2003 January/February, Vol 51 Number 1/2

CONTENT

PAPERS

Analysis of Traditional and Reverberation-Reducing Methods of Room Equalization
Louis D. Fielder   3
Unlike the traditional approach to room equalization, which compensates for the steady-state spectral effects, a true equalization method will become a dereverberator. Such an approach simultaneously removes the acoustic properties of the reproduction environment in both the frequency and time domains. It is a very difficult problem. Although the proposed solution proves to be impractical when employed in a real application, the analysis illuminates several critical criteria for evaluating any solution. New psychoacoustic metrics successfully predicted those degradations that made the system unacceptable.

Kautz Filters and Generalized Frequency Resolution: Theory and Audio Applications
Tuomas Paatero and Matti Karjalainen   27
Most audio signal processing filters use a basic building block containing a delay or a pole, but other choices of orthonormal functions include the use of an all-pass block. When using this type of block, the resulting structures, called Kautz filters, readily allow frequency warping. Although this approach has been overshadowed by the more traditional methods, the authors show that lower order filters are needed when applied to loudspeaker equalization, room response modeling, and guitar body acoustics. The design phase is more complex, but there is no additional computation load at run time.

Horn Acoustics: Calculation through the Horn Cutoff Frequency
Peter A. Fryer   45
The author reconsiders the mathematical approach to analyzing exponential horn loudspeakers above and below the cutoff frequency. This work provides a more solid foundation for the simplified methods of partitioning the mathematics into two regions with different assumptions in each one. By introducing a tiny amount of acoustic loss into the model, the mathematics no longer break down when traversing the transition region at cutoff. The results agree with measured data.

Modified Discrete Cosine Transform-Its Implications for Audio Coding and Error Concealment
Ye Wang and Miikka Vilermo   52
This study of the modified discrete cosine transform (MDCT) explores the implications of audio coding and error concealment from the perspective of Fourier frequency analysis. Subjective coding quality and the tolerance to missing or repeated compressed data blocks often produce contradictory requirements in real applications. Tradeoffs involve the selection of window-sized crossfade transitions between blocks and perception of uncancelled alias components.

CORRECTIONS
Correction to: "On the Use of Time-Frequency Reassignment in Additive Sound Modeling"
Kelly Fitz and Lippold Haken   62

STANDARDS AND INFORMATION DOCUMENTS
AES Standards Committee News   63
MADI; loudspeaker components; peak flutter; tape storage; CD-ROM life; loudspeaker polar data; digital input-output interfacing; digital synchronization; media storage and handling; library and archive systems; forensic audio; audio connectors; shielding and EMC; audio over IEEE 1394

FEATURES

114th Convention Preview, Amsterdam   76
     Calendar   78
     Exhibitors   79
     Exhibit Previews   81
Virtual and Synthetic Audio   93
115th Convention, New York, Call for Papers   112

DEPARTMENTS

News of the Sections   99
Upcoming Meetings   104
Sound Track   105
Available Literature   106
Membership Information   108
Advertiser Internet Directory   109
In Memoriam   111
Sections Contacts Directory   114
AES Conventions and Conferences   120

FPO For the cover:

2003 January/February, Vol 51 Number 1/2

spine: 2003 January/February, Vol 51 Number 1/2