Subject: Re: Wasn't v. Helmholtz right? From: Stefan Uppenkamp <stefan.uppenkamp(at)MRC-CBU.CAM.AC.UK> Date: Tue, 23 May 2000 10:18:15 +0100> > Dear Bill, > thank you for the address! My question rather implicated a judgement by Roy > Patterson, since I was aware of a poster by S. Fobel, St. Uppenkampf, R. > Patterson, and B. Kollmeier, entitled Asymmetry in Perception of Short > Chirp Cignals, presented in Oldenburg, and dealing with two types of > stimuli. Chirps with rising frequency were calculated to compensate for the > dispersion of traveling wave. Dau postulated that this up-chirping will > sound more compact than an ordinary click and much more compact than chirps > with falling frequency. The opposite seems to be true. > Concerning the reply by Jont Allen, I feel a little bit embarrassed. > Fortunately, I have no idea who is suggested to die and why. > - Eckard > Just for clarification, some short remarks on short chirps with rising and falling instantaneous frequency: Dau et al. (JASA 107, pp. 1530-1540, 2000) demonstrated that short frequency chirps with rising instantaneous frequency designed to compensate for BM dispersion will increase wave V of the auditory brainstem response, when compared to click reponses. ABR to time reversed ("down-") chirps shows less synchronisation than a click response. This is a nice and very clean result which can be explained by increased synchronisation of the low frequency channels in case of the up-chirps. It does, however, not include any "postulate" of how these signals sound like (and therefore there is nothing on sound quality in that paper). When listening to these synchronising chirps and its time-reverse, it is obvious that up-chirps sound less like a click than down-chirps, despite the increased synchronisation of BM sections (as reported on that poster at the DAGA meeting in Oldenburg this year). We interprete this finding as indication for a process of temporal integration beyond the structures generating ABR wave V, that transforms the time code into a time-interval code - effectively removing between- channel phase differences while preserving within-channel fine structure. More on this during the forthcoming ISH meeting in Mierlo Stefan Uppenkamp