2006 January/February, Volume 54 Number 1/2 |
CONTENTS
PAPERS
Spatial Impulse Response Rendering II: Reproduction of Diffuse Sound and Listening Tests
Ville Pulkki and Juha Merimaa 3
A proposed technique for spatial impulse response rendering, called SIRR, introduced in a previous article, is examined and evaluated in two listening tests. Based on the analysis presented, a hybrid method consisting of amplitude panning at low frequencies and phase randomization at high frequencies was implemented. In the first test, conducted in an anechoic chamber, the SIRR technique performed best in reproducing a reference virtual reality. In the second test, measured responses of existing spaces were reproduced with SIRR and other techniques. SIRR reproduction was judged most natural, especially at off-center listening.
A Hybrid Time-Frequency Domain Approach to Audio Time-Scale Modification
David Dorran, Robert Lawlor, and Eugene Coyle 21
Increasing or decreasing audio delivery rate without changing the pitch can be achieved by processing the audio in either the frequency or time domain. While time-domain approaches are both efficient and without artifacts, they require the presence of a strong quasi-periodic component. In contrast, frequency-domain approaches work over a wide range of signal types but tend to produce reverberantlike artifacts. A new hybrid approach combines the advantages of both approaches, especially for speech, without requiring an a priori knowledge of the signal.
Effects of Headtracker Latency in Virtual Audio Displays
Douglas S. Brungart, Alex J. Kordik, and Brian D. Simpson 32
In the design of headtracked virtual auditory displays, the latency between head movement and the corresponding image shift can be extremely critical. A series of experiments showed that latency of about 70 ms begins to impair localization ability and becomes noticeable to the listener. While this threshold varies among listeners and types of audio signals, it suggests that headtracker latency is generally not a limiting factor for creating a quality image in state-of-the-art audio display systems, which typically exhibit latency values of less than 30 ms.
ENGINEERING REPORTS
Active All-Pass Crossover Networks with Equal Resistors and Equal Capacitors
René Christensen 45
Active crossover networks can be constructed with all-pass functions and with the additional constraint that resistors and capacitors have the same value. Two such networks are examined: a fourth-order three-way filter and a sixth-order four-way filter. While the transient responses are not perfect, compared to the familiar Linkwitz-Riley configuration, the group delays are significantly smoother for filters of the same order.
CORRECTIONS
Correction to "Spatial Impulse Response Rendering I: Analysis and Synthesis"
Juha Merimaa and Ville Pulkki 54
STANDARDS AND INFORMATION DOCUMENTS
AES Standards Committee News 55
Life expectancy of MO and recordable compact discs; unique identifiers in AES3; loudspeaker
FEATURES
Audio Forensics 56
Audio Networking: Applications and Requirements 62
Reprint of "Speech Reinforcement in St. Paul's Cathedral" 67
29th Conference, Seoul, Call for Papers 95
121st Convention, San Francisco, Call for Papers 96
DEPARTMENTS
News of the Sections 76
Sound Track 80
Upcoming Meetings 82
New Products and Developments 82
Available Literature 83
Membership Information 87
Advertiser Internet Directory 91
Sections Contacts Directory 97
AES Conventions and Conferences 104
FPO For the cover:
2006 January/February, Volume 54 Number 1/2
spine: 2006 January/February, Volume 54 Number 1/2