<html><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /></head><body style='font-size: 10pt; font-family: Verdana,Geneva,sans-serif'>
<p>Trying to implement a classic spectral subtraction algorithm using Opus voice activity detector, and have a couple questions:</p>
<p>1.Should there be any overlap between PCM frames input to the voice activity detector, and if so, what windowing (e.g. Hamming) should be applied to the frames before being input?</p>
<p>2.Is there a threshold value for either the speech activity level of a frame or the probability of speech occurrence that is typically used to assert that speech did or did not occur in the frame?</p>
<p>Thanks</p>
</body></html>