things.

Can the interaction between inaudible frequencies above our range of hearing cause beating effects which are audible?

My answer: Yes. I have explained why I think it is. If you disagree, state specifically why, and then try and be creative enough to come up with conditions it could happen before you say it is impossible. (for example, don't say 'normal speakers can't reproduce it' because there probably are speakers which can. Same rule for whatever other piece of kit. The only piece of kit that should remain standard in this discussion is that there is a human ear. We need to get to the bottom of whether a human ear would be able to perceive audible side effects of ultrasound. I think it is, but I do not know for certain, and the current arguments against it do not convince me for reasons I have made clear.

Do Frequencies above 20khz really matter?

Reply #56 –

Quote
All examples presented, about how ultrasound waves can form byproducts within the audible band, are perfectly explainable with well known properties of waves. No mysterious psycho antennas needed.


I want my friend Rich to see this and really understand that it is this simple.

BTW - who said anything about mysterious psycho-antennae?

Do Frequencies above 20khz really matter?

Reply #57 –
1. The well established phenomenon of perceptible difference tones for middle frequencies:

Audible difference signals (sometimes called Tartini tones) have been widely reported for frequencies in the mid-range of human hearing, e.g. if an amplifier  and loudspeaker system plays a 3000Hz sinewave, and a separate amplifier and speaker system, in the same room,  plays 2500Hz sinewave, a human listener in the room may hear a faint tone at 500Hz. This phenomenon has been explained as being due to non-linearity within the human ear.  [With my own hearing I can hear such difference signals if I am specifically listening for them.]

The phenomenon of the creation of an "undertone" by two higher tones, was reported by the Italian violonist Tartini in 1714.  The bottom of this webpage refers to Tartini tones:  http://www.phys.unsw.edu.au/jw/beats.html

2. A less reported/documented difference tone effect claimed to arise with ultrasonic frequencies:

In contrast to the often cited phenomenon of medium frequencies creating audible difference products, I have not encountered much website material at all to the effect that ultrasonic tones [such as 24KHz and 25KHz] produce an audible difference tone [such as 1KHz].  [In the case of my own hearing, I do not hear difference tones from independent tweeters operating at reasonable power levels at around 25KHz.  Perhaps very high energy levels would be required for an audible difference signal  ... ]

I also note that doubt is raised on this matter at  http://www.silcom.com/~aludwig/Nonlinear.htm :-
[blockquote]
Quote
A intriguing question is: if there are two frequencies above the limit of human hearing, say at 23 kHz and 24 kHz, would the non-linearity cause audible intermodulation products below 20 kHz? Everest states (page 55) that a difference tone of 1 kHz can be heard in this case. I did several tests to see if I could hear such a product, and I could not. Hartmann also states (page 514), without mentioning the frequencies, that a difference tone can be heard for tone levels greater than 50 dB SPL. I also could not hear any difference tones listening to 4 kHz and 5 kHz tones. I thought this indicated that the non-linearities of the ear were virtually inaudible. Not true! It turns out that a difference tone of 2f1-f2 is quite audible, for certain choices of frequencies.
[/blockquote]


3. An astonishing capability of the human brain to perceive beats as between what is presented to the left ear, and what is presented to the right ear; as reported on Wikipedia:

Something that has suprised me greatly is the reported ability of the brain to perceive beats when fed slightly different frequencies through headphones.  This astonishing phenomenon is described at  http://en.wikipedia.org/wiki/Binaural_beats  [I have not had the opportunity to test my own hearing for this; has anyone else?] :-
[blockquote]
Quote
The brain produces a phenomenon resulting in low-frequency pulsations in the loudness of a perceived sound when two tones at slightly different frequencies are presented separately, one to each of a subject's ears, using stereo headphones. A beating tone will be perceived, as if the two tones mixed naturally, out of the brain. The frequency of the tones must be below about 1,000 to 1,500 hertz for the beating to be heard. The difference between the two frequencies must be small (below about 30 Hz) for the effect to occur; otherwise, the two tones will be heard separately and no beat will be perceived.
[/blockquote]

Do Frequencies above 20khz really matter?

