10. Acoustic Models of Sound Production
439
To illustrate the physical basis of melon focal behavior, Aroyan (1996)
plotted the density, velocity, and impedance structure in a two-dimensional
slice through the D. de/phis melon and forehead using the modeling techniques of Section 2.2. These plots reveal an impedance matched melon-seawater interface, confirming the matched impedance hypothesis of Wood
(1964). They also reveal a broad velocity depression in the anterior melon
that narrows posteriorly to form a well defined channel through the rear
throat of the melon. The broad velocity depression is indicative of a lensatic
function for the anterior melon, as conjectured by Wood (1964) and proposed in greater detail by Norris (1968). The posterior velocity channel
appears analogous to a flared wave-guide. The velocity of the melon "wall"
rises somewhat gradually to a maximum of about 20% above the velocity
of the channel in the posterior melon. While a velocity reduction of 20%
within a waveguide of uniform geometry may permit highly efficient
guiding of specific wavefunctions, the short and widely flared throat of the
posterior melon in this dolphin is probably best characterized as a partial
or "leaky" wave-guide of biosonar clicks.
An incomplete model of the nasal air sacs has been shown to improve
the forward reflection of energy projected upward, sideways, and downward
by the skull and soft tissue model. Addition of the air sacs into the model
always raised the directivity indexes of the patterns, although the importance of this contribution varied with frequency. Generally speaking, the air
sac model was effective in forward collimation of energy projected upward,
sideways, and downward by the skull along, while the melon was effective
in narrowing (focusing) the main forward lobe. Both the air sac model and
the melon, however, contributed to both effects. Nevertheless, a precise
evaluation of the relative contribution of the nasal air sac system to beam
formation must await better information on the exact configuration of the
air sacs during click production.
The biosonar signal source tissues in the common dolphin were found to
be localized within a remarkably small volume of the nasal passages. A dramatic clustering of inverse simulation foci for all frequencies tested in the
complete model implicates approximately 1cc of tissue centered about
0.7 cm below the center of the right MLDB complex as the most plausible
"hypocenter" of the dolphin's biosonar clicks. However, the posterior lip of
the MLDB complex is often displaced dorsally with respect to the anterior
lip in postmortem specimens (Cranford et al. 1996). It is possible that some
or all of the soft tissues of the supranarial region were displaced slightly
from their air-metering configuration during biosonar click production. In
the author's opinion, the clustering of focal points best supports the conjecture that the right MLDB complex is the source tissue of the pulsed
biosonar signals of this dolphin. Inverse simulation foci were notably absent
from other previously proposed source locations including the left MLDB,
the larynx, and the right nasal plug node. In the complete model, foci were
found only on the right side of the nasal passageway and not on the left.
The nonviability of the larynx and right nasal plug as source locations for
439
To illustrate the physical basis of melon focal behavior, Aroyan (1996)
plotted the density, velocity, and impedance structure in a two-dimensional
slice through the D. de/phis melon and forehead using the modeling techniques of Section 2.2. These plots reveal an impedance matched melon-seawater interface, confirming the matched impedance hypothesis of Wood
(1964). They also reveal a broad velocity depression in the anterior melon
that narrows posteriorly to form a well defined channel through the rear
throat of the melon. The broad velocity depression is indicative of a lensatic
function for the anterior melon, as conjectured by Wood (1964) and proposed in greater detail by Norris (1968). The posterior velocity channel
appears analogous to a flared wave-guide. The velocity of the melon "wall"
rises somewhat gradually to a maximum of about 20% above the velocity
of the channel in the posterior melon. While a velocity reduction of 20%
within a waveguide of uniform geometry may permit highly efficient
guiding of specific wavefunctions, the short and widely flared throat of the
posterior melon in this dolphin is probably best characterized as a partial
or "leaky" wave-guide of biosonar clicks.
An incomplete model of the nasal air sacs has been shown to improve
the forward reflection of energy projected upward, sideways, and downward
by the skull and soft tissue model. Addition of the air sacs into the model
always raised the directivity indexes of the patterns, although the importance of this contribution varied with frequency. Generally speaking, the air
sac model was effective in forward collimation of energy projected upward,
sideways, and downward by the skull along, while the melon was effective
in narrowing (focusing) the main forward lobe. Both the air sac model and
the melon, however, contributed to both effects. Nevertheless, a precise
evaluation of the relative contribution of the nasal air sac system to beam
formation must await better information on the exact configuration of the
air sacs during click production.
The biosonar signal source tissues in the common dolphin were found to
be localized within a remarkably small volume of the nasal passages. A dramatic clustering of inverse simulation foci for all frequencies tested in the
complete model implicates approximately 1cc of tissue centered about
0.7 cm below the center of the right MLDB complex as the most plausible
"hypocenter" of the dolphin's biosonar clicks. However, the posterior lip of
the MLDB complex is often displaced dorsally with respect to the anterior
lip in postmortem specimens (Cranford et al. 1996). It is possible that some
or all of the soft tissues of the supranarial region were displaced slightly
from their air-metering configuration during biosonar click production. In
the author's opinion, the clustering of focal points best supports the conjecture that the right MLDB complex is the source tissue of the pulsed
biosonar signals of this dolphin. Inverse simulation foci were notably absent
from other previously proposed source locations including the left MLDB,
the larynx, and the right nasal plug node. In the complete model, foci were
found only on the right side of the nasal passageway and not on the left.
The nonviability of the larynx and right nasal plug as source locations for