Subject: Re: Traveling waves or resonance? From: Eckard Blumschein <Eckard.Blumschein(at)E-Technik.Uni-Magdeburg.DE> Date: Thu, 21 Oct 2004 16:49:06 +0200Richard F. Lyon wrote: > I accept > that we have at least a four-way split over how to interpret Ren's > experiment. I'm not saying that any of you is wrong, but that > Ren's data are also very much consistent with the traveling wave > models. Furthermore, those models can accommodate local resonances > that interact with and determine the wave. So I don't see why we need > such a dichotomy between the approaches, when they can be integrated. Martin already subordinates TW but ascribes CFs to the properties of OHCs. Andrew shares my view that TW is just an epiphenomenon. A first reason to do so is the lack of essential energy transmission from base to apex. An additional reasons for me is the fit of my real-valued spectrogram to known data without any problem and without any tweaking. Sophisticated wave equations are a bit too flexible as to be trustworthy. Moreover, speculations on reflected waves and also the argument of phase accumulation have proven untenable. Otherwise, I would be ready to interpret my Fig. 3 in M277 in terms of a TW-model. The epiphenomenal travel inside a standing envelope is well to be seen in this figure. Look at the 3 kHz component showing frequency chirping up with elapsed time. This effect continuously broadens bandwidth as a precondition for subsequent cepstral analysis. Do TW-models exhibit the same? So my spectrogram would be able to replace the subordinated TW in combination with Martin's OHC-hypothesis. Alternatively, it could incorporate any good mechanical or electromechanical model of cochlea. I only expect much additional insight from models that more quantitatively mimick the definitely more involved processes inside cochlea. So far neither the radial coupling between the rows nor the tilted structure of OHC-bundles has been taken into account. Also, Nelson's notch, and many other phenomena are still to be implemented. Even simple lever models might perform better than blind application of differential equations with more input parameters than one can reasonably handle. Anticipating ongoing distrust against my unaccustomed mathematical results, I would like to inform you that I got unexpected support concerning a marginal but fundamental issue: http://www.fh-augsburg.de/~mueckenh/ Eckard