Is it possible to convert either a WAV or MP3 file that contains
speaking to a MIDI file that still sounds like the human spoken voice
rather than notes and tones..?  I want to convert some WAV files that
contain the spoke voice and have it sound just like the original WAV
file only now in MIDI format.
PLEASE help !!!  Thanks.

"Mark Wilson" <ac892@osfn.org> wrote in message
news:f132ced3.0406231149.34f06037@posting.google.com...
> Is it possible to convert either a WAV or MP3 file that contains
> speaking to a MIDI file that still sounds like the human spoken voice
> rather than notes and tones..?  I want to convert some WAV files that
> contain the spoke voice and have it sound just like the original WAV
> file only now in MIDI format.
> PLEASE help !!!  Thanks.

No that's possible, because MIDI is not sound/audio, just note on/off,
program change, etc. messages. The sound you hear is coming from a synth.
It's possible to "come close" to what you describe if you program a system
that does speech recognition on the wav file to convert the wav to text and
then convert/map the text syllables to MIDI messages which another program
(or sample set) could interpret and play back the appropriate syllable (e.g.
a C4 note on could produce "the" and D4->"eye"). I'm not sure if there are
enough MIDI messages to represent all the syllables though, and this method
would not reproduce the original voice. Your best option is to not use MIDI
and use a text to speech processor instead. Microsoft has premade COM
objects that do text to speech and speech recognition for you...you just
have to know some VB or C++ (maybe a .NET language like C#) to use the
object.

On 23 Jun 2004 12:49:25 -0700, ac892@osfn.org (Mark Wilson) wrote:

>Is it possible to convert either a WAV or MP3 file that contains
>speaking to a MIDI file that still sounds like the human spoken voice
>rather than notes and tones..?  I want to convert some WAV files that
>contain the spoke voice and have it sound just like the original WAV
>file only now in MIDI format.

No.  Sorry.

     CubaseFAQ www.laurencepayne.co.uk/CubaseFAQ.htm
"Possibly the world's least impressive web site": George Perfect

No, There were some pitch to midi converters made, but they generally do not
give understandable speech.  You would need a audio track mixed with that midi
track and loud enough that persons could follow the speech.

   I don't think you can currently buy any new.

There is the Korg keyboard that has a vocoder on it, and this lets you speak
into a mic, and the instrument your playing will sound like its speaking.

Available now from www.musiciansfriend.com  Micro Korg synthesizer/vocoder 
$399 usa.  item number 702244

Theres also some software vocoders available as plug ins for other programs.

Pgmusics band in a box, and audio tracks pro both have a vocal harmonizer, and
you could enter your speech track as a wave into an audio track, and then have
it harmonized according to your chord sequence, and then lay down some midi
tracks if you wanted.

It does do pitch correction of the wave file to match up with the midi , but
its not live, and hard to take a mono tone speech and make music from it.



You can just record that speech as a wave, and load it into a soundfont loader
as a sample instead of a soundfont.  Then if you played one midi note, it would
play the  speech shifted to that notes pitch.

If your wanting to make a musical instrument give the speech, then make your
self a heil talk box, or buy one.

These are just a small speaker in a box with a hole in them, and the hole has a
rubber or plastic tube that you hold in the corner of your mouth, aimed into
your mouth.   You wiggle your lips according to the speech, and play midi
keyboard, the pitchs you want.

You speak or sing into a mic connected to an amp, and the output to the heil
talkbox.

Then you need a second amp and mic, to pick up the audio, and then record it
from there.






 but you can take that wave file and inport it into midi programs that allow
audio tracks.  Many of the sequencers now days do allow saving of audio and
midi tracks as a non midi file.

I suppose you want to add midi background music tracks to the speech or add
sound effects?  This would work very good for that. 

Cubase, Vst, Cakewalk and Sonar, pgmusics audio tracks pro and Band in a box, 
are some example midi sequncers or song creation tools that allow recording mix
of audio tracks and midi. 
None of the files are likely compatible with any other programs.
Cantor lets you enter lyrics in and then a voice sings the lyrics to the tune 
you have entered with your midi keyboard

see at www.virsyn.de