The main objective of this thesis is to quantify and compare the relative accuracy of different auditory rendering schemes. These schemes are used in a visual-to-auditory sensory substitution device that renders graphical objects as sound. The device is portable, as it only requires a tablet, stylus and headphones. Current auditory displays perform a horizontal sweep of an image and produce sounds for each column of pixels, with pixel height represented by pitch and brightness represented by loudness. This approach is noisy, has low resolution, and is difficult to interpret. We propose a more efficient encoding in which lines, curves, and shapes are traced out in time by a cursor whose position is represented by sound. This method also provides real-time auditory feedback to the blind user as he draws, thus providing a training mechanism that maps auditory cues to kinaesthetic cues. Different auditory schemes are used to represent the position of the moving stylus or cursor. We compare their relative performance quantitatively using such metrics as perceived position, size, and aspect ratio of sound-rendered points, lines and ellipses. In contrast, previous research on auditory rendering schemes has been more qualitative in nature.