Bioacoustics Laboratory

This new laboratory is focused on understanding the mechanisms underlying vocal production in vertebrates, especially mammals and birds, and in using this understanding to design playback experiments to understand animal communication. We use a diverse set of approaches, including non-invasive imaging techniques (x-ray video, ultrasound, EGG) and excised larynx work with high-speed video to understand the physics and physiology of vocal production. The laboratory is one component of a new multi-institute consortium for the study of bioacoustics led by post-doctoral fellow Angela Stöger-Horwath, in collaboration with ZooVienna and the University of Veterinary Medicine, Vienna.

This project, in collaboration with Dr. Jan Svec at the Palacky  University Olomouc, Czech Republic, is developing new visualization tools to understand vocal fold dynamics in humans and other mammals. These are based on electroglottography and are verified using high-speed video in an excised larynx setup. Other collaborators on this project include Prof. Gerald Weissengruber at the University of Veterinary Medicine, Vienna.

Humidified heated air is blown through an excised larynx mounted on a laboratory bench. Vocal fold vibration and correlated data are documented by simultaneous recording of high-speed imaging data (6000 fps), acoustic signal, sound pressure level, electroglottographic signal, airflow, air pressure etc. Various methods and tools are currently being developed to aid in processing and analyzing the captured data (see sections below). The collected evidence is used to (a) gain a better understanding of vocal fold vibratory features, such as registers and chaos; and to (b) study the voice source properties of animal species not yet researched.

We have developed a new method for analyzing and displaying electroglottographic (EGG) signals (and their first derivative, DEGG): the electroglottographic wavegram. The wavegram provides an intuitive means for quickly assessing vocal fold contact phenomena and their variation over time. Vocal fold closing and opening appears here as a sequence of events rather than single incidents, taking place over a certain period of time, and changing with pitch, loudness and register. Multiple DEGG peaks which can occur during both the concacting and de-contacting phase of vocal fold vibration, are revealed in wavegrams to behave systematically, indicating subtle changes of vocal fold oscillatory regime. As such, EGG wavegrams promise to reveal more physiological information on vocal fold closure and opening events than previous methods.

To construct a wavegram, the time-varying fundamental frequency is calculated and consecutive individual glottal cycles are identified. Each cycle is locally normalized in duration and amplitude, the signal values are encoded by color intensity and the cycles are concatenated to display the entire phonation in a single image, similar as done in sound spectrography.

Only a few non-human mammal species can modify their vocalizations in response to auditory experience. Recently, strong evidence has been published, indicating that elephants are vocal learners. A study published together with international experts in Nature (Poole et al. 2005) provided two examples of vocal imitation in African elephants. The overarching goal of the elephant project is to confirm vocal learning in elephants, and explore its mechanistic basis and functional significance.

Understanding vocal learning and plasticity in a new species requires an understanding of sound production mechanism and dynamic possibilities of the vocal tract; therefore we will apply an acoustic camera and non-invasive vocal production analyzing techniques, in particular electroglottography and ultrasound to captive African elephants, trained to produce infrasonic vocalizations (rumbles) on command, to identify the acoustic source, and investigate vocal tract movements during vocalizations. This project will be the first to investigate mechanisms of infrasonic sound production in any mammal.

Both koalas and Andean bears are highly vocal during the breeding season, indicating an exclusive link between vocal activity and reproduction. Our research program aims to determine the information content of these species calls (and hence, their signalling potential), and use playback experiments to investigate the functional relevance of acoustic variation in reproductive contexts. For the Andean bear studies we will use collars fitted with audio capture devices to obtain recordings of captive and free-ranging animals. This project will be a collaborative effort with the Research Institute of Wildlife Ecology and Schönbrunner Tiergarten here in Vienna, and San Diego Zoo's Institute for Conservation Research in the United States.

An additional aspect of this project focuses on deer vocal communication. Deer will be used as model species for exploring the importance of multi-component signals for female mate choice in mammals. This work will be conducted in collaboration with the University of Sussex in the United Kingdom and the Réserve Animalière de la Haute Touche in France.

We are currently conducting excised larynx experiments on different deer species. These experiments will allow to build up a comparative picture of deer vocal production, and shed some light on the remarkable diversity of vocal behaviour observed across deer. Additional experiments will use Computed Tomography (CT) scans to examine the three dimensional structure of larynges from different deer species. The excised larynx experiments are a collaborative effort between our bioacoustics group, the University of Sussex in the United Kingdom, and Palacky University in the Czech Republic.

Experiments are being conducted in collaboration with the Department of Computational Perception (Johannes Kepler University, Linz) on humans to quantify preferences for one type of musical stimuli over another. This will allow us to explore the possibility that some musical preferences may have evolved through sexual selection and hence, that a possible ancestral role for music may in some cases have been to attract mating partners.

In order to analyze acoustic, electroglottographic and high speed video data, we have developed software tools in Python, FIJI/ImageJ and C++, some of which are availabe to download:

	C4 - Christian's C++ Code Collection - A C++ class library for analyzing sound files, particularly spoken and sung phonations
	FijiVkg - A FIJI/ImageJ plugin for generating videokymographic (VKG) images from high-speed movies of vocal folds during phonation.