Natural Science Forum / Physics / Acoustics / August 2005

Hi,

I am trying to implement an adaptive beamforming algorithm using
artificial delayed two channel data for my class project.

The sampling rate of signal is 8000Hz, I assumed the spacing between
microphones are 17cm. The target speech signal comes from the center of
microphone array (i.e., axis), and white noise comes from 30 degree to
the axis. So I delayed the white noise 2 samples for the second channel
and added it with the speech signal .But my results showed strong
residual noise around 2k-3k Hz at the microphone output.

Is there anybody who can tell me if my assumption is reseanable or not?
Since the spacing is much larger than lambda/2, will this generate
strong aliasing and generate speech distortion? Anybody can suggest me
a reasonable experimental setup for 2 channel microphone array?

Thanks a lot!