CAM10 - module 1: Adaptive Filter
Designed to remove noise from monophonic signals, this offers filter lengths of up to 100,000 taps (although processing constraints may limit the number actually achievable) with user control of the length up to one second. There are user controlled Attack and Release constants, and the filter coefficients are displayed graphically to help the user determine the optimum settings for each job.
There are two output modes provided: the predicted signal, and the filtered signal. You can vary the output mix between the original audio and either of these to maximise the intelligibility of the result.
The Adaptive Filter removes many forms of interference and background noise, making speech signals more listenable and more intelligible, aiding transcription and making recordings more presentable for the courtroom and juries.