1996 JULY/AUGUST, VOLUME 44 NUMBER 7/8 |

## CONTENT

PAPERS

Frequency and Angular Resolution for Measuring, Presenting, and Predicting Loudspeaker Polar Data

Felicity Seidel and Henrik Staffeldt p. 555The relationships between measurement resolution and the accuracy of presented or predicted polar and frequency responses from single or arrayed loudspeakers are presented. Only direct loudspeaker- generated sound fields are addressed. Observations are made of current, commonly used measurement techniques, which can yield results with significant errors. The discussion includes the possible need to reconsider how loudspeakers should be measured, taking into account the increasing improvements in measuring equipment, data processing systems, and storage techniques.

A Model of Nonlinear Wave Propagation in Horns

Keith R. Holland and Christopher L. Morfey p. 569Sound propagation modeling in acoustic horns at high levels is analyzed. This model can predict the degree of nonlinear waveform distortion associated with propagation in the horn of large amplitude waves. The predicted output sound field can then be corrected for nonlinear distortion by appropriately predistorting the input signal, using the modeling-determined data.

Discrete Summation Synthesis of Musical Instrument Tones Using Genetic Algorithms

San-kuen Chan and Andrew Horner p. 581An automated genetic algorithm method determines discrete synthesis parameters for matching an arbitrary harmonic instrument tone. This work synthesizes the spectra of the trombone, English horn, and the Chinese pipa. This unique technique uses discrete summation matching, which requires less computer memory compared to wavetable matching. Although the relative error using a wavetable is always smaller than that of discrete summation, it is a good alternative choice when wavetable memory is limited since it requires only a single sine wave.

ENGINEERING REPORTS

Improving Audio Codecs by Noise Substitution

Donald Schulz 593Separate coding of noise and signal components is proposed for greater data reduction. Combinations of adaptive transform coding and adaptive subband coding using time and frequency domains in subbands distinguish between tonal and noisy audio components. Listening tests have confirmed the superiority of combined algorithms for a variety of different music signals, while the data rate was decreased by approximately 20%.

STANDARDS AND INFORMATION DOCUMENTS

AES Standards Committee Newsp. 599AES3 interface meetings; loudspeakers and microphones

DRAFT AES-5id-xxxx, AES information document for Room acoustics and sound reinforcement system modeling - Loudspeaker modeling and measurement - Frequency and angular resolution for measuring, presenting and predicting loudspeaker polar data(insert)

FEATURES

100th Convention Reportp. 610Exhibitorsp. 624Programp. 627102nd Convention, Munich, Germany: Call for Papersp. 673

DEPARTMENTS

Review of Acoustical Patentsp. 603News of the Sectionsp. 654Upcoming Meetingsp. 656Sound Trackp. 657New Products and Developmentsp. 658Available Literaturep. 662Membership Informationp. 664In Memoriamp. 669Sections Contacts Directoryp. 676AES Conventions and Conferencesp. 680