Reduction of microphone wind noise using local wind velocity measurements.

Abstract

A system for quieting microphone wind noise using local wind velocity measurements is desirable. The concept is shown to be feasible, however, optimization of the cancellation system requires understanding the correlation characteristics between microphone signals (pressure) and anemometer signals (velocity). Wind data was recorded using microphones and anemometers with separation distances from 2.5 to 20 centimeters. The measurements were made in a single plane parallel to the airflow. Both the microphone pressure and wind velocity signals displayed an approximate 1/f frequency characteristic. In addition, correlation strength between wind signals (both pressure and velocity) decreased with increasing separation distance. Correlation estimates from microphone to microphone, anemometer to anemometer, and microphone to anemometer are presented. It is demonstrated that adaptive signal processing algorithms can be utilized to reduce microphone wind noise using anemometer signals as a reference.