Authors:Wun, Simon; Horner, Andrew; Wu, Bin
Affiliation:Department of Computer Science, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong
Spectral centroid and attack time are universally recognized as the two most important perceptual features of acoustic instrument tones. By measuring the strength of higher harmonics relative to lower harmonics, spectral centroid strongly correlates with a tone’s perceptual brightness. Spectral tilting can control the centroid. This study systematically investigates the influence of the spectral centroid on subjective judgments about instrument tones. Subjective tests explored the ability of listeners to detect changes in centroid, the identification of instruments with two types of tilting, and what listeners are hearing with spectral tilting. Discrimination and identification are related.
Download: PDF (HIGH Res) (4.6MB)
Download: PDF (LOW Res) (420KB)
Authors:Brandt, Matthias; Bitzer, Joerg
Affiliation:Jade University of Applied Sciences, Oldenburg, Germany
Hum (power line interference in the form of stationary sine waves and their harmonics) is often introduced into an audio signal during the recording or copying process. These disturbances can distract the listener, affect dynamics processing devices or even overload loudspeakers. The first stage in a removal process requires an accurate detection of the existence of hum and its parameters. This investigation presents an automatic method for detecting hum components with a low false alarm rate. The foundation of the detection algorithm is a statistical analysis of the short-term Fourier transform of the input signal. The algorithm was tested with both artificial signals and real recordings.
Download: PDF (HIGH Res) (3.0MB)
Download: PDF (LOW Res) (861KB)
Authors:Timoney, Joseph; Pekonen, Jussi; Lazzarini, Victor; Välimäki, Vesa
Affiliation:Maynooth University, Music Technology Research Group, Maynooth, Co. Kildare, Ireland; Independent researcher, http://pekonen.cc, Espoo, Finland; Aalto University, Department of Signal Processing and Acoustics, Aalto, Espoo, Finland
When the coefficients of a first-order allpass filter are time modulated by a periodic function, the filter behaves as a dynamic phase distortion device that produces timbral modification of its input. This filter belongs to the class of Periodic Linear Time-Varying (PLTV) systems, and thus it has different properties to the original LTI allpass filter. Such properties include a magnitude response that is not necessarily flat over all time, i.e. not allpass in the usual sense. Many interesting features of these time-varying filters can be discovered by analyzing them in the time and frequency domains. It is important to know too how the time-varying transfer function can be decomposed into recursive and nonrecursive elements, which leads to a criterion for system stability. Additionally, filter topology impacts the output wave shape; only certain configurations produce a low crest factor. These filters offer novel sonic elements for sound designers and the makers of audio effects processors.
Download: PDF (HIGH Res) (1.5MB)
Download: PDF (LOW Res) (849KB)
Authors:Nease, Stephen H.; Lanterman, Aaron D.; Hasler, Jennifer O.
Affiliation:School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA, USA
Because some audio enthusiasts argue that analog systems have more warmth than digital implementations, analog circuits are still of interest in music synthesis. The recent development of Field-Programmable Analog Arrays (FPAAs) offers a way to connect analog components together in an arbitrary fashion on a mixed-signal CMOS chip. This allows for the creation of analog synthesizers with the ease of rapid reconfigurability, a property associated with their digital counterpart. The authors use an FPAA to implement a particular voltage-controlled filter, the transistor ladder. The FPAA consists of three primary blocks: (1) the Computational Analog Block (CAB), a physical grouping of analog circuits that serve as computation elements; (2) the Switch Matrix (SM) that allows local routing between elements inside a CAB, as well as routing between CABs; and (3) the Programmer, which selects a floating-gate device in the SM and allows each devices to be turned on, off, or in between. Multiple CABs and SMs are arrayed in a single FPAA, allowing for large, reprogrammable analog systems.
Download: PDF (HIGH Res) (1.6MB)
Download: PDF (LOW Res) (407KB)
Experts on sound system optimization and tuning met at the 135th Convention, led by Bob McCarthy, to discuss the steps and procedures needed to ensure high-quality results.
Download: PDF (307KB)