Reply #58 –
Quote from: PlayMusic on
We need to get to the bottom of whether a human ear would be able to perceive audible side effects of ultrasound. I think it is, but I do not know for certain, and the current arguments against it do not convince me for reasons I have made clear.


No, you have not made that clear. There are products which use ultrasound interference to direct and localize sound waves. But this is a solely and fully documented physical effect, not a psychoacoustical. The only thing psychoacoustical about it is that it exploits that ultrasonic frequencies are not audible. Because of that energy can be transported silently from the speakers to a target, where you create an interference with the signal of 1-n more speakers. This interference will emit directed and audible < 20 kHz sound waves. Those appear to your ear just as regular sound waves. The hypersonic waves used are just an invisible, physical transport layer.
No interaction between hypersonic content and ear/brain is needed to explain what's happening here. Its just plain physics.

Do Frequencies above 20khz really matter?

Reply #59 –
Quote from: rpp3po on
The only thing psychoacoustical about it is that it exploits that ultrasonic frequencies are not audible. Because of that energy can be transported silently from the speakers to a target, where you create an interference with the signal of 1-n more speakers. This interference will emit directed and audible < 20 kHz sound waves. Those appear to your ear just as regular sound waves. The hypersonic waves used are just an invisible physical transport layer.
Are very high sound pressures involved?  Why can this effect not be replicated by feeding a 25KHz signal to one tweeter, 24KHz to another, and standing a few metres away, straining to hear 1KHz?  I cannot hear 1KHz (or lower frequencies, e.g. 300Hz by using 24.3KHz and 24KHz) with such a simple setup.

Perhaps the setup required is an ultrasonic transducer being fed by say a 50KHz carrier, and with the carrier being amplitude modulated by the desired audio signal.  Is that the technique used?

Do Frequencies above 20khz really matter?

Reply #60 –
Quote from: PlayMusic on
Quote
All examples presented, about how ultrasound waves can form byproducts within the audible band, are perfectly explainable with well known properties of waves. No mysterious psycho antennas needed.


I want my friend Rich to see this and really understand that it is this simple.

BTW - who said anything about mysterious psycho-antennae?


And as I have explained I don't doubt this. What I doubt is that this effect is caused by our perception of sound. As I have explained the sideband frequency is caused by a non-linear system like a Distortion, Hard Limiting, Or some kind of Amplitude Modulation.

Do Frequencies above 20khz really matter?

Reply #61 –
PlayMusic : Hey I'm @Photosounder. I think the problem is that you confuse the beating with distortion on the beating. It seems that you think that a tone plus another tone creates a third tone at the difference frequency. It does not. If you add two tones, you still only get two tones, the only effect is that the result is a signal with an envelope which periodicity is the difference in frequency between the two tones.

What I'm saying is, if you play a 56 kHz tone with a 57 kHz tone, you won't necessarily hear a 1 kHz tone. You'll only hear it if there's distortions. There's still nothing but those two tones as long as there's no distortion, and nothing for a human ear to hear. Here's how the distortion thing does it : imagine you get some hard limiting on a sine. If you look at the difference between the hard limited sine and the original sine, then what you have left has a periodicity equal to the frequency of the sine, which means that for an ultrasonic tone all the distortions will stay ultrasonic.

But if you have beating frequencies, think of it like a single tone that has a sinusoidal envelope, then if you hard limit that well it will have more effect around the top of the envelope, and practically no effect on the lower parts. The result is that the difference between the distorted signal and the original signal is a pattern which periodicity (and therefore its lowest possible frequency component) corresponds to the periodicity of the envelope. When you play a two tones 1 kHz apart, if you can hear the two tones then you can hear the tones beating at a rate of 1 kHz (but no actual 1 kHz signal, just a signal varying with a periodicity of 1 kHz, although I'm not sure you can really perceive a beating of 1 kHz, it should rather sound like two separate tones, but the distinction between envelope and separate frequencies now that's actually a psychoacoustic question) (and if the tones are ultrasounds then you don't get to hear the beating), and if you have distortion (which can occur in the air, in the speakers, with your ear drum, or the microphone, or whatever else that might create such non-linearities) then you'll hear that.

