Hello, I am trying to use the hts and to train a model. I have
run the ready demo HTS-demo_CMU-ARCTIC-SLT. I want to use the other voices of arctic so I have to
create raw files form wav files and lab files from utt files. I have found form a previous email in the forum a
script for converting wav to raw files. for wav ./wav/*.wav do raw=./raw/`basename $wav .wav`.raw sox -c 1 -s -w -t wav -r 16000 $wav -c 1 -s -w -t wav -r 16000 $raw done
the only thing it does, is to change the file
extension .wav to .raw, nothing more. How can I convert the wav files to raw
files? Is there another script? In the readme file in the demo it says that the files
must be: 16kHz raw audio (2-byte short, little endian). I am working on windows and cygwin (windows use big
endian I think). Is there going to be a problem with the raw files? ======================================== Lazaridis Alexandros Electrical and Computer Engineer PhD Student Artificial Intelligence Group Fax: +30 2610 997336 ======================================== |