Posted by jrieger on May 08, 1999 at 13:35:10:
We are trying to obtain information regarding how the voice is
sampled and processed in Dragon Naturally Speaking
When we record a voice in the advanced features - you get a signal,noise, signal to noise ratio and distortion bins reading. Does anyone know the definitions for each of these? We are especially interested in the distortion bin definition since this number can greatly fluctuate and people are interpreting this to mean "a better microphone" or a "worse
voice".
Also, does anyone know what voice models (qualities as in pitch, volume,prosody, etc) Dragon used to create their product?
Does anyone perform a voice quality screen on their clients prior to using voice? or do they use Dragon Nat Speaking to determine if their voice is adequate?
Any information would be greatly appreciated.