A question for the speaker manufacturers on here....

[krazy_olie]

My take is that, if it was there before, then it should still be there after, whether I can hear it or not

Certainly a logical approach, though speaking from an engineers point of view it can sometimes cause more problems than it solves. For a very loose example the hardware that I work on doesn't process things you can't see and massively improves efficiency. In a real world scenario you put resources in to producing things that are irrelevant you inevitably pull resources away from something that is relevant.

This isn't the best example though of that as I stated above I don't think the benefit of a 40khz response has anything to do with hearing 40khz... unless you are a dog.

I'm not an acoustics expert it is quite possible that frequencies above 20khz are used in some way, certainly you won't hear a 40khz tone but I don't know enough to say that it has no effect at all.

Also re the mp3 comment above. I think it has more to do with the fact that the basis on which mp3 makes it's decisions about what to throw away are not perfect and an approximation. Naturally the more bits you give it to play with the better it sounds. A cd will not have anything higher than 22khz in it so mp3 certainly won't top that.

 

[proffski]

First one should learn something about complex waveform reproduction. Starting say with Fast Fourier transform...

Then consider the data here: http://www.cco.caltech.edu/~boyk/spectra/spectra.htm as well as here http://www.ocf.berkeley.edu/~ashon/audio/Ultrasonics.htm, they mirror each other. Most of us will not hear much above 15KHz anyway, the frequencies in question may be many db down, yet they do contribute to the final sound if left unadulterated by digital processing.

 

[CnoEvil]

I can't speak for AEJim or Dean but I'll have a go from my own point of view. To do this, I have to go back to a few basics so if I'm "teaching to suck eggs", apologies in advance. Here's how I see it. bonkers as it may be.

First, ( get yer heads round this one guys ) we don't hear anything!!. We feel it. When a loudspeaker makes a forward stroke, it compresses the air in the listening space which goes down your outer ear and moves the round window i think on the inner eardrum. This is full of fluid which is compressed accordingly and on a bone in the inner ear are a load of tiny hairs.

These hairs are excited by the compression of the fluid and sends an impulse to the brain which is what we perceive as hearing. Unfortunately, from the time we are born, these little hairs are dying off hence when we get to middle age, our hearing tends to 'roll off' due to this.

Frank Harvey is right when saying Harmonics. Your hearing may be rolling off at around 14 to 15kHz but the upper harmonic of that is around 30K and if not addressed, hurts like mad! ( Shrieky treble ).

To prove this, I actually designed a loudspeaker that had a flat response to 20K. Played music through it and nearly blew me head off with the HF yet it measured fine. so I re-designed the treble section by ear and trial and error. Eventually, I got it right. Not just to my ears but to other people as well. It was superb in every way. I then measured it. At 20K, it was nearly 9dB down!.

What I realised was that I had designed the loudspeaker to mimic the response curve of the ear. 9dB at 20K may sound a lot but at the upper harmonic of 40K, the ear can tolerate it. What was this loudspeaker I hear you ask??.

Oooh sorry. It was the Arcaydis AK3. 5 stars twice.

Sorry if I've put you all to sleep guys.

Thank you for your contribution....it sheds light on some of the practical reasoning for implementing such technology....from a speaker designer's perspective.

Cheers

Cno

 

[CnoEvil]

Mate, if you want to understand it, head along to your local uni library and start some research followed by reading. There's heaps of research papers out there covering the topic of hearing, ultrasonics etc.

You won't find an answer on a hifi internet forum, nor, to be truthful, on the web unless unis have publisher their stuff on it. If you do find stuff on the web, it's 99% sure to be someone's whacky opinion backed up by zilch. Even something as basic as cooking, food and nutrition gets mangled by too many crazed idiots masquerading as experts.

Also, you're being given opinions here, and they don't count for anything, because they're just that and may be based on what Fred told you down the local after a couple of tinnies. What you need to do is suss out the facts

AL, I've a feeling that I would find, that one long (boring), highly technical piece of research, would only end up contradicting another....and without the technical background, I would find myself more confused than ever (not difficult to do!).

 

[CnoEvil]

Possible answers as far as I can see.

We can somehow sense the ultrasonic frequencies outside of our hearing range - doubtful, but?

Higher frequencies have an effect on lower frequencies in harmonics - Hugely doubtful, sounds like made up gibberish.

