Is there an audiologist in the house?

[Jezz-the-Fezz]

I'm making a reference to:
Audio Technology 4H52 by serengetiplains. It askes a very subjective question in that from my own personal experience, this part of the audio spectra varies wildly from person to person, and is far from linear. My immediate response to serengetiplains question would be yes; even though it appears to be low level, stacked ontop of the performance of a linear tweeter it will be exposed - just like adding the noise from one component to another. At this frequency, however, we're not talking about sonics that you hear in the same light as presence frequencies; I'd defy anyone to rattle off a tune in the 10-20KHz region! it surely is in the realms of overtone, and although the majority of us (who are under pensionable age and haven't been doing >130dB on regular occaision for long intervals) will pick it up, it's not as clearly defineable as singing the poigniantly audible notes in the sound of a domestic appliance, but it is the kind of tone that can set your teeth on edge.

In short, I'd be really interested to know if any members are audiologists by profession; thus tying the topic of interest all the way from biology to physics.

Cheers - Jezz

[lineup]

My hearing stops at about 11-12.000 Hz.

It is is not difficult to test this.
Use a variable frequency signal generator
feeding your headphone amplifier
and your good quality headphones.

If you havent got a signal generator
you can use a test CD with different frequencies
covering the spectrum.

Start at 1-2.000 Hz and adjust level for normal or very high volume.
Use this level setting for the rest of test.

[maxro]

Check this out:

http://www.phys.unsw.edu.au/~jw/hearing.html

Max

[thalis]

This break-up hump at 13KHz, more than 25 dB down, is insignificant.
We have here certain information about a characteristic, about a ‘problem’.
The difficult part in audio isn’t knowledge of various information, but construction of an appropriate value scale for our information. Most times we have related information that helps determine the relevant value of a new one.
In this case, we can think like this: There is no midrange driver with any break-up near the one octave region on top of its crossover frequency. If we regard the 2nd acoustical order as most common, we end up with Loudspeakers with similar break-ups 24dB down or even less, and we are speaking about break-ups in lower frequencies (most likely in the region of 4 to 7 KHz) which are for sure more critical for our ears. This means that there is practically no loudspeaker out there which hasn’t got this ‘problem’, mostly in a greater degree. So, Break-up in 13 KHz, >25dB down shouldn’t be problematic. If we tried to get rid of it, we would probably end up with other kind of problems, most likely of greater importance.
To deal with this issue from the audiologist’s or psychoacoustics aspect, is very difficult. We don’t possess enough positive (scientific) knowledge regarding these matters, we just experiment and run into assumptions. We can’t easily answer if something is critical or not, based merely on our existent knowledge of the ear-brain mechanism.
For instance, it is true that most of us (I’ve run some tests to 5-10 people for this) can’t hear tone signals above 17-18KHz at 95dB SPL, some don’t hear even 15KHz. At 13KHz, sound should have considerable pressure for us to be able to here the tone. Since this break-up is more than 25 dB down, if you hear from this speaker at about 95dB and since also spectra of music isn’t flat (like pink noise), I would presume that at 13KHz you wouldn’t hear much. On the other hand, the presence of ultra-sonic frequencies is perceived by our ear. If you hear to a cembalo from your system (this instrument has about 2% of its spectrum above 20KHz and goes up to 50KHz or more, some wind instruments also extend up to the 40KHz region or more) with a super tweeter for the 20 to 50KHz frequencies, and without it, you will here differences. But you can’t hear these frequencies from the super tweeter alone.
So, my answer is that this beak-up is probably perceived by our ears, but its relevant value shouldn’t be of great importance.
(My english isn't good - I Hope it is acceptable:-)

Regards,
Thalis.

[filgor]

Hi all,

I'm not an audiologist I am a biologist and now work in electronics. I know a reasonable amount about how our ears detect sound and how sound is interpreted.

I'm sure of this much: if you have a high enough amplitude sinwave signal ie from an audio generator fed through a decent amp and sweep the frequency on the generator up and down around 13kHz you would most likely notice the amplitude going up and down .
However, when you are listening to any program material ie music these peaks and resonances, being so low in relative amplitude and so marginal to our audible frequency range, would become completely inaudible.

Even if we can theoretically pick up the signal our brain tends to focus on the louder frequencies amplitudes and transients especially when the sound field is very rich.

The compression of MP3s capatilises on this and similar principles.

P.S. I'm not saying it will make music sound like an MP3, far from it

[serengetiplains]

Thalis, I like how you differentiate between information and a value scale for that information. I think many disccussions on this website regarding the audibility of this or that element of the audio chain (pick your favourite: cables, resistors, capacitors, etc.) would perhaps not devolve into flame wars, or not so quickly, if that distinction were more clearly drawn. (I often find myself defending positions that fall on the "information" side, leaving the "value" assessment to later discussion and personal preference.)

From the website posted by Maxro, it looks like the Fletcher Munson curves enhance the AT's natural upper frequency roll-off by a factor of 2-5dB.

[thalis]

Well, I think that all our efforts should be associated with the construction of value-scales and comparative evaluations. I think that knowledge is mostly about that and not about information, in general.

Yes, you are right. I was partially exaggerating about our ears loss of sensitivity in the 13 KHz region, as I also was a bit absent-minded when I stated about the second order 24 dB (evidently 12dB) decrease in an octave :-)
My main reasoning about the 13 KHz bump though, was that every loudspeaker has similar breakups in greater amplitude and lower frequency from this driver, so we should not bother and deteriorate other more critical quantities trying to eliminate it, except if we had other, more critical reasoning to add a lowpass filter.

Elsewhere in the net resources, these curves look somewhat different, with the 13-15KHz region to mid frequencies ratio a considerably bigger (ca 20dB), for instance:
http://www.webervst.com/fm.htm
http://www.sfu.ca/sonic-studio/handb..._Contours.html
These are the real Fletcher-Munson 1933 curves.
These curves where substituted with the Robinson-Dadson 1956 curves, to which your Link is referred, and there has also been a very recent revision in 2003 which I couldn’t find in the Net. The 1933 and 1956 curves can de seen here, together:
http://www.sengpielaudio.com/Fletche...son-Dadson.pdf
BTW the 1956 curves also reveal a second ‘pole’ of sensitivity in the 400 Hz region, in the 40 to 110 dB amplitudes.
In the 2003 revision, we (as I recall) find again significant differences in the upper frequencies, though I don’t remember now if it is ‘for good’ or ‘for bad’.
All these curves come from statistical data, were a (ideally) large sample of people are tested. Our ears sensitivity could be influenced by season (temperature) and by moisture (don’t forget that we have a light diaphragm in our ears), and also because of our residence: People in town live in a sound ‘basis’ (threshold) of 40 or more dB, people in villages should have something like 20 or less, for their entire life, and this should also lead to somewhat different sensitivity curves as I presume.
In general, I believe that sensitivity in the upper frequencies (before we stop hearing) is a very ‘unstable’ condition, as it is with every order which turns from measurable to immeasurable.
Another quantity which is very unstable is the Frequency limit of our listening capabilities. It is referred everywhere as 20 to 20KHz, but I think that this is perhaps a convenient number-rounding. Having ‘measured’ some ears, I believe (though not sure) that most of as most likely hear (pure tones – not just presence of harmonious content) from 15-16Hz to 17-18KHz, with the upper frequency limit deteriorating with our aging.