ClearSpeech® Adaptive Noise Suppressor (ANS) is an algorithm designed to reduce signals that are below a given threshold. For moderate levels of both varying and stationary background noise, ANS can be used to improve the signal to noise ratio.
As the background noise changes, ANS adapts to the changes and continues to reduce the noise. The algorithm can be used either on its own or in combination with NCT’s Adaptive Speech Filter (ASF/ASFII) or Voice Recognition Enhancer (VRE) algorithms to further enhance communications.
A fact sheet that explains this technology in more detail is available to download:
(pdf format)
Adaptive Noise Suppressor (ANS)
back to top>>
ClearSpeech® Intelligent Squelch Filter (ISF) is an algorithm designed to reduce unwanted signals such as pops and pre-speech noise experienced in some radio communications.
Squelched transmissions such as amateur radio and air-ground communications can contain unwanted pops at the start and end of transmission when the ‘push-to-talk’ button is pressed. In addition, noise is often transmitted prior to the start of speech. Both the pops and pre-speech noise can be reduced by ISF improving the quality of the received signal. ISF is an adaptive algorithm with both time and frequency domain components. It operates on a single channel of audio.
A fact sheet that explains this technology in more detail is available to download:
(pdf format)
Intelligent Squelch Filter (ISF)
back to top>>
ClearSpeech® Referenced Noise Filter (RNF) is an algorithm designed to
remove noise from speech where a reference signal from the noise-producing source is available.
The reference signal can be a direct tap from the noise source or a signal from a reference microphone placed close to the localized noise source.
Effective against a wide variety of noises such as background television, radio, machine and siren noise, RNF can be used to dramatically improve the performance of communication devices such as emergency service radios and covert surveillance equipment. The algorithm can also be used for de-reverberation of Public Address (PA) systems and to enable voice control of audio systems.
By continuously tracking changes in the noise or changes in the acoustic path between the noise source and the communications microphone, this algorithm adapts for consistent performance.
A fact sheet that explains this technology in more detail is
available to download: (pdf format)
Referenced Noise Filter (RNF)
back to top>>
ClearSpeech® Speech in Noise Detector (SND) is an algorithm designed to detect speech in a wide range of noise environments.
Producing a Binary value depending on whether or not speech is detected, SND can be used as a noise-robust voice activity detector (VAD). In addition to the Binary value, SND also provides an audio output for which non-speech segments have been muted. Operating over a wide range of signal-to-noise ratios (SNRs), SND can be optimized to detect either individual words or phrases.
A fact sheet that explains this technology in more detail is available to download:
(pdf format)
Speech in Noise Detector (SND)
back to top>>
Demonstration Wave files are available
upon request for each of the above products. Please contact our
sales department for more information at
sales@nctclearspeech.com.
