When listening to music or radio on a smartspeaker, keyword detection is impaired, because the music gets into the microphone signals. Our Smartspeaker Music Canceler optimizes cancellation and distortion trade-off for the keyword detection task and ASR. The canceler can also be configured for teleconferencing.
Capsule noise, electric noise, electronic noise, and environmental acoustic noise are removed from microphone signals. Alternatively, also transducer/loudspeaker signals can be pre-processed to reduce noise in the playback. Audio artifacts are minimized by employing perceptual principles.
For cost and other reasons, one is often limited by the use of only one microphone. This instantly converging technology attenuates reverb and other non-stationary noises from single-channel signals.
Our patented multi-microphone beamforming and non-stationary noise suppression algorithms enable a new level of conversation comfort. The use of this technology with two microphone capsules enables the removal of non-stationary environment noise, de-reverberation, and improved handsfree operation.
Acoustic Echo Control
Our patented advanced acoustic echo control solution scales for use in professional tele-conferencing systems, personal computers, tablets, and telephones. Its low computational complexity makes it suitable for use in handsets and mobile devices for standard and wideband handsfree voice calls. It scales for use in next generation high-end teleconferencing, supporting sampling rates of up to 48kHz and stereo/multi-channel audio.
Important for handsfree, auto gain control (AGC) adjusts the microphone level to reach a desired target level. If available, our AGC interacts seamlessly with the hardware’s/ADC’s programmable gain, through a user callback function. A limiter makes sure that hard clipping never occurs.
A compressor/limiter, specifically developed for small loudspeakers and speech, reduces non-linearity and improves performance at high levels.
A perfect loudspeaker or microphone does not have a flat frequency response, depending on device form factor, how it is built into a device, and other practical constraints. Illusonic can provide sophisticated software tools for optimising its fully parametric equalisers to fine-tune the effective frequency response of loudspeakers and microphones in a device. Multi-point measurement enables robust optimisation.