This page indexes experiments related to performance analysis.
- onset -- Given two nearly simultaneous events,
what is the time sensitivity to perceiving their correct order?
The experiment shows that ordering accuracy is very high if the
two events are separated by more than 60 milliseconds. The 75% JND
point for identifying the correct ordering is 22 milliseconds. And
the perception for simultaneity is centered very close to the
note attack onset.
- tempojnd -- Given a constant
acceleration of a sequence of taps, what is the minimum change in tempo
which can be perceived as an acceleration or ritard and how does this
relate to the length of the sequence of taps?
If the starting tempo is 60 MM, then given three taps (two time
intervals), the JND is 63.3/56.9 MM. This means that if one beat
is played at 60 MM, and the next is played at 63.3, the second beat
will just barely be preceivable as being faster. If the second beat
is played at 56.9 MM, the beat will just barely be preceived as being
Maximum sensitivity in tempo change occurs when about 6-8 taps
(5-7 time intervals) are given. For example, for 5 taps starting at
60 MM, the JND is 62.0 as the ending tempo. For four taps, the JND
would be 62.1. And for 10 taps, the JND ending tempo is 62.2.
Another way to describe the tempo JND is by duration differences.
At 3 taps (2 time intervals), if the first time interval is 1000
milliseconds, then the second time interval can just barely be perceived
as shorter when it is shorted by 26.5 milliseconds.