So when you play two ultrasound tones and that you hear anything, it's coming from distortion, not directly from beating frequencies, you can have ultrasonic beating frequencies without hearing anything as long as there's no distortion.

Do Frequencies above 20khz really matter?

Reply #62 –
Quote from: rpp3po on
Quote from: PlayMusic on
We need to get to the bottom of whether a human ear would be able to perceive audible side effects of ultrasound. I think it is, but I do not know for certain, and the current arguments against it do not convince me for reasons I have made clear.


No, you have not made that clear. There are products which use ultrasound interference to direct and localize sound waves. But this is a solely and fully documented physical effect, not a psychoacoustical. The only thing psychoacoustical about it is that it exploits that ultrasonic frequencies are not audible. Because of that energy can be transported silently from the speakers to a target, where you create an interference with the signal of 1-n more speakers. This interference will emit directed and audible < 20 kHz sound waves. Those appear to your ear just as regular sound waves. The hypersonic waves used are just an invisible, physical transport layer.
No interaction between hypersonic content and ear/brain is needed to explain what's happening here. Its just plain physics.


That's fine, like I said, let's just move on from my definition of psychoacoustics I have which could be wrong.  The thing is, in that paragraph of yours, you confirm and agree with my entire argument in this thread - i particularly like the wording "The hypersonic waves used are just an invisible, physical transport layer", because it helps simplify the explanation for people who think that the entire concept is impossible purely because the original two tones are inaudible.

MLXXX, you've presented the best post of the thread with the summaries of the various phenomenon in discussion here.

To respond to a few points:

Quote
[In the case of my own hearing, I do not hear difference tones from tweeters operating at reasonable power levels at around 25KHz. Perhaps very high energy levels would be required for an audible difference signal ... ]


Good on you for trying it out, I wonder if your ears were exactly in line with the crossing point of the tweeters on your speakers - purely because frequencies this high would be ultra-directional, probably in a straight line out the front of the tweeter (assuming the speakers can actually represent freqs. this high correctly).

I would love to know if the original testers you refer to also kept this in mind during their testing.

As for binaural beats - I know little about these, and although the theory is similar to 'normal' (or apparently, monaural) beats, the difference is, all points of physical interaction between the waves are cancelled out - for exampl, there is no opportunity for speaker, air, or even your own basilar membrane to get traditional beats. The interesting thing is that the beating somehow occurs within the brain, after somehow summing the signals from your ears. The speculation on the effect binaural beats can have is rife, probably full of a lot of bollocks, but I would be interested to know if there were genuine positive or negative effects from binaural beats.

Oh, and this is one psychoacoustic definition we can't argue with - binaural beats are entirely to do with a phenomenon happening inside our brain, after our ears in the 'signal chain'.

Do Frequencies above 20khz really matter?

Reply #63 –
Quote from: A_SN on
PlayMusic : Hey I'm @Photosounder. I think the problem is that you confuse the beating with distortion on the beating. It seems that you think that a tone plus another tone creates a third tone at the difference frequency. It does not. If you add two tones, you still only get two tones, the only effect is that the result is a signal with an envelope which periodicity is the difference in frequency between the two tones.


Yes, I am glad you said that because I originally thought this (it acts much like an LFO affecting the amplitude of the original tones), but after reading this:

http://hyperphysics.phy-astr.gsu.edu/hbase.../subton.html#c1 (scroll down

Because, it would act only as amplitude modulation if the beating frequency were under, say, 20-30Hz. Same as if you just blended in a synthesised sine in your DAW (right?).

But if the beating frequency is, say, 1KHz, then is it audible as a third tone? It doesn't matter that it is still just acting as a fast amplitude modulator, because if you had blended in the 1KHz as a sine in the DAW anyway, it would be doing both things - affecting the amplitude of the existing two tones, and also be audible as a sine, because remember it is still something happening 1000/sec - it's getting hard to explain, I could be confused so if this is totally wrong the correct me.

