10 – 13). For each sequence heard by a subject we studied the amplitude and phase (relative
to the acoustic AM) of the aSSR in contiguous 2-s epochs from each channel.kThat is, one
amplitude and phase value of the aSSR was obtained from each successive 2-s epoch of the
channel’s brain signal via a Fourier transform, yielding approximately 30 amplitude and
phase values 148 channels per sequence.
Since a good deal of our analysis concerns phase information, it is worth giving a brief
conceptual explanation of phase. By amplitude modulating our tone sequences at 41.5 Hz,
we are introducing an oscillatory signal into the brain at that same frequency. This causes
an oscillatory response (the aSSR) at that frequency in certain brain regions. The relative
337Figure 21.7 Examples of pitch contours from the different conditions used by Patel and Balaban^44 (A) Random
sequence, (B) 1/fsequence, (C) 1/f^2 sequence, (D) scales. All sequences had ~150 tones of 415 ms each, with no
pauses between tones. The sequences in this figure can be heard in sound examples 6–9.
Semitones re A4 (440 Hz)Time (s)ACBD- 10
010- 10
010- 10
010- 10
0100 1020304050600 1020304050600 1020304050600 102030405060kTwo-second epochs were chosen on the basis of pilot analysis suggesting that this was the shortest epoch
we could study and still maintain a desired signal to noise ratio.