Record companies and speaker manufacturers are in cahoots to sell HD music to help squeeze the last drop of money out of the market, after all it's no use having a download that can go to 48 Khz and speakers that go to 20 - plausible but maybe a bit ott.

My personal favourite is the Ferrari analogy given in the other thread (ie make something that performs well way above the level it needs to so it isn't struggling lower down - probably complete bonkers, but sounds good.

Ultimately though, It feels as if people, with no more than a basic understanding of the science involved, enjoy coming up with x files style theories along the lines of spontaneous human combustion and telekenisis all wrapped in a hugely pseudo-scientific wrapper to make it sound good.

If I was to hazard a guess, is it's simplythat drive units are measured for frequency response, and why not claim that which they can be measured to? It doesn't matter that no human alive can hear above around 22 Khz, and the vast majority not much over 18 Khz, if a drive unit, under testing, shows flat to 40 Khz, it looks like a bigger (is better) number.

ie, which sounds better on the specs?...

22Hz - 20 KHz

18Hz - 22 KHz

18Hz - 44 Khz

I've even chosen headphones with these measurements in mind, not really thinking about it...

The answer is definately in there somewhere.

 

[Anonymous]

Higher frequencies have an effect on lower frequencies in harmonics - Hugely doubtful, sounds like made up gibberish.

Really? It's only gibberish if you have no understanding on how sounds are constructed.

Have a look into the background behind Fourier analysis (see if you can fin Marcus De Sautoy's recent Radio 4 podcasts about mathematics, one of them deals with Fourier, actually it might have been on his BBC2 show as well, with pictures, which makes it easier to visualize the waveforms), every sound you hear can be constructed out of combinations of basic sine waves of different frequencies, you look at an oscilloscope trace of a complex sound, such as brass instruments and it looks a mess, but break it down and you find that it's just combinations of sine waves of different frequencies. The point of upper harmonics is that even though we can't hear them directly they will add to the harmonic frequencies that we can hear (if you add two sine waves of the same frequency together they combine and increase, the same applies to harmonic frequencies). Of course if they're out of phase they'll subtract. In either case the presence of the harmonic frequency will alter the sound of the frequencies we can hear. 40kHz is a harmonic of 20kHz (and 10kHz and 80kHz...).

This may (note: may) explain why people don't like mp3 as much as lossless audio, mp3 specifically "throws away" the frequencies we can't hear, if that removes harmonics it's possible people could subconsciously notice it. That would particularly apply to orchestral music, you're much more likely to get those sort of upper harmonics from strings and brass instruments.

It is quite correct to say that any soundwave can be though of as a combination of pure sine waves, of different frequency, phase and amplitude.

Where this all falls down is that if the transducer (your ear) can't respond to the higher component frequencies, either because of age or too many Van Halen concerts, then you can't hear it.

It doesn't matter that if you could hear it, it would change the shape of the resultant waveform, you can't hear it, your ear isn't up to it, and you can't hear it.

So, while it is interesting to speculate about what would happen if you could hear 40KHz, you can't. Any impact that a 40KHz harmonic would have on the sound of lower frequencies, however different the resultant waveform would look on a 'scope is lost to you because you can't hear it.

I really would suggest Google-ing some of this stuff rather than speculating. Most universities publish their papers, and with a bit of digging you can find a wealth of excellent analysis.

 

[The_Lhc]

It doesn't matter that if you could hear it, it would change the shape of the resultant waveform,

That isn't what I said, the higher frequencies add to and subtract from the soundwaves of frequencies you CAN hear, that's the point.

 

[Anonymous]

It doesn't matter that if you could hear it, it would change the shape of the resultant waveform,

That isn't what I said, the higher frequencies add to and subtract from the soundwaves of frequencies you CAN hear, that's the point.

The changes that the higher frequency waves make to the soundwaves you CAN hear are undetectable by your ear, because your ear cannot respond to these higher frequencies. It was if the higher frequencies weren't there in the first place, you can't detect them, and therefore you can't detect the effect they have on lower frequecies you can detect.