EDIT: Also, the way you have described the distortions has given me a new angle on this, which I am grateful for, if anyone did describe that before then thank you but it has been explained better now. I will have to go for a think on this...

P.S. Can anyone expand on the theory on how those ultra-directional speakers work?

Do Frequencies above 20khz really matter?

Reply #64 –
Just been reading the post about the distortions through more, and yes this is a better way of explaining what I was talking about already. Should have made it easier for people to understand by not always referring to it as just 'beats', because I do understand that the beat is a distortion.

I still wonder about the question of whether ultrasonic freqs will cause this distortion in a human ear - I guess it will happen in the air by the ear first, something I was unsure about in an ealier post - but the final effect would be the same, right? So the plausibility of disguising audio in ultrasound is still entirely realistic. My descriptions haven't necessarily been wrong, I'm going to have to read it all back to see

Do Frequencies above 20khz really matter?

Reply #65 –
Quote from: PlayMusic on
Yes, I am glad you said that because I originally thought this (it acts much like an LFO affecting the amplitude of the original tones), but after reading this:

http://hyperphysics.phy-astr.gsu.edu/hbase.../subton.html#c1 (scroll down


I'm afraid the content of this page should be taken with a pinch of salt. I don't see how two flutes playing at the same time could sound like three, the description of multiphonics don't match what I could read on the Wikipedia article and as for the missing fundamental thing I don't think that looking at it is the way to go, it's more like seeing the regular spacing in the frequency domain, just like a cepstrum algorithm does. So yeah, I wouldn't rely on the claims made on that page.

Quote from: PlayMusic on
Because, it would act only as amplitude modulation if the beating frequency were under, say, 20-30Hz. Same as if you just blended in a synthesised sine in your DAW (right?).

Well, like I said, that's mainly a psychoacoustic question, because mathematically there's no real difference, but yeah, basically.

Quote from: PlayMusic on
But if the beating frequency is, say, 1KHz, then is it audible as a third tone?

Nope, all you hear is two tones. Think of it in the frequency domain if it helps, each since being a peak. Addition in the time domain is addition in the frequency domain. So when you add a peak at a frequency with another peak at another frequency, well you still only get two peaks. And as long as there's no distortion, that's all you'll hear. Now if you look at it in the time-frequency domain, you can either choose to see it as two separate tones, or something like one beating tone, or even perhaps two beating tones, it doesn't really make any difference.

Quote from: PlayMusic on
It doesn't matter that it is still just acting as a fast amplitude modulator, because if you had blended in the 1KHz as a sine in the DAW anyway, it would be doing both things - affecting the amplitude of the existing two tones, and also be audible as a sine, because remember it is still something happening 1000/sec - it's getting hard to explain, I could be confused so if this is totally wrong the correct me.

Not sure what you're saying here, you might be confusing modulation (multiplication) and addition. If you add then you just add, with no other result than a change in the sound's envelope (but no new frequency component). If you modulate, then it's a bit more complicated, but basically modulation (multiplication in the time domain) is convolution in the frequency domain. A sine in the time domain is represented by a peak at the positive frequency and another peak symmetric around 0 Hz (DC) for the negative frequency (real signals, that is, signals that are not made of complex numbers,  have positive and negative frequencies. Positive frequencies are what you know, and negative frequencies are just a mirror of that). So when you convolve two such things, well basically that's like replacing the each peak of the first sine with the two peaks of the second since, with the 0 Hz of the second since being place where the peaks of the first since were. So, if you multiply/modulate a 10 kHz sine with a 1 kHz sine, all you have in the end is a 9 kHz sine and a 11 kHz sine. So yeah, at the end of the day it's all about independent sines that mixes together produce envelopes, unless you don't want to think of them as envelopes and prefer to think of them as independent sines.

Quote from: PlayMusic on
P.S. Can anyone expand on the theory on how those ultra-directional speakers work?

