Higher than 22k, or ultrasonic?

[Andy G]

Or just become an audio engineer !!

 

[Nanook]

ultra sonic...

there is some merit to ultra wide bandwidth for human hearing. we cannot actively hear to that frequency of course, but there are "impressions" left that can somehow seem more natural than filtered output or playback.

put yourself in a room with a supertweeter at 50k for any period of time (or a long period of time). An impression will be left.

don't underestimate the ability of the human mind to somehow process the information. We just might not have the language, or jargon or highly specialized field of study to explain it, 'cause humans only "hear" to 20k

I'd love to set up a study of a live performance with overtones that could be predicted and compare to a sound that was filterred. hmmm, psychoacoustics?

anyone have an idea of the highest fequency overtones possible with acoustic instruments?

 

[Scottmoose]

Even a budget moving coil cartridge will trace beyond 25k, and the mastering cutter heads are good to 50k. Moving magnet cartridges have problems above 20k because of inductance, but the better ones, with lower output, go to 30k or more.

Pity the LPs aren't. About the only ones that did were CD4.

Budget moving coils with good tracing at 25KHz? Which?

 

[DAVOhorn58]

Dear All,

I use fostex and TAD tweeters which go to 40khz and 100khz respectively.

Many ribbons go to 30-40khz and the Eton tweeters i use go to 40khz.

The old KEF T27 did 40khz i believe and also the COLES 4001 did 30khz.

So there is no shortage of drive units doing in excess of 20khz.

Can you here a difference?

I believe so.

Also as an aside the QUAD 57's struggled past 16khz but everybody loves their top end.

It isnat always what you do but how you do it.

Oh yeah cheap and cheerful Piezzo's manage in excess of 30khz. So for a few dollars you could add a pair of piezzo's and suck it and see.

regards David.

 

[dnsey]

anyone have an idea of the highest fequency overtones possible with acoustic instruments?

Surely, as with any complex waveform, they are theoretically infinite (but the level usually diminishes as a function of the order).

 

[Cal Weldon]

Can someone please tell me what we are listening to above 15K?

 

[OzMikeH]

In my very limited experience: Mostly evident with vocals: T and S sounds. Breathing or breathy vocals. whispers.

Obviously cymbals and just about any metal percussion/bells/chimes. Acoustic guitar, the first bit of the Twang as the finger leaves the string. The airflow noises from flute and woodwinds.

Essentially the last 2% of signal content that takes you over the goosebumps threshold. The difference between a good sounding recording and a realistic one.

When I disconnect my supertweeters it's like listening to music with a severe head cold or un-popped ears. My main drivers drop 10dB between 14 and 15kHz. Some people can't tell the difference if they're connected or not.

 

[tinitus]

I have at some point managed to make my speakers so dry and "short" sounding by means of the xo, that the result was that the musicians sounded like they played a bit false

Some times when I play violin in my listening room, sound is so "short" and soft that its hard to make it sound good ... but when I play in my kitchen where there is a bit echo due to hard sufaces the violin sound much better and in tune

 

[planet10]

OK guys, chew on this:

dave

Quote from Allen Wright's "The SuperCables CookBook" (recommended http://www.vacuumstate.com/)

OK, OK you say, great story but I'm not a baboon -- I can't even hear 18 kHz anymore so what's the problem? What's phaseshift at 80khz got to do with me?

Well, he went on to tell about an experience he had in hearing a bass clarinet player hit a single note while he was setting up microphones for a recording session and haering each finger hole down the clarinet 'Light up' as the initial air pressure wave ran down the inside of the body of the instrument. He was just in the right place at the right time to hear this effect and was so stunned by it that he had the guy hit it again and recorded it -- and when analysed on a fast storage oscilloscope he found that this supposedly bass instrument was making measurable and repeatable harmonics at 18khz and 24kHz, and that the initial transient was at 48kHz.

Now unlike the big monk you may not be able to hear 48kHz but i trust you can hear 6kHz, and that's what's in the air when those 18 and 24kHz harmonics beat together. And if the 48 and 24kHz harmonics are not exact octaves of each other due to normal constructional variations in the instrument, you will hear (or at least sense) the differences between them -- which might be back down in the bass region again.