To use an electronics analogy, if you have a signal containing 2 frequencies, one at 100Hz, one at 300Hz and look at it on a 'scope it produces a complex waveform. The 300Hz signal has undeniably made a difference to the lower frequency signal and produced a complex result. If you pass the mixed signal through a low pass filter with a cuttoff at 150Hz all you end up with is a pure 100Hz signal. It was as if the 300Hz signal had never been added in the first place. Your ear is the same, its mechanical components form a low pass filter. Higher frequencies added to a lower frequency sound just dont make a diference because they are undetectable.

 

[krazy_olie]

It doesn't matter that if you could hear it, it would change the shape of the resultant waveform,

That isn't what I said, the higher frequencies add to and subtract from the soundwaves of frequencies you CAN hear, that's the point.

The changes that the higher frequency waves make to the soundwaves you CAN hear are undetectable by your ear, because your ear cannot respond to these higher frequencies. It was if the higher frequencies weren't there in the first place, you can't detect them, and therefore you can't detect the effect they have on lower frequecies you can detect.

To use an electronics analogy, if you have a signal containing 2 frequencies, one at 100Hz, one at 300Hz and look at it on a 'scope it produces a complex waveform. The 300Hz signal has undeniably made a difference to the lower frequency signal and produced a complex result. If you pass the mixed signal through a low pass filter with a cuttoff at 150Hz all you end up with is a pure 100Hz signal. It was as if the 300Hz signal had never been added in the first place. Your ear is the same, its mechanical components form a low pass filter. Higher frequencies added to a lower frequency sound just dont make a diference because they are undetectable.

>>THIS

It's sort of counter inuitive, the higher frequency affects the signal and you end up with something more complex but once passed through a low pass filter it makes no difference whether they were there or not. to start with. Easier tto understand if you look at frequency domain graphs rather than time domain

 

[CnoEvil]

AEJim, since you are about these parts at the moment, I would really like to hear your thoughts on this topic....if it's not too controversial.

Cheers

Cno

 

[AEJim]

AEJim, since you are about these parts at the moment, I would really like to hear your thoughts on this topic....if it's not too controversial. Cheers Cno

Hi CnoEvil,

Well, from my point of view the main benefit of having tweeters reaching to 40kHz+ is largely for the region you CAN actually hear, up to 18-20kHZ or so max. All drivers have break-up points and the further you can push them from their main operating bands then the smoother their response should be. The old Porsche going 70mph is much more comfortable than an 800cc Mini kinda thing. However, as others have said, harmonics play a part and also in listening to a supertweeter attached to one of our conventional models there were definite benefits in terms of "air" and space to the sound, a more "live" effect so to speak. This could of course have been in part to it bolstering output into the normal hearing band...

At the end of the day, having tweeters going to these rather extreme frequencies does seem to have benefits and certainly does no harm so I'm sure manufacturers (ourselves included) will continue to use them - at the same time something like our old Aego 2 only went to 15kHz max and no-one seemed to notice, most finding the sound very enjoyable!

 

[altruistic.lemon]

Hey cno, still on this, huh?

What sources do you have that can reproduce over 22khz??

 

[CnoEvil]

Hey cno, still on this, huh?

What sources do you have that can reproduce over 22khz??

....'fraid so. First time round, Richard Allen was the only one (from a manufacturer) to respond.

The whole point of the thread was to understand the approach of different manufacturers....which now only leaves Dean Hartley (whose company has taken the most extreme approach).

 

[CnoEvil]

Hi CnoEvil,

Well, from my point of view the main benefit of having tweeters reaching to 40kHz+ is largely for the region you CAN actually hear, up to 18-20kHZ or so max. All drivers have break-up points and the further you can push them from their main operating bands then the smoother their response should be. The old Porsche going 70mph is much more comfortable than an 800cc Mini kinda thing. However, as others have said, harmonics play a part and also in listening to a supertweeter attached to one of our conventional models there were definite benefits in terms of "air" and space to the sound, a more "live" effect so to speak. This could of course have been in part to it bolstering output into the normal hearing band...

At the end of the day, having tweeters going to these rather extreme frequencies does seem to have benefits and certainly does no harm so I'm sure manufacturers (ourselves included) will continue to use them - at the same time something like our old Aego 2 only went to 15kHz max and no-one seemed to notice, most finding the sound very enjoyable!

Thank you for taking the time to respond - much appreciated.

One more question (and it's a b**?!er), if I may. In your opinion, is there a point, above which there is no point in going (what is it)....or is it a case that the higher you go the better?