Hi,

I'm working on a speech recognition system and I'd like to use a 24 bits audio signal as input for HCopy (HTK) in order to keep the full dynamic range. Can I avoid to downsample the signal to 16 bits before computing the MFCCS?

Thank you

--- (Edited on 12/20/2011 10:02 am [GMT-0600] by Visitor) ---