Hi, I'm doing a project on Voice Morphing in JAVA. I am wondering if
you could help me with the concepts behind this.
Voice morphing/conversion converts the words of the source speaker to
be spoken with the voice of the target speaker.
e.g. input: Source speaker says "Hello"
input: Target speaker says "World"
output: Using the voice quality of the Target Speaker, Hello is
pronounced
What I've done so far is to get the AudioInputStream
and with the code audioInputStream.read( audioBytes );
I get the byte[] audioBytes ranging from -128 to 127. How do I use the
byte[], is it the amplitude in time domain? I am even worried if I am
asking the right questions.
Can you please explain to me the ff and with their important
parameters, as from what I've read will be useful in voice morphing:
Linear Predictive Coding(LPC)
order of the LPC
LPC residual
All-zero digital filter
All-pole digital filter
Unvoiced - white noise
pre-emphasis
de-emphasis
Windowing
LP-PSOLA
Hai..more to ask... hopefully you can clear this terms to me. and
hopefully im in the right track
Thank you very much...
-Drace