 

[Salas]

Your hearing must suck. I can hear above 22kHz, though not as well as I can at, well, lower than 22kHz. In one of my electronics classes a few years back, we hooked a signal generator up to a small speaker and just went from like 10kHz up. Most people could hear to 18-20kHz; I went up to about 23kHz without having to turn up the amplitude (again, at a reduced volume though).

However, either way, I don't think it's an important range.

I used to conduct such experiments in a psychoacoustic seminar. -Special CDs and tested headphones.- Ages from 20 to 50yrs. ALL had to double the volume to faintly listen to the 18kHz and 20kHz tones, when the 15kHz tone was discernible using the same volume as with the sub 10kHz tones. Did you go up to 23kHz using the same volume you had for 15kHz? This is extraordinary.

 

[Salas]

OK guys, chew on this:

dave

Quote from Allen Wright's "The SuperCables CookBook" (recommended http://www.vacuumstate.com/)

The beating products fall into the normal range as the quoted text indicates. Its well understood that supersonic beatings can create sonic products. But this has to do with producing sound, not reproducing sound. In the text example, if the reproduction chain can replay the recorded normal sonic range including the beating products that fall in it, why would it have to extend where they have been originally born?

 

[Svante]

My view on this is this:

We typically cannot hear frequencies above 20 kHz, most people don't hear anything above 16-18 kHz.

Most people listen to CDs, which should not produce anything above 22050 Hz. If they do, they are faulty.

However, if a loudspeaker is made to reproduce up to 40 kHz, there is at least a slight chance that the response in the 10-20 kHz region is more well behaved.

On the other hand, there is also a risk that quality in the audible band will have to be sacrificed in order to reach the 20k+ frequencies.

IMO things like these takes the focus away from what is really important; the audible range. Frequency response cannot be simplified to an upper and a lower limit, there is a whole curve from 20 to 20000 Hz that should be sort of flat. Actually there is a whole bunch of such curves in different directions. Finding the right balance in these in the audible range is a far more challenging task than extending the range above the audible frequencies.

 

[Scottmoose]

Interesting thoughts all. FWIW, I seem to remember a study done a few years back showing harmonics of a trumpet went up past 100KHz. Now all we need is a format that can actually do it, if we go along with the thought that if you chop off something, the consequencies will always be felt further down the chain. You lop off the last foot of a python, methinks the rest might have a reaction. Does it apply to audio? I've no idea. Most supertweeters are resonating / breaking up something chronic above about 25KHz anyway, rather than behaving in a pistonic fashion. But I suspect if it's implemented well, it will have an effect on the audible band. The frequent reaction to adding a supertweeter to a system is not that the HF has become more extended, but the midband & even the upper bass has cleaned up.

Speaking generally, I have a suspicion a little too much attention is given to upward extension though sometimes, when the main audio-band hasn't been properly sorted. Look at all those dodgy commercial speakers with supertweeters, and a deeply mediocre / dodgy response in the telephone-band, worse still, they've usually got an XO slap-bang in the middle of that region too. Sigh. Bit like a medieval eunuch in a brothel, there is little point to the exercise.

 

[chrisb]

very little point, indeed (has my wife been talking to you?)

 

[Salas]

Interesting thoughts all. FWIW, I seem to remember a study done a few years back showing harmonics of a trumpet went up past 100KHz. Now all we need is a format that can actually do it, if we go along with the thought that if you chop off something, the consequencies will always be felt further down the chain. You lop off the last foot of a python, methinks the rest might have a reaction. Does it apply to audio? I've no idea. Most supertweeters are resonating / breaking up something chronic above about 25KHz anyway, rather than behaving in a pistonic fashion. But I suspect if it's implemented well, it will have an effect on the audible band. The frequent reaction to adding a supertweeter to a system is not that the HF has become more extended, but the midband & even the upper bass has cleaned up.

Speaking generally, I have a suspicion a little too much attention is given to upward extension though sometimes, when the main audio-band hasn't been properly sorted. Look at all those dodgy commercial speakers with supertweeters, and a deeply mediocre / dodgy response in the telephone-band, worse still, they've usually got an XO slap-bang in the middle of that region too. Sigh. Bit like a medieval eunuch in a brothel, there is little point to the exercise.

Adding a very fast small transducer to a system, alters two main things. 1. The rise time of its impulse response. 2. Its last octave beaming. It makes a system rising faster and beaming wider. This has enough impact so to alter system's impression on listeners. I doubt its just a matter of sheer HF extension if at all.