In This Section
Journal of the AES
2011 April - Volume 59 Number 4
Room reverberation changes perception in a signal-dependent way. A reverberation masking index (RMI), which relates to room size and listening location, represents the way in which reverberation decay produces perceptual smearing and masking. These influences are time varying and signal dependent rather than a signal-independent description of room acoustics. Informal listening tests show a direct relationship between the RMI and subjective experience. The index also provides guidance for preprocessing the audio signal for a particular acoustic environment.
Comparing Sound Radiation from a Loudspeaker with that from a Flexible Spherical Cap on a Rigid Sphere
Modeling a real loudspeaker as a spherical cap on a rigid sphere is often more useful than the conventional model of a loudspeaker as a rigid piston in an infinite baffle. This new model can predict such parameters as sound pressure, baffle-step response, sound power, directivity, and the acoustic center. This research extends earlier work without depending on a more complex scheme. Future work using this approach will show the relevance to other topologies.
Detecting applause in both audio recordings and real-time performances is relevant in such applications as music information retrieval and spatial audio coding. A combination of mel-frequency cepstral coefficients and low-level descriptors yielded the best classification performance in the experiments. Low-pass filtering of the feature time series leads to the concept of sigma features. Binary misclassification occurs more often when applause and nonapplause with similar amplitudes are simultaneously present.
Comparison of Receiver-Based Concealment and Multiple-Description Coding in an 802.11-Based Wireless Multicast Audio Distribution Network
Because data packets of compressed audio are sometimes lost in wireless distribution networks, some kind of concealment strategy is needed to mitigate the perceived degradation. In selecting a strategy additional bandwidth, processing requirements, and latency must also be considered. Simulations of several strategies based on psychoacoustic models are used to evaluate the audio quality with packet loss. The advantage of a multiple-description system is the graceful degradation of audio quality with increasing packet loss.
Standards and Information Documents
AES Standards Committee News
High-resolution audio on Blu-ray Disc; core audio metadata; metadata for audio object structures for preservation and restoration; acoustics and sound source modeling; grounding and EMC practices
42nd Conference Preview, Ilmenau
There’s a growing diversity of methods and technologies devoted to transferring audio over digital networks, both wirelessly and over physical connections. Experts on audio coding and networked audio applications gathered in San Francisco at the AES 129th Convention to explain a number of the most recent developments in the field.
45th Conference, Helsinki, Call for Papers