Let's Hear How Big You Are

Philippe Roux
(UCSD - Marine Physical Laboratory, La Jolla CA 92093-0238)

David A. Demer
(NOAA - Southwest Fisheries Science Center, 8604 La Jolla Shores Drive, La Jolla, CA 92037)

Julien de Rosny
(ESPCI - Laboratoire Ondes et Acoustique, 10 rue Vauquelin, 75005 Paris, France)

Popular version of paper 2pPA12
Presented Tuesday afternoon, November 11, 2003
146th ASA Meeting, Austin, TX

What is the effect of the human body on an acoustic field in the audible regime? Is absorption or scattering dominant? Using acoustic measurements of a human walking in a reverberant room, we show that the human body scatters sound predominantly and that sound absorption is due to the clothing only. Besides the first-ever measurement of the acoustical scattering properties of the human body, these results can lead to interesting considerations for the design of concert hall.
 
  

Figure 1 - Scattering and absorption measurements. The human moves in the reverberant room between the shots k and k+1 (a and b). The reverberated acoustic waves are recorded on the microphone (hk(t) and hk+1(t), c and d), and the coherent intensity corresponding to the acoustic waves reflected on the boundaries of the room (e, Sc(t) blue) can be separated from the incoherent intensity due to the human motion (e, Si(t) green). The ratio S(t) of the coherent and incoherent intensity (e, red) decreases linearly with time. The slope depends on the scattering cross section of the human.

This work is based on the following idea. When an acoustic pulse travels inside a very reverberant room in the presence of a human body, the resulting signal received on a microphone is composed of numerous echoes (Figure 1). Among those echoes, some have only been reverberated by the room walls, floor and ceiling (blue line) while others have also been scattered by the person inside the room (green line). Those echoes vary of course in amplitude and time with respect to the position of the person inside the room. As a consequence, if several acoustic pulses are emitted sequentially while the person is moving inside the room, the respective recordings on the microphone are composed of both a constant background signal corresponding to the reflections on the room boundaries only and a shot-dependent signal that is related to the scattering properties of the human body.

From an ensemble of reverberation time series, we have shown that the ratio of the coherent intensity and the incoherent intensity (Fig. 1e) inside the room provides an absolute measurement of the scattering cross section of the human body. Moreover, the comparison of the intensity in the empty room, prior to the human entering, and the room with the human provides the value of the absorption cross section of the body. Interestingly, these measurements do not depend on the room properties or on the human behavior inside it. What makes the measure robust and independent of the room is that, from pulse to pulse, the body is not encountered only once by the acoustic wave, but multiple times and from multiple directions due to the high reverberation in the room.
 

Naturally, this time-domain technique can be applied using pulses with different frequency contents. This enables the measurement of the frequency-dependent acoustic scattering cross section of the human body. As expected, we show then that the human body frequency response is comparable to one from a hard ellipsoid with same volume (Figure 2), and a height to width ratio of the ellipsoid similar to that for a person. One can think of an ellipsoid as either an elongated sphere, or a 3D version of an ellipse.

Figure 2 - Scattering cross section σT spectra of multiple humans (grayscale lines) measured between 0.1 and 3 kHz, rescaled by the geometric limit for the high frequencies σT∞. The averaged spectra in ka domain (dashed red) is similar to the spectra for an ellipsoid (dashed green). k is the ratio between the frequency and the sound speed, and a the characteristic length of the ellipsoid (minor axis).

 

Figure 3 - Variations of the scattering (σT, red) and absorption (σa, blue) cross sections for one human wearing different amounts of clothing.

The acoustic absorption of the body was measured for a human wearing different amounts of clothing (Figure 3). Wearing underwear, the absorption cross section was measured at 0.003 m2, close to null. Then, by increasing the number of clothing layers, the absorption cross section of the same human increased to a maximum of 0.86 m2, while the scattering cross section always remained around the same value of 0.5 m2.

This study shows how to measure the absolute scattering and absorption cross section of any object in a reverberant room, independently of the room characteristics. Applied to humans, it confirms at the same time that the body scatters sound like an ellipsoid of same volume, while absorption is due to clothing only. It was well known that the response of a concert hall can be very different depending on the number of spectators. We show that it also depends on the amount or type of clothing they are wearing. If the hall is crowded or if only few people are attending, its scattering and absorption acoustic properties might be modified significantly, but a power adjustment could be realized to compensate for the effect of the humans.

References:
J. de Rosny, P.Roux, J. Acoust. Soc. Am. 109, 2587 (2001).
J. de Rosny, P. Roux, M. Fink, J.H. Page, Phys. Rev. Lett. 90, 9-4302 (2003).

Acknowledgments:
We would like to thank W.A. Kuperman for his support in this experiment, Larry Robertson of SWFSC and the RIMAC Center staff from UCSD for allowing us to conduct these experiments in the fallout shelter and the squash court respectively.  We also would like to thank the 27 people who graciously participated in these experiments.