Subject: Re: New Scientist question and Pitch and pressure effects From: Daniel Rowan <dr(at)ISVR.SOTON.AC.UK> Date: Thu, 22 Nov 2001 11:09:42 +0000Interestingly, I am just writing up a research project that has specifically looked at the issue of pitch shifts with changes in ear canal pressure. A study by Fritze (1995) in Acta Otolaryngol. [Stockh] measured interaural pitch matching and varied the ear canal pressure in a single ear using a tympanometer (0 vs +/- 4kPa). He reports an upward pitch shift of around 0.7% using a 1kHz 60dB(A) pure-tone as the standard (in the ear that the prerssure was varied) with +/- pressure. SOAE frequency is also dependent on ear canal pressure and there is a school of thought that the mechanical properties of the oval window, which are dependent on tran-tympanic pressure, help determine the frequency of SOAE, maybe by via an effect on the BM (?) (maybe some of you OAE modellers can help on this one). Anyway, I have done an experiemt looking at the influence of ear canal pressure on 0.5kHz and 4kHz pure tones, also using interaural pitch matching. No pitch shift at 4kHz although there was an effect on thresholds (thus presumably the middle ear). At 0.5kHz one or two subjects had a downward shift in pitch by approx 0.3%. Hopefully full details of experiment will be published in due course. Myself and my colleagues here have tried the jaw clenching thing. Haven't observed anything. I think there is something in pitch shifts due to direct middle ear influences on cochlea mechanics, although I'm not convinced we are looking at an effect the size described by Corey (as much as a semi-tone). If we limit ourselves to two options, either the frequency of the tone changes on route to the cochlea (not likely) or the coding of frequency is influenced. I am convinced that the pitch-intensity effect was well controlled in my study. So, if it is an influence on frequency coding is it place or periodicity? I can see no reason to suspect periodicity. So then perhaps these middle ear influences do affect place coding. If so, given that most of the evidence is from pitch shifts at low frequencies, maybe this also indicates that place coding is important below 2kHz after all? More questions than answers unfortunately. This area certainly needs more systematic research. If anyone else is working on this area or has got any theoretical comments, please get in touch!!! Daniel Rowan Institute of Sound and Vibration Research University of Southampton Southampton UK dr(at)isvr.soton.ac.uk