Re: Autocorrelation

[Christian Kaernbach <chris@xxxxxxxxxxxxxxxxxxxxxxxxxx>]

Dear Jont,

> I would like to suggest that you place some samples on the web
> that demonstrate the effects you are talking about, along with
> a short summary of your paper(s), describing and summarizing
> your point of view. Many of us are interested (some at a superficial
> level) and we would like to better understand the arguments, and
> get up to speed with all of this discussion, without spending a day,
> or days, doing it. If, by chance, you already have such a web site,
> then this would be a great time to give us the URL.

There is such a URL:
        http://www.uni-leipzig.de/cognition/kaernbach/pitch.htm
This site was done more than a year ago, and in quite a hurry. I should
rework it. But it could at least serve to illustrate the main point. I
would appreciate if receiving hints as to how improve this web site.

> One short question: Do you, Christian, feel that the tails of tuning
> curves is relevant to your mechanism B, or to mechanism A for that
> matter? If so, would you please articulate why you feel that way?

I don't think so. The low-pass noise is indispensible to isolate
mechanism B, but not essential to its working. The tails would (as any
property of the tuning curve) have influence on the resolvability and
hence on the operative range of mechanism A.

- Christian Kaernbach