Site Map
In almost any application related to sound capture and audio communication, the microphone not only picks up the signal of interest, but also picks up noise that lowers the quality of the signal of interest for the far-end user. In the acoustic environment there are many sources of noise consisting of different spectrum characteristics, either time-invariant or time-varying. Noise reduction (noise control) can come in several different manners and is dependent on the information that is available:
The classical approach for a single-channel noise reduction (noise control) application is spectral subtraction. In spectral subtraction, the idea that the noise is additive and thus it can be subtracted from the disturbed signal. This concept is relatively simple but relies solely on an accurate model of the noise spectrum. Therefore, the main intellectual property of a single-channel noise cancellation system is in obtaining an accurate time-variant noise model of the system.
In more modern systems, components related to speech are also estimated and enhanced to provide a higher overall signal-to-noise ratio(SNR). For example, some speech enhancement routines estimate the pitch frequency of a voiced section of speech and then apply a comb filter to remove spectral components not associated with the pitch frequency and its harmonics.
In acoustic environments in which the sound source and microphone are located far apart, reverberation is most likely going to be a problem. The concept for dereverberation via deconvolution is to create a model of room impulse response and filter it from the received signal. The problem is much more difficult than acoustic echo cancellation because in this scenario we do not have access to the original sound source. One approach to this problem is to use linear prediction analysis. Based on the knowledge of the human speech production system, only the LP residuals will be affected by reverberation. This information can than be applied to an update of the adaptive filter for deconvolving the room impulse response from the received signal.
In multichannel scenarios, such as active noise cancellation (sometimes referred to as active noise control), usually a second microphone is placed near the noise source and far enough away from the sound source so the signal from the second microphone can be used as the noise estimate in the spectral subtraction technique described above. In beamforming the spatial relationship between the multiple microphones is used to reduce noise. If the direction of sound source relative to the microphone array is known than a beamformer can be designed to pass signals coming from the sound source and filter out signals coming from different directions. This form of noise reduction is most applicable to an acoustic situation in which multiple people are talking but only one is desired to be heard.
Addtional information regarding VOCAL's noise reduction is also available: