10. Acoustic Models of Sound Production
437
mainly produces scattered downward and sideways emissions. Sources
placed in the right nasal plug node projected sound forward at rather high
vertical angles (average e = 38°) and somewhat left of forward. The left
MLDB source location projected beams averaging about 26° right of
forward. The right MLDB source location produced approximately centered beams with some secondary lobe energy into higher vertical angles.
4. Discussion of Results
As with all bioacoustic tissue simulations, the results of this study are dependent on both the accuracy of the wave propagation techniques and on the
fidelity of the tissue models to in vivo tissue geometry and acoustic
properties. Although shear wave modes were ignored, the accuracy of the
compressional wave propagation techniques was demonstrated in Aroyan
(1996) to be excellent. Tissue geometry is generally well represented in
high-resolution X-ray CT data, but some displacement of the supranarial
soft tissues may have occurred through air sac collapse. However, the scan
methodology of Cranford (1992) was developed to minimize gross postmortum tissue distortion. While approximate, the technique used to model
the acoustic properties of soft tissues is based on established tissue mass
density scan practices and on the measured acoustic properties of mammalian and unique delphinid tissues. In addition, simulation results were
found to be quite robust with respect to biologically reasonable variation
of the HU-to-density and HU-to-velocity model mappings. This may in part
be due to the relatively short tissue propagation paths (generally less then
10 wavelengths) involved in the simulated emission processes. While many
opportunities exist to refine the analysis, the work presented here is sufficient to reach a number of interesting conclusions.
The role of the skull in acoustic beam formation has been investigated
using both acoustic sources and ray tracing techniques (Evans et al. 1964),
using lightbulb-as-source techniques (Dubrovskiy and Zaslavskiy 1975),
using acoustic sources with and without soft tissue present (Romanenko
1973), and using two-dimensional bioacoustic simulations (Aroyan 1990;
Aroyan et al. 1992). The results of the three-dimensional skull-only forehead simulations have corroborated and extended the conclusions of these
previous studies. The common dolphin's skull plays a predominant role in
beam formation by establishing the basic focal geometry of the forehead.
The skull by itself is capable of forming significant forward beams solely by
reflection off of its upper surfaces. Focal maxima were found only in the
immediate vicinity of the asymmetrically enlarged right narial depression,
and not elsewhere. The fact that these maxima occurred only on the right
side of this depression and not on the left suggests that the focal geometry
of the skull is functionally related to its asymmetry. Because air sacs cover
portions of the skull surface in the vicinity of the nasal passages, the focal
437
mainly produces scattered downward and sideways emissions. Sources
placed in the right nasal plug node projected sound forward at rather high
vertical angles (average e = 38°) and somewhat left of forward. The left
MLDB source location projected beams averaging about 26° right of
forward. The right MLDB source location produced approximately centered beams with some secondary lobe energy into higher vertical angles.
4. Discussion of Results
As with all bioacoustic tissue simulations, the results of this study are dependent on both the accuracy of the wave propagation techniques and on the
fidelity of the tissue models to in vivo tissue geometry and acoustic
properties. Although shear wave modes were ignored, the accuracy of the
compressional wave propagation techniques was demonstrated in Aroyan
(1996) to be excellent. Tissue geometry is generally well represented in
high-resolution X-ray CT data, but some displacement of the supranarial
soft tissues may have occurred through air sac collapse. However, the scan
methodology of Cranford (1992) was developed to minimize gross postmortum tissue distortion. While approximate, the technique used to model
the acoustic properties of soft tissues is based on established tissue mass
density scan practices and on the measured acoustic properties of mammalian and unique delphinid tissues. In addition, simulation results were
found to be quite robust with respect to biologically reasonable variation
of the HU-to-density and HU-to-velocity model mappings. This may in part
be due to the relatively short tissue propagation paths (generally less then
10 wavelengths) involved in the simulated emission processes. While many
opportunities exist to refine the analysis, the work presented here is sufficient to reach a number of interesting conclusions.
The role of the skull in acoustic beam formation has been investigated
using both acoustic sources and ray tracing techniques (Evans et al. 1964),
using lightbulb-as-source techniques (Dubrovskiy and Zaslavskiy 1975),
using acoustic sources with and without soft tissue present (Romanenko
1973), and using two-dimensional bioacoustic simulations (Aroyan 1990;
Aroyan et al. 1992). The results of the three-dimensional skull-only forehead simulations have corroborated and extended the conclusions of these
previous studies. The common dolphin's skull plays a predominant role in
beam formation by establishing the basic focal geometry of the forehead.
The skull by itself is capable of forming significant forward beams solely by
reflection off of its upper surfaces. Focal maxima were found only in the
immediate vicinity of the asymmetrically enlarged right narial depression,
and not elsewhere. The fact that these maxima occurred only on the right
side of this depression and not on the left suggests that the focal geometry
of the skull is functionally related to its asymmetry. Because air sacs cover
portions of the skull surface in the vicinity of the nasal passages, the focal