GALLERY OF VOCALIZATION BEHAVIORS
90Hz
16Hz
RAPID “B” CALLING
TYPICAL “A”-”B” CALLING
“B” CALLING, THIRD HARMONIC ONLY
C
C
C
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
WHALE GROUPING WITH “D” CALLS
The call behavior seen here has been identified as a blue whale calling behavior in the literature (see Aroyan et al.). The only time it occurs in Pioneer Seamount data is in day 245 of 2001, between 4 am and 2 pm. These “D” type vocalizations appear highly variable and seem to occur within the proximity of multiple individuals.
Blow-up of the repeated type “B” call.
This blue whale “conversation” took place the day after the installation of the underwater cable.  (Time axis in seconds.)
SOUND PRODUCTION MECHANISMS
It has been proposed by Aroyan et al. that in vocalization by blue whales the air-filled cavity formed by the laryngeal sac and nasal passages effectively forms a Helmholtz oscillator, as shown above.  Air flows through an oscillating valve driving the resonant oscillation of the cavity.  In order to produce sustained low-frequency moans -such as the “B” call - this model suggests that the necessary differential pressure on the lungs must come from the act of diving.
Blue whales typically produce sequences of several equally spaced calls or call pairs, separated by a breathing interval (McDonald et al., shown to the left).  In the case of the A-B calling sequences, a reasonable hypothesis would be that one call is produced during the descent, when air passes from the collapsing lungs into the nasal passages and laryngeal sac, and the paired call is produced as the whale returns to its original depth.  CritterCam observations (shown to the left) indicate a pattern during feeding with a rapid descent and ascent, followed by a pause.  While calling may be separate from feeding, the (descent, ascent, pause) pattern could correspond to the (A, B, pause) pattern of the AB calling sequence.
Several other acoustic features of the calls may have implications for the mechanism.  The A call starts and ends with a tone very similar to the B call; in between, short segments at the B-call frequency alternate with pulses at 90 Hz – some sort of valving mechanism is suggested.
Graphics from Aroyan, McDonald, Webb, Hildebrand, Clark, Laitman and Reidenberg, “Acoustic Models of Sound Production and Propagation,” in Hearing by Whales and Dolphins, ed. Au, Popper and Fay (Springer,2000) p. 409-469.
FREQUENCY CONTROL
The most remarkable result of our study is the reproducibility of the “B” call frequency.  (The characteristics of the B call make it possible to measure its central frequency very accurately.)  This frequency is the same, within the accuracy of the measurement, for all calling whales (large and small) and over a space of a year!  Why do they need to call at the same frequency?  And how do they do it?
While we do not have the expertise to propose a model of cetacean hearing, we note that a process of observing beats is suggested.  The typical deviation of frequency that we observe is 0.09 Hz.  This corresponds to a one-cycle error in phase over 12 seconds.  And frequency-matching by observing beats has a sensitivity of roughly one cycle over the period of observation.
The “A” call
“C” and “B” call
The “C” call is always seen preceding the “B” call. The “C” call seems only to occur in association with the A-B call pairs, but rarely contains a significant amount of energy so it is not certain how often this call appears.
This spectrogram of a typical “A” call has been de-noised, revealing the time- frequency characteristics of each pulse. The pulses are comprised of a ~1 sec upward sweep at the same central frequency of the “B” calls followed by a 90Hz “squeak” of about 0.4 sec. The whale seems to ramp up the power of its low-frequency call to hit this non-harmonically related 90Hz note. The call then finishes with a “B” type drone as if it’s clearing the air out of its lungs and preparing for the “B” call or series of ‘B” calls which always seems to follow an “A” call.
“C sharp” call ?
This spectrogram, showing 20 minutes of the loudest call sequence observed, reveals a feature not seen in calls from more distant whales. The feature corresponds temporally to the “C” call but occurs in the 250-350 Hz range. This feature is only visible at these large scales because the signal is so weak and seems to be dispersed over the 10 second length of the “C” call.
250Hz
350Hz
The spectrogram shows half an hour of the most typical calling behavior. An “A” call seems always to be followed by a “B” call 20-25 seconds later. A sequence of 5-6 A-B call pairs is repeated at intervals of 125 seconds, between three-minute breathing intervals. This type of behavior occurs for hours on end and long after the whale is nearly out of range. Notice the faint calling of more distant whales. The third-harmonic band seems to be filled with “B” calls…
This spectrogram, 30 minutes long, shows another abundant type of calling behavior. A faint “A” call is followed by 7-16 “B” type calls spaced 50 seconds apart. These “B” type calls are distinct because they are broken into two sections separated by a 2 second gap. The spacing between each call is much shorter compared to that of the typical A-B call pairs seen above.
90Hz
16Hz
Rapid “B” calling
“D” call
“D” call