"Picket Fence" effect
Miller and Licklider (J. Acoust. Soc. Am. 22: 167-173,1950)
combined interrupted speech and interrupted
noise maskers to create an interesting effect they described as the
"picket fence" effect. In this situation,
speech and noise are temporally interleaved to form a continuous signal.
They reported that perceived
continuity of the speech signal was increased by this manipulation,
although intelligibility was unaffected.
Bashford, Reiner and Warren (Percept. Psychophys. 51:
211-217,1992) have found that the addition of
noise to fill the silent gaps can improve intelligibity if the stimuli
are complete sentences, rather than isolated
monosyllables as used by Miller and Licklider.
The picket fence effect is illustrated with audio examples below. The
first example is of a male voice
saying "The watchdog gave a warning growl" in quiet. This is followed
by several processed versions
in which 50% of the speech waveform is alternated with white noise
at a signal-to-noise ratio of 0 dB.
The audio examples illustrate speech-noise alternation at rates of
1, 2, 4, 8, 16, 32, 64, 128, 256, and 512 Hz.
The figure shows the waveform of the sentence processed with a 1-Hz
alternation rate.
.WAV format
.AU format
-
Original sentence
Original sentence
-
1 Hz
interruption rate
1 Hz
interruption rate (see figure)
-
2 Hz
interruption rate
2 Hz
interruption rate
-
4 Hz
interruption rate
4 Hz
interruption rate
-
8 Hz
interruption rate
8 Hz
interruption rate
-
16
Hz interruption rate
16 Hz
interruption rate
-
32
Hz interruption rate
32 Hz
interruption rate
-
64
Hz interruption rate
64 Hz
interruption rate
-
128
Hz interruption rate
128
Hz interruption rate
-
256
Hz interruption rate
256
Hz interruption rate
-
512
Hz interruption rate
512
Hz interruption rate
Email:
assmann@utdallas.edu
School of Human Development
The University of Texas at Dallas, Richardson,
Texas, 75083, USA
Tel: (214) 883-2435
FAX: (214) 883-2491