AES New York 2009
Poster Session P14

P14 - Signal Processing

Sunday, October 11, 10:00 am — 11:30 am
P14-1 A New Distance Measurement Method for Distance-Based Howling Canceller—Akira Sogami, Arata Kawamura; Youji Iiguni, Osaka University - Toyonaka, Osaka, Japan
In this paper we propose a new distance measurement method for a distance-based howling canceller. We have previously proposed a howling canceller that uses only distance information between the loudspeaker and the microphone. The howling canceller suppresses howling based on the distance measured by a sonic wave. The conventional measurement method however has a noise while on distance measurement. To solve the problem we propose a new distance measurement method that suppresses the noise. Simulation results in a practical environment show that the proposed distance measurement method can almost exactly estimate the distance more silently than the conventional method.
Convention Paper 7918 (Purchase now)

P14-2 New Technology for Hearing Stimulation Employing the SPS-S Method—Andrzej Czyzewski, Gdansk University of Technology - Gdansk, Poland; Henryk Skarzynski, Institute of Physiology and Pathology of Hearing - Warsaw, Poland; Bozena Kostek, University of Gdansk - Gdansk, Poland, Institute of Physiology and Pathology of Hearing, Warsaw, Poland; Piotr Odya, Piotr Suchomski, Gdansk University of Technology - Gdansk, Poland; Piotr Skarzynski, Sense Organs Institute - Nadarzyn, Poland
A prototype of a the new Compact Audio Therapy Unit (CATU) is presented that can process any audio signal inside a very compact device working in real time, employing advanced digital filtration, signal keying, manipulating playback rate, various spectral modifications of the signal, repeating phrases, and others. It was designed to provide a platform for the therapy with the new Method of the Aural Perception Stimulation (SPS-S). The design for wearability allows one to use the device effectively in normal everyday life conditions, e.g., outdoors. The compact and versatile processing device can potentially open a new era in patients and trainees mobility.
Convention Paper 7919 (Purchase now)

P14-3 Frequency Characteristics Measurements of 78 rpm Acoustic Record Players by the Pulse-Train Method—Teruo Muraoka, Takahiro Miura, Tohru Ifukube, University of Tokyo - Tokyo, Japan
The authors have been engaged in the research for the restoration of seriously damaged audio signals, employing Generalized Harmonic Analysis (GHA). In this research, it is important to know frequency characteristics of sound reproducing equipments to realize proper sound reproduction. However, frequency characteristics of the ancient acoustic record players such as ”Credenza,” etc., are significant but not clear: especially, the frequency characteristics when records are actually reproduced. However it can solely be measured with frequency record and vibrator-method is not used any more. In fact, no shellac-made 78 rpm frequency record can be manufactured today: the traditional measurement techniques are inapplicable. On the other hand, one of the authors previously developed Pulse-Train measurement for phonograph cartridges, in order to obtain their frequency characteristics of amplitude and phase performances. This method is applicable so long as any pulse waveform is curved on record surface. Thus the authors employed this method. Radial directional groove was curved on a surface of shellac disc-record, and Pulse-Train response is obtained by reproducing the record with an acoustic record player. Some examples will be exhibited in the report.
Convention Paper 7920 (Purchase now)

P14-4 MDCT for Encoding Residual Signals in Frequency Domain Linear Prediction—Sriram Ganapathy, Johns Hopkins University - Baltimore, MD, USA; Petr Motlicek, Idiap Research Institute - Martigny, Switzerland; Hynek Hermansky, Johns Hopkins University - Baltimore, MD, USA
Frequency domain linear prediction (FDLP) uses autoregressive models to represent Hilbert envelopes of relatively long segments of speech/audio signals. Although the basic FDLP audio codec achieves good quality of the reconstructed signal at high bit-rates, there is a need for scaling to lower bit-rates without degrading the reconstruction quality. Here, we present a method for improving the compression efficiency of the FDLP codec by the application of the modified discrete cosine transform (MDCT) for encoding the FDLP residual signals. In the subjective and objective quality evaluations, the proposed FDLP codec provides competent quality of reconstructed signal compared to the state-of-the-art audio codecs for the 32 - 64 kbps range.
Convention Paper 7921 (Purchase now)

P14-5 State-Space Biquad Filters with Low Noise and Improved Efficiency for Pipelined DSPs—David McGrath, Dolby Laboratories - Sydney, NSW, Australia
A State-Space filter structure is presented, along with simplified equations for mapping the coefficients of arbitrary biquad filter coefficients to the State-Space structure. This procedure allows low noise implementation of an arbitrary second-order filter transfer function. A block-processing variant of the State-Space structure is described, with the added benefit that greater efficiency can be achieved on some classes of modern pipelined DSP processors.
Convention Paper 7922 (Purchase now)

P14-6 A Bandlimited Oscillator by Frequency-Domain Synthesis for Virtual Analog Applications—Glen Deslauriers, Colby Leider, University of Miami - Coral Gables, FL, USA
Problems posed by the bandlimited synthesis of audio signals have long been addressed by the music and audio engineering communities. However, few of the proposed solutions have the flexibility necessary to accurately model and produce the variety of waveform functions present in an analog oscillator. Preferably, an additive technique would be employed as the ideal method of alias-free synthesis. Inverse Fourier Transform synthesis is one method that is often discussed but less-frequently utilized. Here we propose modifications to the method and implementation of Inverse Fourier Transform synthesis as a viable basis for the creation of a software oscillator for a Virtual Analog instrument. Design results show the quality to outperform a variety of currently implemented methods.
Convention Paper 7923 (Purchase now)

P14-7 Digital Simulation of Phonograph Tracking Distortion—Richard Tollerton, Isomorphic Software, Inc. - San Francisco, CA, USA
Phonograph tracking distortion results from the misalignment of a playback cartridge with respect to the cutting head. While it has been researched for decades, it remains a source of mystery: it cannot be isolated, it has not been accurately simulated, and its importance remains undecided. Here, a PCM simulation of horizontal and vertical tracking distortion of extremely high quality is presented, operating on the principle of phase modulation, allowing tracking distortion to be evaluated in isolation with real musical content. In this context, tracking distortion is equivalent to digital audio sampling jitter, with the jitter spectrum equal to the signal spectrum. Implications of this connection, as well as simulation accuracy, preliminary listening test results, and potential applications are discussed.
Convention Paper 7924 (Purchase now)