AES E-Library

Frequency-Domain Processors for Efficient Removal of Noise and Unwanted Audio Events

Filtering heavily polluted audio signals with low-cut, high-cut or notch filters helps only if the spectrum of disturbances does not overlap the desired signal. Unfortunately, this is rarely the case in real situations. Therefore using frequencydomain methods is the only way to achieve dramatic improvements when cleaning up noisy signals. Spectral denoising works well in the case of broadband, smoothly changing noise, however, the situation become hopeless when the noise character changes abruptly and the S/N ratio is around zero decibels. We present a unique tool, reNOVAtor, working in 3D domain (time, frequency, amplitude) allowing efficient removal of tenacious disturbances. An intuitive manmachine interface allows localization, identification, and very precise removal of unwanted audio events. Special functions like automatic detection of clicks, tones, and harmonics significantly accelerate the workflow. A special difference function allows monitoring of the signal portion being removed.

Authors: Hatje, Ulrich; Musialik, Christoph M.
Affiliations: Charles M. Salter Associates; EBB Consult(See document for exact affiliation information.)
AES Conference:26th International Conference: Audio Forensics in the Digital Age (July 2005)
Paper Number:3-2 Import into BibTeX
Subject:Audio Forensics in the Digital Age

Click to purchase paper or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!

This paper costs $20 for non-members, $5 for AES members and is free for E-Library subscribers.

Learn more about the AES E-Library

E-Library Location: (CD 26thPapers)   /26/aes26-000027.pdf

Start a discussion about this paper!