Sonically spatialized screen reading: Aiming to restore spatial information for blind and low-vision users

Abstract

Blind and low-vision (BLV) individuals utilize systems called “Screen Readers” to navigate the World Wide Web. These systems retrieve the text information from the web pages accessed by a browser and convert this text to speech. However, typical screen readers process only one channel of audio and, therefore, convey the information from the web page in single-track, linearized format, stripping from it the two-dimensional spatial information available to sighted users of the same web page. There is an interest in presenting the sound from screen readers in multiple “virtual auditory columns,” aiming to restore the perception of two-dimensional placement of items on contemporary web pages. In this paper, we explore alternative methods to spatialize sound over five horizontal positions, in front of a computer user. An experimental comparison of two of these sound spatialization methods was statistically evaluated. The results indicate that there is not a statistically significant difference in accuracy achieved by application of the Vector-Based Amplitude Panning (VBAP) or the Linear Panning (LP) methods.