Human echolocation: Acoustic gaze for burst trains and continuous noise Bo N. Schenkman a,b,⇑ , Mats E. Nilsson c , Nedelko Grbic d a CTT – Centre for Speech Technology, Department of Speech, Music and Hearing, Royal Institute of Technology, Stockholm, Sweden b Psychological Sciences Research Institute, Université catholique de Louvain, Louvain-la-Neuve, Belgium c Department of Psychology, Stockholm University, Stockholm, Sweden d Department of Electrical and Information Technology, Lund University, Lund, Sweden article info Article history: Received 2 July 2015 Received in revised form 7 December 2015 Accepted 14 December 2015 Keywords: Blind Echolocation Bursts Noise Orientation abstract This study explored the ability of blind and sighted listeners to detect reflections, ‘‘echoes”, of burst trains or continuous noise. Echo detection was compared by presenting 5 ms bursts, rates from 1 to 64 bursts, with a continuous white noise, all during 500 ms. Sounds were recorded in an ordinary room through an artificial binaural head, the loudspeaker 1 m behind it. The reflecting object was an aluminum disk, diameter 0.5 m, placed at 1 m. The sounds were presented to 12 blind and 26 sighted participants in a laboratory using a 2-Alternative-Forced-Choice methodology. The task was to detect which of two sounds contained an echo. In Experiment 2, 1.5 m distance sounds were presented to the blind only. At 1 m, detection for the blind increased up to 64 bursts/500 ms, but for the sighted up to 32 bursts. At 1.5 m, the peak performance for the blind was at 32 bursts. At the 1 m, but not at the 1.5 m distance, the blind performed best with continuous white noise. The overlap in time of signal and echo at 1 m for 64 bursts was 60%, but at 1.5 m 82%. Avoiding an overlap between emitted bursts and returning echoes seems important for echolocation, indicating that an acoustic gaze, analogous to in echolocating animals, may also exist in humans. Ó 2015 Elsevier Ltd. All rights reserved. 1. Introduction This paper investigates how burst train compares to continuous noise in influencing echolocation for humans. Human echolocation pertains to the structure and composition of the reflecting object, to the characteristics of the room, to the abilities and experiences of the blind person, the properties of the signal itself and if the sound was emitted by the individual’s own voice, or by using an external source, such as tapping with a long cane on the ground. The sound may be a short burst, or a longer sounding noise. This study evaluates how echolocation varies with the rate at which a sound is repeated over a specific period of time and how varying burst rates compare to continuous noise. The success of echolocation by blind people depends on a num- ber of factors. Research has shown that in some cases, the spatial acuity of blind people may approach that of bats [44]. The human capacity for echolocation has presumably evolved by other ways, and is dependent on other factors. Studies on bats have shown that they modulate the outgoing signals according to the environmen- tal demands (e.g. [22]), many species for example change the signal when they approach a prey. We suppose that blind people act in a somewhat similar adapting way, when they are close to or far from an object, when there are many reflecting objects, or when there is much ambient noise. They may need to change and adapt the sig- nals to their emitted rate, intensity or frequency content. The avoidance of overlap of signals is prevalent among echolo- cating animals. Madsen and Surlykke [21] used the concept of acoustic gaze adjustments when echolocating animals update their acoustic sampling of the world. Both bats and toothed whales have to wait for their echoes to return before emitting the next sonar pulse (see also [5]). If they emit them too fast, range ambiguity will occur. This happens when they emit a new sonar pulse before pre- vious generated echoes have arrived. Seibert et al. [39] for one bat species they studied, compared each pulse-pair to a visual saccade and regarded the sonar beam movements between pulses as acoustic gaze saccades. To the adjustments belonging to acoustic gaze, Madsen and Surlykke [21] lists hearing, call rates, levels, fre- quencies and beam width of the emitted sound pulses. We believe that similar processes to avoid overlap also take place for human http://dx.doi.org/10.1016/j.apacoust.2015.12.008 0003-682X/Ó 2015 Elsevier Ltd. All rights reserved. ⇑ Corresponding author at: CTT – Centre for Speech Technology, Department of Speech, Music and Hearing, Royal Institute of Technology, Stockholm, Sweden. E-mail addresses: bosch@kth.se (B.N. Schenkman), mnn@psychology.su.se (M.E. Nilsson), Nedelko.grbic@eit.lth.se (N. Grbic). Applied Acoustics 106 (2016) 77–86 Contents lists available at ScienceDirect Applied Acoustics journal homepage: www.elsevier.com/locate/apacoust