Aubauer R, Au W W
Hawaii Institute of Marine Biology, University of Hawaii, Kailua 96734, USA.
J Acoust Soc Am. 1998 Sep;104(3 Pt 1):1165-70. doi: 10.1121/1.424324.
In behavioral experiments where real targets are used to investigate dolphin echolocation, it is often very difficult to extract the relevant echo parameters that the animals use to discriminate or classify. The complex relationship between the physical dimensions and the reflection characteristic of real targets prevents separate control of various echo parameters of the stimuli presented in an echolocation experiment. A new echo simulation method presented in this paper avoids this problem. Dolphin echolocation sounds are transformed with the target impulse response into artificial echoes, which are played back to the animal. The phantom echo system is implemented on a digital signal processing board and gives an experimenter fully programmable control over the echo generating process and the echo structure itself. Echoes of several underwater targets were simulated to evaluate the quality of the method. A comparison of simulated echoes with the original echoes demonstrated very good agreement independent of the incident signal (cross-correlation coefficient > 0.95). The method has tremendous potential for investigating animal echolocation and understanding biosonar signal processing.
在使用真实目标来研究海豚回声定位的行为实验中,通常很难提取动物用于辨别或分类的相关回声参数。真实目标的物理尺寸与反射特性之间的复杂关系使得在回声定位实验中无法分别控制所呈现刺激的各种回声参数。本文提出的一种新的回声模拟方法避免了这个问题。海豚回声定位声音通过目标脉冲响应被转换为人工回声,然后回放给动物。虚拟回声系统在数字信号处理板上实现,使实验者能够对回声生成过程和回声结构本身进行完全可编程的控制。模拟了几个水下目标的回声以评估该方法的质量。将模拟回声与原始回声进行比较,结果表明无论入射信号如何,二者都具有非常好的一致性(互相关系数>0.95)。该方法在研究动物回声定位和理解生物声纳信号处理方面具有巨大潜力。