v7.1, 20040930, me
Friday, October 29, 2:00 pm 6:00 pm
Session H MULTICHANNEL SOUND
Chair: Geoff Martin, Bang & Olufsen A/S, Struer, Denmark
H-1 Customization for Personalized Rendering of Motion-Tracked Binaural SoundJoshua Melick, V. Ralph Algazi, Richard Duda, Dennis Thompson, University of California, Davis, CA, USA
Motion-tracked binaural, or MTB, recordings preserve the dynamic sound localization cues provided by voluntary head motion, making MTB less sensitive than other binaural methods to mismatches between the HRTF of the listener and the HRTF of the recording system. However, MTB performance can be improved by customizing the reproduction process to the listener. In this paper the different types of mismatch and their perceptual consequences are identified. Techniques are presented for partially or completely correcting the mismatches, and properties of these techniques are described.
Convention Paper 6225
H-2 5.1 and 22.2 Channel Multichannel Productions Using an Integrated Surround Sound Panning SystemKimio Hamasaki, NHK Science & Technical Research Laboratories, Tokyo, Japan, Kyushu University, Fukuoka, Japan; Setsu Komiyama, NHK Science & Technical Research Laboratories, Tokyo, Japan; Hiroyuki Okubo, NHK Science & Technical Research Laboratories, Tokyo, Japan; Koichiro Hiyama, NHK Science & Technical Research Laboratories, Tokyo, Japan; Wataru Hatano, Tamura Corporation, Tokyo, Japan
5.1 surround sound productions and broadcasts have become popular in Japan, and NHK has developed 22.2 multichannel sound system for ultra-high definition video. While two-dimensional or three-dimensional sound reproduction is possible by these multichannel sound systems, the production of contents is more complicated and time consuming than two-channel stereo production. In productions using a conventional surround sound mixing tool, in particular, much time is needed for creating two-dimensional sound effects. Therefore, an integrated surround panning system was developed to enable various surround sound effects to be created easily. This paper introduces the newly developed integrated surround sound panning system, which has innovative functions such as a distance control and an integrated sound source movement control, and discusses various issues on multichannel sound production.
Convention Paper 6226
H-3 Examination of Multichannel Sound Field Recomposition Utilizing Frequency-Dependent Interaural Cross Correlation (FIACC)Teruo Muraoka, Tomoaki Nakazato, Masaki Ichikawa, Musashi Institute of Technology, Tokyo, Japan
Locations of loudspeakers were examined utilizing frequency dependent interaural cross correlation (FIACC) for optimum sound field recomposition in the multichannel recording and reproducing process. Experiments were conducted by comparing pairs of FIACC, where one of a pair was measured in an original sound field, and the other was measured in a reproduced sound field. Conclusively, it became clear that ITUs recommendation to the speaker arrangement in 5 channel system is reasonable.
Convention Paper 6227
H-4 Modeling Auditory Localization of Subwoofer Signals in Multichannel Loudspeaker ArraysJonas Braasch, William L. Martens, Wieslaw Woszczyk, McGill University, Montreal, Quebec, Canada
For economical reasons, home entertainment surround sound systems are usually equipped with a single subwoofer channel. The main argument for this procedure is the believed inability of the auditory system to localize low frequencies in small reverberant rooms. However, a psychoacoustic localization test that was conducted using a standard 5-channel set-up with subwoofers showed that the listeners were able to determine the lateral displacement left, center or right of the loudspeaker presenting the test stimulus (an octave-band noise burst at 31.5-Hz, 63-Hz or 125-Hz center frequency). Using a binaural model simulating human perception, recordings of subwoofer signals at different positions were analyzed. As expected, the interaural level differences remained nearly constant for different subwoofer positions in the low frequency range. On basis of interaural time differences, however, the model was able to predict the position of the loudspeaker regarding the left/right dimension, verifying the outcome of the listening test. The results indicate the importance to consider more than one subwoofer in multichannel audio systems.
Convention Paper 6228
H-5 Identification and Discrimination of Listener Envelopment Percepts Associated with Multiple Low-Frequency Signals in Multichannel Sound ReproductionWilliam Martens, Jonas Braasch, Wieslaw Woszczyk, McGill University, Montreal, Quebec, Canada
Multichannel surround sound systems that meet the 5.1 channel ITU standard may use a single subwoofer to reproduce signals below 120\Hz. Yet the ITU standard allows for presentation of five different full-range signals to the listener, each extending to a low-frequency limit of 20 Hz. Including multiple low-frequency signals in sound reproduction enables the creation of auditory spatial imagery that features variation in perceptual attributes such as auditory source width (ASW), auditory source distance (ASD), and listener envelopment (LEV). The experiments reported in this paper were designed to determine how identifiable and how discriminable the listener envelopment percepts are in the auditory images resulting from decorrelation between low-frequency signals presented via selected pairs of five loudspeakers positioned according to the ITU standard configuration.
Convention Paper 6229
H-6 Multichannel Sound Recording Using 3-, 4-, and 5-Channel Arrays for Front Sound Stage CoverageMichael Williams, Sounds of Scotland, Paris, France
In previous papers the Multichannel Microphone Array Design (MMAD) procedure has been used mainly to determine the design of arrays giving complete 360 degree coverage of the sound field. Many sound recording engineers, however, use the main microphone array to cover only the front sound stage, and add in early reflections and reverberation either by artificial means (electronic generation) or by using a second array in the reverberation field. This paper describes MMAD procedure applied to only front coverage of the main sound stage using 3-, 4- or 5-channel microphones, covering any desired angle within the front hemisphere and for the usual 1st order microphone directivities. Various array alignments are described in the form of the arc-of-a-circle with different radius. All arrays described are critically linked (seamless) within the front hemisphere.
Convention Paper 6230
H-7 Designing High Spatial Resolution MicrophonesArnaud Laborie, Rémy Bruno, Sébastien Montoya, Trinnov Audio, Paris, France
Multichannel recording is one of the most important challenges of todays audio techniques. A good surround recording should provide good envelopment feeling, accurate localization, a large sweet spot, and respect for tonesall at the same time. The Fourier-Bessel theory and advanced signal processing allows you to obtain directivities designed from panning laws, which have been designed to optimally drive any multichannel layout. This paper presents the underlying concept of high spatial resolution, the spatial equivalent for high fidelity, and points out why this is a key point to achieve high spatial quality. A very flexible and scalable technology providing high spatial resolution, as well as a high-performance 5.0 microphone featuring a compact and light array of 16 omnidirectional capsules are also presented.
Convention Paper 6231
H-8 Capturing Manipulation and Reproduction of Sampled Acoustic Impulse ResponsesRonen Ben-Hador, Itai Neoran, Waves Audio, Tel Aviv, Israel
We discuss the capturing manipulation and reproduction of impulse responses (IRs) of acoustic spaces. While trying to maintain the accuracy of an IR, other factors such as sound quality and musical character of sound should also be considered. Furthermore, IRs are not limited to preserving the sound of venues but also as a tool in music production. Therefore, the IRs are converted to standard multichannel reproduction formats, such as stereo and ITU 5.0. In order to obtain a flexible reverberation tool, the IRs are manipulated to modify acoustic properties such as reverb time and interchannel decorrelation. A new real-time audio plug-in was developed for which IRs of venues and devices were recorded worldwide. The IRs are convolved with dry audio. The plug-in supports mono, stereo, and surround, at sample-rates up to 96 kHz.
Convention Paper 6232