[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Besides HRTF

Forwarded message:
& From caron Tue May 30 12:52:28 1995
& Subject: Re: paper.
& To: bofriis@VNET.IBM.COM (Jens Bo Friis)
& Date: Tue, 30 May 1995 12:52:28 -0400 (EDT)
& From: Francky goes to Hollywood <caron@vertex.INRS-Telecom.UQuebec.CA>
& In-Reply-To: <199505300629.AA18745@velcro.inrs-telecom.uquebec.ca> from "Jens
 Bo Friis" at May 30, 95 08:27:43 am
& X-Mailer: ELM [version 2.4 PL21]
& MIME-Version: 1.0
& Content-Type: text/plain; charset=US-ASCII
& Content-Transfer-Encoding: 7bit
& Content-Length: 2719
& & "Localization using nonindividualized HRTF," JASA, july 1993 by Wenzel
& & Where can I get a copy of this paper?
&       In general, big compagnies has specialized library in which you can
& access to all the conferences and proceedings and all that engineering stuff
& around many subject related to the state of the art in signal processing,
& for example.
&       If not, well, I got around 30 references here. All about it since
& 1989 to summer 1994. If you can read french, I got here my report that give
& you a nice snap-shot (30 p) including all the references.
& & Do you have any tricks that make perception better? I thinking of
& & things that I could implement besides the HRTF filters that would
& & make front-back localisation better?
& 1-    Besides HRTF, the most important factor is MOUVEMENT. It got to
& move, because you're not just friendly (and the HRTF is not perfectly
& custumized to you) with this synthetic environment.
&       Adding mouvement to HRTF in the way that you need it to localize the
& fixed source, is a hell of programming. At this point, you need real-time
& computing environment with a head-localization loop feeding your program.
& The program will adjust the azimut and elevation of the source according to
& the next position of your head. Wightman, F.L., Kisler, D. were talking
& about that at the 127th meeting of Acoustical Society of America on june
& 6-10th 1994. And I strongly beleive in that, I really do.
&       I've done it in a simple way. In simulation, I've just played in a
& sequence, many processed sound files that are just beside one to each other.
& It was just for that, much better than in a random way.
& 2-    Beside HRTF again, you GOT to simulate a real environment because
& you're not just familiar within an anechoic room (infinite room). Just go
& into an anechoic room, and you'll be lost for a moment. You won't react the
& same way to any quind of sound coming to you; your head is gonna move toward
& it, I'm telling you.
&       So reverberation is also important, and that gives you nice sound
& files with many images ( high reflexion order )
&       I could talk to you about ... but I just can't type all about it!
&       For now, I am working on something else (Echo Canceller) but I hope,
& I'll have the chance to work on that again. See, this tends toward
& psycho-acoustic research. I love it but for now, I got to learn more about
& adaptive signal processing... ;-)
& &
& & Regards,
& & Bo

Francois Caron                  //// //// ////     caron@inrs-telecom.uquebec.ca
INRS-telecommunications        //// //// ////           @ (514)-761-8603
16 Place du Commerce          //// //// ////            @ (514)-362-9015
Ile des Soeurs, Montreal               ////             @ (514)-761-8501 Fax
CANADA           H3E 1H6