Klarinet Archive - Posting 000781.txt from 2003/08
From: Tony@-----.uk (Tony Pay)
Subj: [kl] Difference tones again
Date: Thu, 28 Aug 2003 01:15:45 -0400
On Tue, 26 Aug 2003 13:13:29 -0400, joseph.wakeling@-----.net said:
> Tony Pay wrote:
>
> > We had a similar confusion of levels here when we talked about
> > 'difference tones'. Some people thought that the variation in
> > amplitude associated with the sum of two sine waves was
> > 'appreciated' by the eardrum. And, I have to admit, I thought so
> > too for many years.
> >
> > But in fact, the eardrum is 'too busy' -- har har -- 'obeying' the
> > forces involved.
>
> What a wonderful coincidence you should mention that. I am in
> Copenhagen at the moment at a conference on complex systems science
> and have just attended a wonderful talk on the very issue of how
> so-called "difference" tones are perceived.
Well, first of all, I should say that my remark above that you quote
should be read in the context of the discussion about difference tones
that we'd been having -- which was, do difference tones 'exist', in the
sense that physically subtracting out the sine waves (say, by adding two
more, but 180 degrees out of phase with the original ones) would still
leave the air vibrating at the 'difference' frequency? Or are they
'real only to us', in the sense that they are perceived by us as an
output of our auditory processing systems?
Pretty clearly, the answer is the second. As your pal puts it, we hear
something that isn't *there*.
> In fact "difference" tones is a serious misnomer. There is
> psychophysical data from the 1960s that shows that, if you start with
> two tones of a given frequency difference (and a resulting perceived
> pitch), and then raise the two pitches while *keeping the frequency
> difference between them the same*, a change in pitch is actually
> observed. But Helmholz' ideas have had so much influence that this is
> still not widely appreciated.
It's taken me quite a while since your post to sort out in my mind where
I was going wrong in siding with Helmholtz against Young over
'difference tones' -- although of course, Young's wasn't really a
'theory' of difference tones. You have to explain the neural mechanism
by which variations of amplitude come to be associated with pitch
perception. I suppose I was saying that producing that diagram of the
sum of two sine waves doesn't (yet) constitute an explanation; see:
http://www.woodwind.org/Databases/Logs/2001/02/000320.txt
Of course, I don't have a proper background in any of this. But
Schouten et al I thought was to do with what's called 'the missing
fundamental illusion'. That is, if you have a sound that is 'harmonic',
ie consisting of whole number multiples of a fundamental frequency, then
we perceive the pitch of the complex as of a note at the fundamental
frequency. And this remains true *even if you filter out the
fundamental frequency*.
Helmholtz had explained this by invoking non-linearity of the ear,
saying that heterodyne relationships between the remaining harmonics
generated by that non-linearity effectively replace the fundamental.
That idea I'm told fails because the effect would appear to be too
small, plus Schouten et al.
But -- am I right that Schouten et al did their experiment with a
complex consisting of several harmonics, not just two? After noting the
presence of the 'missing fundamental', they shifted each harmonic by the
same number of vibrations, and noted that the 'missing fundamental'
moved too, contra Helmholtz.
That didn't surprise me when I first read about it, because that meant
that their complex was then not in simple harmonic relationship with
itself, and I knew at least *about* the difficulties of pitch assignment
in those circumstances. Like, how difficult it is to play in tune with
'non-harmonic' instruments like bells -- or even pianos! At least a
part of that is to do with the fact that, in those circumstances, with
sounds made up of non-harmonic partials, identity of pitch assignment of
two notes played one after the other may not coincide with a 'zero-beat'
effect when they are played together -- so the perceived fundamental
*would* move. And that's because our systems 'try to fit' non-harmonic
partials into a harmonic template in some way yet to be explicated --
which explication I take it Chialvo has contributed to.
Now, for some reason, I thought that the 'missing fundamental'
phenomenon was different from the 'difference tone' phenomenon, and
that Helmholtz stood up on that last one. (Which is clearly pretty
stupid -- both must be related to the survival advantage of being able
to sort out 'a jungle full of whoops and screams' -- ie, harmonic
events.) Somehow I thought that 'just two' tones had to be harmonically
related. But when I tried to wave my pen to show why, it didn't work.
So now, when you describe Schouten's result above as applying to two
tones, not three -- *does* it?
I suppose it must, thus making Helmholtz's explanation of the
'difference tone' effect wrong too.
But I don't find the reference.
> Anyway, the talk I attended was by a friend of mine called Dante
> Chialvo who trained as a medical doctor and has since had a wide and
> colourful research career spanning a great variety of biological
> subjects. Preprints of his papers explaining the "difference tone"
> phenomenon can be found here:
>
> http://www.apkarianlab.nwu.edu/~dchialvo/psprints.html
>
> The top two papers in the list are the ones in question.
Thanks, Joe.
Tony
--
_________ Tony Pay
|ony:-) 79 Southmoor Rd Tony@-----.uk
| |ay Oxford OX2 6RE http://classicalplus.gmn.com/artists
tel/fax 01865 553339
....... I almost had a psychic girlfriend but she left me before we met
---------------------------------------------------------------------
Klarinet is supported by Woodwind.Org, http://www.woodwind.org/
