AES London 2010
Paper Session P25
P25 - Listening Tests and Evaluation Psychoacoustics
Tuesday, May 25, 14:00 — 18:00 (Room C5)
Chair: Natanya Ford, Buckinghamshire New University - UK
P25-1 Toward a Statistically Well-Grounded Evaluation of Listening Tests—Avoiding Pitfalls, Misuse, and Misconceptions—Frederik Nagel, Fraunhofer Institute for Integrated Circuits IIS - Erlangen, Germany; Thomas Sporer, Fraunhofer Institute for Digital Media Technology IDMT - Ilmenau, Germany; Peter Sedlmeier, Technical University of Dresden - Dresden, Germany
Many recent publications in audio research present subjective evaluations of audio quality based on the Recommendation ITU-R BS.1534-1 (MUSHRA, MUltiple Stimuli with Hidden Reference and Anchor). This is a very welcome trend because it enables researchers to assess the implications of their developments. The evaluation of listening tests, however, sometimes suffers from an incomplete understanding of the underlying statistics. The present paper aims at identifying the causes for the pitfalls and misconceptions in MUSHRA evaluations. It exemplifies the impact of falsely used or even misused statistics. Subsequently, schemes for evaluating the listeners’ judgments that are well-grounded on statistical considerations comprising an understanding of the concepts of statistical power and effect size are proposed.
Convention Paper 8146 (Purchase now)
P25-2 Audibility of Headphone Positioning Variability—Mathieu Paquier, Vincent Koehl, Université de Brest - Plouzané, France
This paper aims at evaluating the audibility of spectral modifications induced by slight but realistic changes in the headphone position over a listener’s ears. Recordings have been performed on a dummy head on which four different headphone models were placed eight times each. Musical excerpts and pink noise were played over the headphones and recorded with microphones located at the entrance of the blocked ear canal. These recordings were then presented to listeners over a single test headphone. The subjects had to assess the recordings in a 3I3AFC task to discriminate between the different headphone positions. The results indicate that, whatever the headphone model or the excerpt, the modifications caused by different positions were always perceived.
Convention Paper 8147 (Purchase now)
P25-3 Objectivization of Audio-Video Correlation Assessment Experiments—Bartosz Kunka, Bozena Kostek, Gdansk University of Technology - Gdansk, Poland
The purpose of this paper is to present a new method of conducting audio-visual correlation analysis employing a head-motion-free gaze tracking system. First, a review of related works in the domain of sound and vision correlation is presented. Then assumptions concerning audio-visual scene creation are briefly described. The objectivization process of carrying out correlation tests employing gaze-tracking system is outlined. The gaze tracking system developed at the Multimedia Systems Department is described, and its use for carrying out subjective tests is given. The results of subjective tests examining the relationship between video and audio associated with the video material are presented. Conclusions concerning the new methodology, as well as future work direction, are provided.
Convention Paper 8148 (Purchase now)
P25-4 A New Time and Intensity Trade-Off Function for Localization of Natural Sound Sources—Hyunkook Lee, LG Electronics - Seoul, Korea
This paper introduces a new set of psychoacoustic values of interchannel time difference (ICTD) and interchannel intensity difference (ICID) required for 10°, 20°, and 30° localization in the conventional stereophonic reproduction, which were obtained using natural sound sources of musical instruments and wideband speech representing different characteristics. It then discusses the new concept of ICID and the ICTD trade-off function developed based on the relationship of the psychoacoustic values. The result of the listening test is conducted to verify the performance of the proposed method is also presented.
Convention Paper 8149 (Purchase now)
P25-5 Effect of Signal-to-Noise Ratio and Visual Context on Environmental Sound Identification—Tifanie Bouchara, LIMSI-CNRS - Orsay, France; Bruno L. Giordano, Ilja Frissen, McGill University - Montreal, Quebec, Canada; Brian F. G. Katz, LIMSI-CNRS - Orsay, France; Catherine Guastavino, McGill University - Montreal, Quebec, Canada
The recognition of environmental sounds is of main interest for the perception of our environment. This paper investigates whether visual context can counterbalance the impairing effect of signal degradation (signal-to-noise ratio, SNR) on the identification of environmental sounds. SNRs and semantic congruency between sensory modalities, i.e., auditory and visual information, were manipulated. Two categories of sound sources, living and nonliving were used. The participants’ task was to indicate the category of the sound as fast as possible. Increasing SNRs and congruent audiovisual contexts enhanced identification accuracy and shortened reaction times. The results further indicated that living sound sources were recognized more accurately and faster than nonliving sound sources. A preliminary analysis of the acoustical factors mediating participants’ responses revealed that the harmonic-to-noise ratio (HNR) sound signals was significantly associated with the probability of identifying a sound as living. Further, the extent to which participants’ identifications were sensitive to the HNR appeared to be modulated by both SNR and audiovisual congruence.
Convention Paper 8150 (Purchase now)
P25-6 The Influence of Individual Audio Impairments on Perceived Video Quality—Leslie Gaston, University of Colorado, Denver - Denver, CO, USA; Jon Boley, LSB Audio LLC - Lafayette, IN, USA; Scott Selter, Jeffrey Ratterman, University of Colorado, Denver - Denver, CO, USA
As the audio, video, and related industries work toward establishing standards for subjective measures of audio/video quality, more information is needed to understand subjective audio/video interactions. This paper reports a contribution to this effort that aims to extend previous studies, which show that audio and video quality influence each other and that some audio artifacts affect overall quality more than others. In the current study, these findings are combined in a new experiment designed to reveal how individual impairments of audio affect perceived video quality. Our results show that some audio artifacts enhance the ability to identify video artifacts, while others make discrimination more difficult.
Convention Paper 8151 (Purchase now)
P25-7 Vertical Localization of Sounds with Frequencies Changing over Three Octaves—Eiichi Miyasaka, Tokyo City University - Yokohama, Kanagawa, Japan
Vertical localization was investigated for sounds consisting of 22 tone-bursts ascending/descending along the whole-tone scale between C4 (262 Hz) and C7 (2093 Hz). The kinds of the sounds used were pure tones, one-third octave band noises, and piano-tones. The sounds were presented through a fixed loudspeaker (SP-A) set up just in front of listeners with numbered cards set perpendicularly (case-1) or with seven dummy loudspeakers attached to the numbered cards (case-2). The results show that most observers perceived the locations of the sound images moved upward from a loudspeaker around SP-A for the ascending sounds or downward for the descending sounds in both cases, although the sounds were radiated through the fixed loudspeaker (SP-A).
Convention Paper 8152 (Purchase now)
P25-8 Variability in Perceptual Evaluation of HRTFs—David Schönstein, Arkamys - Paris, France; Brian F. G. Katz, Université Paris XI - Orsay Cedex, France
The implementation of the head-related transfer function (HRTF) is key to binaural rendering applications. An HRTF evaluation and selection is often required when individual HRTFs are not available. This paper examines the variability in perceptual evaluations of HRTFs using a listening test. A set of six different HRTFs was selected and was then used in a listening test based on the standardized MUSHRA method for evaluating audio quality. A total of six subjects participated, each having their own recorded HRTFs available. Subjects performed five repetitions of the listening test. While conclusive HRTF judgments were evident, a significantly large degree of variance was found. The effect of listener expertise on variability in perceptual judgments was also analyzed.
Convention Paper 8153 (Purchase now)