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

Re: MAA and stream separation


>>         ... Maybe an absurd example would help to see why.
>>         ... What if our M(inimum)A(udible)A(ngle) were, instead of
>>         1-2 spatial degrees, actually much larger ... say, 100 degrees !
>>         ... i.e., TWO speakers would be perceived as ONE speaker
>>         unless they were separated by MORE THAN 100 degrees !!
>Let's take another absurd example.  You are listening to two talkers
>through a *single* loudspeaker.  You would almost never think they were
>one speaker, even though the angle of separation is zero.
        ... Let's be a bit more accurate.
        ... I agree we would not think those two talkers coming from the
        single LoudSpeaker were ONE talker.
        ... But, that is NOT the point !!!
        ... You would NOT be able to separate the TWO talkers 100% of
        the time, i.e., focus on either - at will - as we do in the Cocktail
        Party Effect.

>Spatial separation is useful, but is far from the defining feature of stream
>segregation, for the simple fact that in highly reverberant environment,
>the necessary information (time of arrival, phase)  can often be ambiguous.
        ... No.
        ... In a system that time-locks to Sources (as you will no doubt
        agree binaural hearing does!) local reflections are actually fused
        with THOSE Sources as per the Precedence Effect.
        ... Your concern should be: "How can I do what Hearing does?"

Would you like to try a very surprising stereo experiment that will completely
UNMIX those TWO talkers played-back thru that single LoudSpeaker ???


McGill is running a new version of LISTSERV (1.8d on Windows NT). 
Information is available on the WEB at http://www.mcgill.ca/cc/listserv