grab wrote:No, I'm putting myself up as someone capable of checking whether I might be being sold a line.
But have you actually checked?
Ben Duncan is certainly no idiot and books like his are peer reviewed before publishing -- I've done a fair few for Focal myself! I see little reason to doubt his claims but, as always, it pays to make independent checks when facts like this seem so surprising.
So... since you're capable of checking, what is the radius of a hydrogen atom?
And... what is the average deflection of a human tympanum for a given sound pressure level, and how does that relate to the deflection at the threshold of hearing?
I've made some quick checks myself and as a result I am pretty confident that BD's claim is broadly correct.
He's made one further claim which is plausible but has no reference to source studies: 0.1% of people hearing significantly outside the normal range.
There have been several learned papers on ultrasonic hearing in humans that I've seen. Not all have withstood scrutiny and not all have had independently verified results... but some have shown fairly convincingly that some people can perceive something related to ultrasonic pressure waves -- although usually only at very high SPLs if I recall correctly.
But the ear is a very non-linear organ and it is quite possible that non-linearities cause ultrasonic signals to be detectable in some fashion or other -- if not as sound then at least as a sensation that some people can recognise reliably.
I've not searched for evidence to support his figure here, but 0.1% doesn't seem unreasonable to me. 0.1% of the population can run at extraordinary speeds, jump extraordinarily high, lift extraordinary weights or are extraordinarily intelligent or talented in some other way...
[*] He's then stated the audio bandwidth to be *way* outside generally-accepted human norm, with no reference to source studies or in fact any justification at all:
Again, there certainly are many papers that discuss this. Some I've read claim that human hearing perception is based at least in part on the time domain analysis of signals rather than the frequency domain, and that the temporal accuity of human hearing appears to be the fourier equivalent of something well in excess of 50kHz -- even though we can't detect 50kHz sine waves.
Okay, so that doesn't confirm BD's claim of 200kHz, but it is much higher than the 'accepted' limit of 20kHz and I dare say if I looked around hard enough I'd eventually discover the same literature that BD has read!
Of course there's a bell curve of hearing performance - the question is where the bell becomes zero at each end. This is a fairly standard thing for audiologists to check, so I thought someone might have actual figures for it.
There are plenty of published figures for simple sine-wave auditory frequency testing. But as we all know, few audio signals are simple sine waves. Most are harmonically complex and most have extremely complex transients which can not be analysed using simple fourier analysis techniques.
Although I think the jury is still out on this, I think it wise to maintain an open mind as to the actual range of audio signals we can perceive and which might be important.
I have in the back of my mind some research I read years back trying to explain how we could hear stereo imaging with the accuity we can, when the time of arrival differences were significantly smaller than the notional response time of normal brain nerve cells!
Thing is, if your amp is linear and stable from infrasound to ultrasound then you're likely to have rather good performance in the 20-20k region.
Very true...
I'm *very* suspicious of outrageous and unsubstantiated claims.
It's always wise to be cynical and inquisitive, but equally it is important to confirm whether such cynicism is valid before pronouncing upon it...
hugh
