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

Re: Gaussian vs uniform noise audibility

The auditory system (and auditory neurons) can certainly distinguish
between stimuli based on their amplitude distribution. Uniform noise
sounds more 'smooth' than gaussian, and gaussian sounds more 'smooth'
than heavier-tail distributions like Cauchy. At the limit, consider the
fact that a Poisson sequence of clicks has a flat spectrum, like white
gaussian noise, but sounds completely different.
       Amplitude distributions in time are related to correlations across
frequency. In gaussian noise, the amplitudes and phases of the different
frequency components are independent. In non-gaussian noise, even with
white power spectrum, amplitudes and/or phases are correlated.
       The auditory system is sensitive to correlations across frequency. For
amplitudes, consider for example comodulation masking release and
related work about monaural correlation detection (there's a nice paper
of V. Richards in JASA many years ago about that). We are also sensitive
to correlation in phases across frequency components that fall within
the same peripheral filter, and to the phases of envelopes even across
wide frequency separations (see some interesting modeling work in:
Carlyon RP, Shamma S., An account of monaural phase sensitivity. J
Acoust Soc Am. 2003 Jul; 114(1): 333-48).


Israel Nelken
Dept. of Neurobiology
The Alexander Silberman Institute of Life Sciences
Edmond Safra Campus, Givat Ram    | Tel: Int-972-2-6584229
Hebrew University                 | Fax: Int-972-2-6586077
Jerusalem 91904, ISRAEL           | Email: israel@md.huji.ac.il