Next: Results
Up: Simulations of segregation
Previous: Simulations of segregation
For experimental stimuli, we will assume f1(t) is a sinusoidal signal
and f2(t) are two types of noise where f21(t) is an AM
bandpassed noise and f22(t) is a bandpassed random noise, as follows:
f1(t) |
= |
|
(26) |
g1(t) |
= |
|
(27) |
f21(t) |
= |
|
|
|
|
|
(28) |
f22(t) |
= |
|
|
|
|
|
(29) |
where
is a bandpass filter with a center
frequency f0 and a bandwidth of 23 Hz (as shown in
Fig. ), f0=600 Hz,
,
and R(f) is an uniform random within
.
Here, EM(t) and ER(f,t) are amplitudes in which white noise
lowpass filtered at 30 Hz and added to the bias value to prevent
over modulation.
Note that the amplitude ER(f,t) fluctuates independently in
frequency region f, and that the power of noise is adjusted so that
.
Here, the bandwidth of noise is 1 kHz and the SNR between a sinusoidal
signal and the bandpassed noise is -8.5 dB.
These mixed signals are shown in Fig. .
The two types of mixed signals are
fM(t)=f1(t)+f21(t) when
f2(t)=f21(t) and
fR(t)=f1(t)+f22(t) when
f2(t)=f22(t).
When a human hears these mixed signals, he can hear the sinusoidal
signal from fM(t) caused by CMR; however cannot hear the sinusoidal
signal from fR(t) caused by Masking.
These mixed signals are shown in Fig. .
In this simulation, segregation is done using 10 mixed signals which
are made by varying the onset of f1(t),
in Eq.
().
Next: Results
Up: Simulations of segregation
Previous: Simulations of segregation
Masashi Unoki
2000-10-26