Well, if your speaker has a certain size compared to the wavelength of the sound it emits, then the sound is emitted more or less directionally. The larger the speaker wrt the wavelength, the more directional. Since ultrasound have very tiny wavelength, you can get a very directional sound with a relatively modest speaker. As for how they transmit sound, well, they rely on ultrasounds getting distorted (the non-linearities) through the air or whatever, so they are generated in a way so that once they're distorted as a result of the distortion they produce the desired audible sound. I wish I knew how they compute their ultrasounds, and how it sounds like (seems to me like it would be hard to use distortions to create a clean sound reproduction), and I'm not too clear on where and how the distortions take place, I read about that a long time ago on Slashdot and don't remember too well. But yeah, it uses the distortions on what you can think of as the beat frequencies, although the ultrasonic signals must be nowhere as basic as that, so I guess you can rather think of it as some complicated ultrasonic signal getting distorted into producing "artifacts" that happen to be on a frequency range you can hear.

Do Frequencies above 20khz really matter?

Reply #66 –
Quote from: PlayMusic on
What is not strictly true? My engineering experience not being a science? I am referring to the studio craft of recording being science, in response to someone saying engineering is not a science. Much of it may be subjective, but it is mostly scientific methods and techniques which lead to the desired results. But I'm preaching to the converted now.

And Rich, I'm not claiming to be an authority on anything, this is a debate. I asked a lot of open questions in my first post on the first page.



The studio craft of recording is certainly *not* science -- unless REs are now employing scientific method.  Nor are science and engineering the same thing (nor is engineering a subfield of science). 

And again, it's eminently possible for great engineers, and great scientists, to make silly claims about things that are not directly in their field of expertise...as Linus Pauling did with nutrition, and Neve has done re: psychoacoustics.  It is an occupational hazard.

Btw, your attempt to sciencify Neve's/Emerick's/Fletcher's expertise by calling it mastery of the 'how to control voltage specifically for use in audio'  was wonderful.

Do Frequencies above 20khz really matter?

Reply #67 –
Quote from: boomtopper on
And here is more proof that our ears do not repond to frequencies beyond 20k.



No, it is evidence they they can, to a  point, if the SPL is high enough.

Before people repeat this 'there is no hearing beyond 20kHz ' mantra again, please note that the highest *documented* frequencies heard at 'normal 'levels are , IIRC, ~24 kHz, particularly in the very young.  *Adults* (at least in industrialized nations) cannot be expected to hear such frequencies at normal levels, though -- and most can be expected to have lost much of their sensitivity to frequencies a few kHz *below* 20 as well.

Do Frequencies above 20khz really matter?

Reply #68 –
Quote from: krabapple on
Before people repeat this 'there is no hearing beyond 20kHz ' mantra again, please note that the highest *documented* frequencies heard at 'normal 'levels are , IIRC, ~24 kHz, particularly in the very young.  *Adults* (at least in industrialized nations) cannot be expected to hear such frequencies at normal levels, though -- and most can be expected to have lost much of their sensitivity to frequencies a few kHz *below* 20 as well.


In this situation we are talking generally and not about individuals.

Do Frequencies above 20khz really matter?

Reply #69 –
Quote from: PlayMusic on
Quote
[In the case of my own hearing, I do not hear difference tones from independent tweeters operating at reasonable power levels at around 25KHz. Perhaps very high energy levels would be required for an audible difference signal ... ]


Good on you for trying it out, I wonder if your ears were exactly in line with the crossing point of the tweeters on your speakers - purely because frequencies this high would be ultra-directional, probably in a straight line out the front of the tweeter (assuming the speakers can actually represent freqs. this high correctly).

I did the experimenting in June 2008.  The tweeters were quite capable of operating at 25KHz (they were rated to 35KHz, and I verified they were working at 25KHz, using a microphone and oscilloscope).  I moved the speaker cabinets close together and aimed them slightly inwards, and listened at the same height as the tweeters and a metre or so away, where the tweeter axes intersected.  The 25KHz was not audible at all so I thought a 1KHz tone resulting from non-linearities in the air or my ears would have been easy to hear, given how sensitive human hearing is at 1KHz. But I heard nothing.

