Audio noise reduction


Alexey Lukin, Maxim Tkachenko

A spectral subtraction method for reduction of a stationary noise in audio signals leaves a well-known artifact called "musical noise" — a series of random energy bursts on a time-frequency plane. We are developing 2D smoothing methods for prevention of this artifact, and using multiresolution STFT techniques to improve time and frequency resolution of the method.



Lecture slides on spectral subtraction (Power Point, in Russian) from the laboratory's introductory course on DSP.

Homework assignment on noise reduction for the course, includes programming materials.



Online Audio Denoiser (by Maxim Tkachenko), featuring multiresolution processing and musical noise suppression by a 2D smoothing.

Comparison of Noise Reduction Systems (by Alexey Lukin), on a speech sample with broadband noise and electrical buzz.



  • M.S.Tkachenko, A.S.Lukin "A multiresolution spectral subtraction algorithm for noise suppression in audio signals" // Proceedings of 12-th International Conference and Exhibition "Digital Signal Processing and its Applications" (DSPA'2010), vol. 1, p. 226. Abstract: PDF, 44 Kb, full text in Russian: PDF, 131 Kb
  • A.Lukin, J.Todd "Suppression of Musical Noise Artifacts in Audio Noise Reduction by Adaptive 2D Filtering" // 123rd Audio Engineering Society Convention, New York, USA, October 2007, paper #7168. PDF, 395 Kb
  • A.Lukin, J.Todd "Adaptive Time-Frequency Resolution for Analysis and Processing of Audio" // 120th Audio Engineering Society Convention, Paris, France, May 2006, paper #6717. PDF, 555 Kb
  • A.Lukin, D.Kalinkina, D.Kubasov "Adaptive Multiresolution Filter Banks for Image and Audio Processing" // Proceedings of GraphiCon'2005, Novosibirsk Akademgorodok, Russia, 2005, pp. 312-315. PDF, 205 Kb