A dolphin had previously been trained to perform a cross-modal matching-to-sample task. In one version of this task the animal had to investigate a sample object that was concealed in a box through its echolocation sense alone, then select the correct match among up to three alternative objects visually in air. Given the frequency range of a dolphin click and the limited number of sensors that the dolphin receives the sonar returns with, the dolphin should have difficulties resolving the details of the object. We suggested earlier that the dolphin might be using synthetic aperture to gain a higher resolution of the stimulus. To test this hypothesis we proposed to restrict the movement of the dolphin by stationing him on a bite-plate that was fixed in front of the box that contained the sample object. We trained the dolphin to station on the bite-plate while performing the cross-modal task (echolocation to vision) while recording the sound field around the dolphin through a 16-hydrophone array that was placed in a variety of positions and configurations between the object and the dolphin stationed on the bite-plate. The acoustic data were recorded at 500 kHz and later analyzed. To our surprise the dolphin was still able to perform the discrimination task. In this paper, we present the analysis of the data collected and show that the dolphin employs techniques such as beam steering and beam shaping while acoustically interrogating the object. This suggests that while the dolphin might still employ a synthetic aperture when possible, he might not need it to resolve the details of the object. We are planning to extend the range of objects to new and unfamiliar objects to explore whether the dolphin is indeed able to resolve details of the object acoustically without the need for synthetic aperture.
展开▼
