1 Spectral Discrimination Thresholds Comparing Audio and Haptics Lorenzo Picinali Department of Media Technology De Montfort University, Leicester, UK lpicinali@dmu.ac.uk Brian FG Katz LIMSI-CNRS Orsay, France brian.katz@limsi.fr ABSTRACT Previous studies have been carried out concerning the ability of individuals with normal hearing to discriminate auditory stimuli which have the same fundamental frequency but different spectral content [5]. This study concerns to what extent is it possible to perform the same differentiation considering vibratory tactile stimuli. Being aurally able to distinguish between an A3 note played by a flute and the same note played by a clarinet, the question posed is whether or not it is possible to observe a similar ability in terms of vibratory tactile stimulations. In trying to address these questions, a research project is currently ongoing consisting of a series of perceptual tests to be carried out in an attempt to compare discrimination thresholds in terms of spectral differences between auditory and vibratory tactile stimulations. Author Keywords Audio, haptic, tactile vibration, haptic-audio transfer function, spectral discrimination threshold. ACM Classification Keywords H.5.2 [Information interfaces and presentation]: User interfaces - Haptic I/O. H.5.2 [Information interfaces and presentation]: User interfaces - Auditory (non-speech) feedback. INTRODUCTION This paper presents an ongoing research collaboration between LIMSI-CNRS (Orsay, France) and the Fused Media Lab (DMU, Leicester, UK) based on perceptual studies for comparing detection thresholds for spectral differences of auditory and vibratory tactile stimulations. The cooperation between these two laboratories has already been presented through a series of conference publications in the same research field between 2009 and 2010 [2][3]. In the following sections a brief overview will be carried out of the objectives of the project and of the procedures for the tests that will be performed, illustrating possible guidelines for the analysis of data resulting from the experiments, and possible outcomes. OBJECTIVE Within multimodal rendering applications, in order to be able to use the same models and algorithms for the haptic feedback that are used for audio, it is essential to understand how these two modalities are perceived in terms of amplitude, frequency, time and spectrum. The main objective of the current project is to acknowledge how spectral variations are perceived within a tactile vibratory stimulation, and to compare these results with those measured for auditory stimulations. This will possibly bring to light the identification of a series of differences between auditory and tactile vibratory perceptions; differences that could then be used for the implementation of transfer functions to be applied in order to directly use an auditory stimulus as a haptic one. BACKGROUND It is well known that differences in the frequency content and spectral envelope in auditory signals are often perceived as timbre variations, allowing for the differentiation between stimuli with the same fundamental frequency, loudness and duration, while having different spectrum [5]. But, is it possible to observe similar abilities via the perception of vibratory tactile stimuli? Previous studies have shown the distinction between different tactile vibratory stimuli and perceived intensity: in [8] the subjective perceived intensity as a function of the vibration frequency have been studied and compared in relation to contact area. In [9] studies on frequency and amplitude discrimination for tactile stimuli have outlined the differences between tactile and auditory perceptions. In [6], vibration detection thresholds were compared using different haptic devices. The ability to discriminate between different signals and design parameters for the generation of tactile feedback has been investigated in [4]. Nevertheless, basic studies, performed using pure and/or very simple tones, and comparing the discrimination thresholds for spectral variations between audio and tactile stimuli, have not be found in the literature. EXPERIMENTAL PLAN The experimental apparatus will be composed by a software component, a computer, an audio interface, an audio amplifier and two 8” loudspeaker woofer drivers: one of these will be modified by removing the actual cone, and leaving only the magnet and the coil. A coupling system will be created in order to transfer the vibrations of the coil to the hand, possibly using a small rigid flat surface on which the fingers of the subjects will be placed. Using