(Georgia Institute of Technology, 2001-07)
Walker, Bruce N.; Lane, David M
Magnitude estimation was used to determine preferred datato- display mappings, polarities, and psychophysical scaling functions relating data values (like temperature) to underlying acoustic parameters (like pitch, tempo, or spectral brightness) for blind and visually impaired listeners. The resulting polarities and scaling functions were compared to findings with sighted participants. There was general agreement between the polarities obtained with the two listener populations, with some notable exceptions. There was also evidence for strong similarities with regard to the magnitudes of the slopes of the scaling functions. The results indicate that sonification designers will need to consider whether their intended listeners are visually impaired or not. However, conclusions from this study are limited by the small sample of visually impaired participants. Further research is necessary to arrive at more definitive recommendations.
(Georgia Institute of Technology, 2001-07)
Walker, Bruce N.; Lane, David M
It is not always clear to a display designer how to use sound most effectively to represent data values. This is where the cumulated experiences of the auditory display community could greatly assist in the creation of more effective and more useable sonifications and auditory displays. However, there is no one central place to find recommendations about which sounds are best for a given data type. Past results in this field have been presented in a broad range of journals, publications, and conference proceedings, making i t difficult to search for design guidance. In the absence of such guidance, a designer might wish to try out several competing approaches to determine the best performer. Unfortunately, schedule and budgetary constraints often make this impractical, if not impossible. Thus, there is a clear need for a central repository where researchers can contribute their experience and wisdom about data sonification and auditory display design. Walker [1][2] has pointed out that in order to begin creating a successful sonification it is critical for a designer to know at least three things about the data-to-sound mapping: 1. The nature of the mapping. Which data dimension (e.g., temperature, pressure, velocity) is mapped onto, or represented by, each acoustic parameter (e.g., frequency, loudness, tempo)? For example, temperature may be represented by the frequency of the sound, so that as temperature changes, the frequency of the sound changes. 2. The polarity of the mapping. When the temperature increases, does the frequency of the sound increase or decrease? In the case of temperature-to-frequency mapping, it is likely best to use an increasing-toincreasing (or up-up) polarity. However, in the case of size and frequency, it is likely best to use an increasingto- decreasing (up-down) polarity. 3. The scaling of the mapping. If the temperature increases by 30 degrees, how much change must you make in the frequency to convey that temperature change to the listener?
(Georgia Institute of Technology, 2000-04)
Walker, Bruce N.; Kramer, Gregory; Lane, David M
We determined preferred data-to-display mappings by asking experiment participants directly and then examined the psychophysical scaling functions relating perceived data values to underlying acoustic parameters. Presently, we are extending and validating the scaled mappings in practical data interpretation tasks. The resulting scaling functions, in conjunction with the experimental paradigm developed here, should spark further research in this area and have implications for the design of future sonifications.