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

Re: comparison of vision and auditory system

>Jont Allen <jba@RESEARCH.ATT.COM> wrote:
>> ... The frequency JND is very small, but the acuity of temporal events
>> is suprisingly bad.

I find this point cited often, but the notion of temporal events is not
usually clearly pinned down in the discussion.

For low-level temporal events, such as waveform peaks as resolved by the
cochlear filtering, the auditory system has an amazingly precise ability
to represent time.  Binaural comparison of event times has JND around
10 to 20 MICROseconds.  Monaural detection of displaced pulses in regular
click trains is not much worse.  Generally, the time JND for low-level
events is much better than the reciprocal of the frequency JND for
sine waves--as opposed to "suprisingly bad."

For higher-level events, such as separately perceived sounds, or gaps
in speech, the time JND is indeed much larger.  But I don't see why that
is surprising, as there's no reason to expect these higher-level processes
to response on the same scale as the low-level sound detection levels
of the auditory nervous system.  They respond on a scale appropriate to
the kinds of events that they extract.