1996 OCTOBER, VOLUME 44 NUMBER 10

CONTENT

PAPERS

Nonlinear System Identification for Horn Loudspeakers
Wolfgang Klippel p. 811
The dominant nonlinear mechanisms in horn loudspeakers when excited by small amplitude signals can be modeled using a class of polynomials known as SM systems. The amplitude and phase responses of linear subsystems are estimated in the SM system function by linear regression. The analysis gives qualitative clues to the dominant nonlinearities as well as quantitative estimates of relevant loudspeaker model parameters. This information is valuable for improving horn loudspeaker design and for developing correction filters with which to compensate for horn loudspeaker nonlinearities.

Design and Commissioning of Sound Reinforcement Systems for the Australian Parliament A Holistic Approach
Glenn Leembruggen and David Connor p. 821
Speech reinforcement in the Australian Parliamentary Chambers presented unusually difficult conditions because of wide dynamic range, large orator-to-microphone distances, high levels of audience noise combined with architectural and acoustical constraints. The solution to providing high intelligibility with faithful spectral reproduction included the use of original analytical software and unique tapered line-array loudspeakers. Commissioning the system involved numerical optimization of the gain structure and a novel acoustic loop-gain measurement technique to predict equalizations with which to optimize the acoustic gain margin.

Experimental Auralization of Car Audio Installations
Emmanuel Granier, Mendel Kleiner, Bengt-Inge Dalenbäck, and Peter Svensson p. 835 Currently available architectural modeling software is not usable for studying automotive listening environments since these programs are optimized for large rooms. By employing other models the binaural room impulse response for low frequencies can be predicted in conjunction with an augmented geometrical acoustical model for high frequencies. Software for combining the separate low- and high- frequency modeling results is described and the findings are compared to preliminary listening tests.

Analyzing Listening Tests with the Directional Two-Tailed Test
Les Leventhal and Cam-Loi Huynh p. 850
Researchers typically analyze double-blind listening tests with a one-tailed significance test, which provides for deciding whether performance is better than chance. The little-known directional two- tailed test provides for deciding whether performance is better or worse than chance, and may be more useful to some investigators. These two tests are compared and further discussion is included on when and how to use them effectively. Extensive statistical tables are provided for conducting either test without the need for calculation.

LETTERS TO THE EDITOR

Addendum to 'On the Acoustic Impedance of Baffled Strip Radiators'
Stanley P. Lipshitz, Tony C. Scott, and Bruno Salvy p. 864

STANDARDS AND INFORMATION DOCUMENTS

AES-2id-1996 AES information document for digital audio engineering - Guidelines for the use of the AES3 interface p. 865 AES Standards Committee News p. 893
Request for models; AES-24 protocol; digital audio measurements

FEATURES

Review of Society's Sustaining Members p. 898

DEPARTMENTS

Upcoming Meetings p. 921
News of the Sections p. 922
Sound Track p. 926
New Products and Developments p. 927
Available Literature p. 929
Membership Information p. 931
In Memoriam p. 935
Audio Engineering Society Application Instructions p. 937
Section Contacts Directory p. 940
AES Conventions and Conferences p. 944