Hi Miguel,
Voiced speech consists of fundamental frequency + harmonics called formants. So
a simple way to clear a majority of the speakers voice would be to remove
formants in the frequency domain.This would involve using either
transforms or filterbanks and then manually detecting formants and clearing
them. You would still be left with some amount of un-voiced speech or plosives.
Regards,
Amit
----- Original Message ----
From: "k...@hotmail.com"
To: a...
Sent: Friday, 13 June, 2008 5:45:26 PM
Subject: [audiodsp] vocal detection and removal
Hi everybody,
I'm a student who is working in a school project; honestly I don't
have to much experience with DSP or filters. First I want to detect the
fundamental frequency of voice in an audio file. In the second part to suppress
or remove the voice of an audio file, not necessary of a song. For example to
record a conference and you want to suppress the voice of the speaker.
I know that I can use matlab to read the audio file and store it in a data file.
But I don't know if there are some functions that allow me to extract or
suprime the audio.
Can anybody help me to orient me in my project?
Thanks and best regards
Miguel