In This Section
Journal of the AES
2009 January/February - Volume 57 Number 1/2
The following text and diagram (next page) describe how papers are currently being reviewed by the Audio Engineering Society. If you have any comments on the process, you can let us know by submitting them to the Journal Forum.
Hierarchical Bandwidth Limitation of Surround Sound—Part II: Optimization of Bandwidth Allocation Strategy
A technique for band limiting a multichannel surround sound, previously described, requires an optimization strategy based on subjective testing. In this technique, the 5 channels are transformed into a hierarchical domain where each transformed signal is assigned some percentage of the available bandwidth. Two cases with total bandwidth of 40 and 60 kHz were tested using a wide range of audio samples, and the resulting audio quality on a 100-point scale was rated at 87 and 93, respectively. This technique was superior to traditional bandwidth-limiting algorithms.
While parametric equalization and dynamic processing have been around since the early twentieth century, and while they are both a mainstay of audio engineers, they are seldom combined into a single operation. A digital system is described in which dynamic processing is designed to operate with the building blocks of a parametric equalizer. This provides to the user time-varying amplitude control in selected frequency bands. Audio mastering and audio reinforcement would benefit from the integration of equalization and dynamics as a single tool.
Creating an arbitrary length audio background sound, called sound texture, from a small original source is a form of sound synthesis. For example, in video games it may be desirable to continue the aural impression of blowing wind or background traffic for as long as the player remains in a scene that requires that background. Using a block segmented overlap and add algorithm with crossfades, the resulting sound preserves the subjective characteristics of the original yet without being perceived as repetitive. The original source is partitioned into blocks, which are then randomly sequenced. Block length of 2 seconds is optimum.
While experimental prototypes of digital loudspeaker arrays usually use a PCM format, preliminary results show that sigma–delta technology is also a viable alternative. In the test configuration, an FPGA is connected to 32 autonomous true-digital amplifiers driving a corresponding number of miniature moving-coil loudspeakers organized in an array. The sigma–delta approach produces comparable results but with lower harmonic distortion, especially off axis and at mid-frequencies.
Standards and Information Documents
AES Standards Committee News
Shielding of cables; angular measurement of polar radiation; forensic audio; loudspeaker modeling and measurement; audio connectors
Recent developments in materials and miniaturization have led to the possibility for novel forms of audio transducers. Advances have also been made in direct digital transduction and in the handling of signals at extremes of the audio frequency range. In this article we provide a summary of papers from the last two AES conventions describing the latest work in loudspeaker and microphone designs that depart from conventional designs in one way or another.
14th Regional Convention, Tokyo, Call for Papers
127th Convention, New York, Call for Papers