Subject: Re: AUDITORY Digest - 27 May 2003 to 28 May 2003 (#2003-107) From: Jont Allen <jba(at)auditorymodels.org> Date: Thu, 29 May 2003 06:57:44 -0400Robert There is a record that came out in JASA, produced by Ed Villchur. Mead Killion would have a copy, and I suspect could give you wave file versions. > > ---------------------------------------------------------------------- > > Date: Wed, 28 May 2003 13:48:12 -0400 > From: Bob Masta <masta(at)UMICH.EDU> > Subject: Hearing Loss Simulations > > Does anyone know where to obtain (hopefully on the Web) > .WAV files simulating what would be heard by people with > various hearing losses, including cohlear implant users? > > Thanks! > > > Robert Masta > tech(at)daqarta.com > > D A Q A R T A > Data AcQuisition And Real-Time Analysis > Shareware from Interstellar Research > www.daqarta.com > > ------------------------------ Fa-Long Please stop sending HTML format to this list (or any list for that matter). A good coder does, essentially, replace the FFT magnitude with an excitation pattern. The accuracy in doing this however is no better that our understanding of the excitation pattern, and the details of how it is generated. There are several issues. For example, (1) what exactly is the critical bandwidth of the cochlear filters, as a function of the intensity and level of the stimulus. We dont yet know. (2) what is the magnitude of the neural noise associated with each filter? Is it is Poisson, as has been proposed? (3) We need a detailed model of detection of signals at threshold, in the presents of other masking tones. This model must accurately model the upward spread of masking for complex signals. I dont think we are there yet. The best model to date is the one by M. Heinz (ARLO). We have a way to go yet before we can accurately calculate the masking pattern for a complex stimulus. Jont Allen > > Date: Wed, 28 May 2003 10:46:17 -0700 > From: Fa Long Luo <falong.luo(at)QSTECH.COM> > Subject: Can Audio be further Compressed by Replacing FFT Magnitude Spectrum > with Excitation Pattern > > > Dear Experts and Lists, > I would be most grateful if you could give any comment, advice and = > further information about the idea: further audio compression with = > excitation pattern replacing the FFT magnitude spectrum used in = > calculating masked threshold and the related bit allocation in current = > audio codecs such as AC3 and AAC. . The motivation for this thought is = > on the basis of the fact that excitation patterns are much flatter than = > the magnitude spectrum. > Thank you for your kind attention. > I am looking forward to hearing your Reponses. > > Fa-Long Luo > >