As is well-known, in the measurement of low frequency sound, a main spectrum component of the wind noise exists almost within the same frequency bands as the low frequency sound. Accordingly, it is essentially difficult to eliminate the effect of wind noise by use of the analog filter based on the usual electrical circuit. Sometimes, this effect of wind noise can be partly decreased by using the windscreen or putting the microphone near the ground surface. However, they are not a drastic countermeasure. From the above point of view, in this paper, two kinds of typical countermeasure methods which are in striking contrast in removing the above wind noise are proposed, especially by focussing the central point of present study on its principle of method and practical usefulness. Concretely, two different measurement methods based on the principle of Bayes' filter and synchronous detection are employed in the low frequency sound. Both methods use basically the information of linear correlation between the actually observed value and the buried signal in the countermeasure of wind noise, and their practical estimation algorithms are explicitly proposed corresponding to the above two contrastive methods. Finally, the effectiveness of these two estimation methods of removing the wind noise is experimentally confirmed too by applying them to the actual data observed in a free sound field.
抄録全体を表示