Quote from: A_SN on
Quote from: PlayMusic on
Yes, I am glad you said that because I originally thought this (it acts much like an LFO affecting the amplitude of the original tones), but after reading this:

http://hyperphysics.phy-astr.gsu.edu/hbase.../subton.html#c1 (scroll down


I'm afraid the content of this page should be taken with a pinch of salt. I don't see how two flutes playing at the same time could sound like three, ...
Audible Tartini tones from independently sounding instruments (such as a pair of flutes, or recorders) are faint.  Many if not most members of an audience might not notice the additional tones at all, even if hearing a duet at high volume.

Do Frequencies above 20khz really matter?

Reply #70 –
Quote from: A_SN on
I'm afraid the content of this page should be taken with a pinch of salt. I don't see how two flutes playing at the same time could sound like three, the description of multiphonics don't match what I could read on the Wikipedia article


I wouldn't take anything on a website as total gospel, but in fairness it is a website run by a university, whereas wikipedia is open to anyone to edit - the last poster is right that it is a useful and innovative system (and debatably the same accuracy as encyclopaedia britannica - see the great book wikinomics).



Quote
Quote from: PlayMusic on
But if the beating frequency is, say, 1KHz, then is it audible as a third tone?

Nope, all you hear is two tones. Think of it in the frequency domain if it helps, each since being a peak. Addition in the time domain is addition in the frequency domain. So when you add a peak at a frequency with another peak at another frequency, well you still only get two peaks. And as long as there's no distortion, that's all you'll hear. Now if you look at it in the time-frequency domain, you can either choose to see it as two separate tones, or something like one beating tone, or even perhaps two beating tones, it doesn't really make any difference.
Yes, but ultimately, when the two sines are played then there WILL be distortion somewhere along the line, am I right? So in the end, it depends if test subjects do hear a third tone. It is subjective, but that is probably why they are sometimes referred to as subjective tones.

Regarding possible confusion between addition and multiplication: Good point, I absolutely had not considered that.

Thanks for your views on that directional speaker system, you mirror some of my ideas on how it works, and I equally would love to know how they compute the ultrasonic sounds - some mad algorithm, and things must get super complex (maybe impossible) with translating complicated sounds. Maybe the speaker system doesn't even use this theory, but it is the only way I can see non-directional freqs being pinpointed at someone. As for where the distortions take place, my theory is that it happens in the ear of the target, but it is again confusing about how the air is actually reacting in this process.



Quote from: krabapple on
The studio craft of recording is certainly *not* science -- unless REs are now employing scientific method.  Nor are science and engineering the same thing (nor is engineering a subfield of science).


Really? I would consider that a fair argument, this isn't CERN, but I think the fairest way to describe audio engineering is to use scientific method to reach a subjective aim - for example, carefully getting mics in phase on a guitar amp, is something of a scientific method in my opinion. But to be anal about it, I suppose it might be fair to say that engineering was quite casual compared to high end, serious scientific investigations.

Quote
Btw, your attempt to sciencify Neve's/Emerick's/Fletcher's expertise by calling it mastery of the 'how to control voltage specifically for use in audio'  was wonderful.


LOL  okay that line sounds ridiculous, but I made an effort to point out how the whole story is ultimately anecdotal, and non-scientific, and I never tried to claim Fletcher's scientific ability, just that he was the one who had in turn recalled the story from Neve, and that he likely hadn't bullshitted and made up that Neve said it. Little unfair for you to call me on that when I emphasised this in two separate posts.

On the topic, do you really think the story deserves no extra credibility for Emmerick and Neve having heard it? Because the story is true - we don't know that they heard the ultra high frequency for sure, but that the resistors were genuinely broke shows there turned out to be something to their hunch about the dodgy channels.

In fact an online friend, whose opinion carries some weight for me (look him up), said this the other day: http://twitter.com/RonanCMurphy/status/3668645488

Just had a thought - We know they probably never heard the +3db 54KHz, for obvious biological reasons, but if not, then what else COULD it be?

I wonder - BIG maybe - if harmonics interacting with this peak were in turn creating slight peaks within the audible range? Kind of like the distortion/beats discussed in this thread? This could be really silly. Just putting it out there.

