In Ham Radio we normally cut off (filter out) everything below 300 hz and in SSB (voice) work it is common to filter out everything above 3000hz. This does result in noticeable distortion of the speech but it is still very clearly understandable. In contesting and other situations where the band is very crowded or there is lots of noise tighter filters are sometimes used. I have used as tight as 300-2100hz, this does produce a very significant change in the sound of the speech but the tradeoff of cutting out noise is often worth it, i.e. readability improves. The problem for speech recognition is that changing the bandwidth will probably require retraining the system. michael N6CHVwrote: > > >For reference, middle C is about 256 Hz; most men speak > >about a fifth to an octave below that, > It is remarkable that speech experts themselves cite a wide range of > ranges for the fundamental. The union of the various ranges they cite is > 50-500 Hz (just over three octaves), -- Subcription/unsubscription/info requests: send e-mail with subject of "subscribe", "unsubscribe", or "info" to
Wear-Hard Mailing List Archive (searchable): http://wearables.ml.org
From Wear-Hard Mailing list Archive (WH)
Maintained by R. Paul McCarty
Archive created with babymail