Peter, You are right that pinna-related spectral notches would not be preserved through a headset array. But in terms of enhancing speech-to-noise ratio through spatial filtering, the loss of directional filtering from the pinna would be more than compensated by the directivity of the array. Loss of pinna cues may be more important for localization, especially in the vertical plane. Noble, Byrne and colleagues at NAL have documented such effects with different types of hearing aids. Patrick M. Zurek President Sensimetrics Corporation 48 Grove St. Somerville, MA 02144 Tel: 617-625-0600 x237 Fax: 617-625-6612 email: pat@sens.com At 03:08 PM 4/1/2004, Peter Lennox wrote:
Am I right in thinking that a headset-type array cannot incorporate any of the pinnae filtering part of the HRTF? - isn't this particularly important for the speech-in-noise problem? Even using a capsule either side of the head, so that at least you have duplex components and so some of the hrtf would behave properly, surely the lack of pinnae effects would be serous? - sorry if this is a naive question. It just seems to me that ideally, a hearing aid would make up for amplitude-with frequency loss within the context of the HRTF - ideally, wouldn't one wish to exaggerate pinnae notches to make up for the loss, or is this simply not possible? regards ppl