EDIT: just seen this:

Quote
I did the experimenting in June 2008. The tweeters were quite capable of operating at 25KHz (they were rated to 35KHz, and I verified they were working at 25KHz, using a microphone and oscilloscope). I moved the speaker cabinets close together and aimed them slightly inwards, and listened at the same height as the tweeters and a metre or so away, where the tweeter axes intersected. The 25KHz was not audible at all so I thought a 1KHz tone resulting from non-linearities in the air or my ears would have been easy to hear, given how sensitive human hearing is at 1KHz. But I heard nothing.


Big props for actually having tested this. I really thought that the theory would have been sound here, but this is a credible test. So now we ask, why not?

Do Frequencies above 20khz really matter?

Reply #71 –
Off-topic discussion over whether engineering is or is not science or a subfield of science has been moved here.

The topic at hand is whether frequencies above 20 kHz really matter.

Do Frequencies above 20khz really matter?

Reply #72 –
Quote from: boomtopper on
I found this interesting post on gearsluts ? The general gist is that if we add 2 very high frequency sines together they can create a beating effect that is audible to our ears.
Here is the quote of the example:-
Quote
the "beat" created between these two ultrasonic tones is equal to 1Khz (57K - 56K), which is a frequency right in the middle of the human hearing range and reproducible on pretty much any speaker...There is a mathematical propensity for harmonics to converge about three and a half octaves above the fundamental


I'd love to hear what people think of this?


If you ever actually set the stage for this sort of thing to happen, well it happens.  You may obtain an audible sound in the  midrange of the usual audible frequeny range, where the ear is most sensitive.

If I were doing this I'd pick the two tones 3-4 KHz apart, because this is where most people's ears are most sensitive according to Fletcher and Munson.

What you need is:

(1) Two strong tones or ther sounds with tonal components 500-8000 Hz apart in frequency, presumably both at very high frequencies outside of the audible range.

(2) Program material that doesn't mask sounds in the intended frequency range.

(3) A signal path that has signficiant nonlinearity and response at the chosen high frequencies.

One reason why people have such a hard time finding audible differences between music band-limited to 22 Khz, and music band-limited to like 50 or 100 KHz, using good audible systems to listen, is that items 1-3 above rarely if ever happen at the same time. 

Items 1 and 2 rarely happen in nature, due to the frquency characteristics of music.

Even if you get items 1 and 2 to coincide, parctical experience suggests that item 3 needs a signal path with 0.1% or more nonlinear distortion, and that is not a slam dunk with good modern equipment.



Do Frequencies above 20khz really matter?

Reply #73 –
Quote from: PlayMusic on
Yes, but ultimately, when the two sines are played then there WILL be distortion somewhere along the line, am I right? So in the end, it depends if test subjects do hear a third tone. It is subjective, but that is probably why they are sometimes referred to as subjective tones.

Regarding possible confusion between addition and multiplication: Good point, I absolutely had not considered that.


This distortion could occur in a couple of places. It could also occur in the power amp especially if it is other then Class A. Also Speakers are non linear too could be attributed to it. But I'm sure there could be other things too that I may not have thought of.

You have to imagine it a little like this:-
sin(x) + sin(y)

That would give you 2 harmonics in the frequency domain. This would be linear.

what if we were to do:-
sin(fc) * sin(fm)

Now this is ring modulation. And would therefore output 2 sidebands that are equal to:-

F(out) = F(fc + fm) + F(fc - fm)

Now this would be an example of a non-linear system as the output is not the sum of the ouput signals produced when the input signals are applied seperatly.
In the case of the start post something similar to this would have happened.

Do Frequencies above 20khz really matter?

Reply #74 –
Quote from: PlayMusic on
I really thought that the theory would have been sound here, but this is a credible test. So now we ask, why not?

The intermodulation tones come about due to distortion. Distortion comes about due to system non-linearities (and there's no such thing as a perfectly linear system). The intermodulation distortion we're talking about in this two-speaker setup occurs in the air and/or in the ear. If you want to hear it you either need to turn it up and/or move to (higher) frequencies where non-linearities are more pronounced.

There are commercial products that exploit this phenomenon (e.g. reference earlier in this thread) so there's no point arguing whether or not the phenomenon exists.