Create a vocal harmony with words and have MuseScore SING it back to you... what would it take to do this?
I've heard vocoded voices in music for years. What if MuseScore has some sort of vocoder capability that sang your song back to you?
Vocoders combine a modulator wave (something like a human voice talking) and a carrier wave (some sort of pad, or simply a sine wave) into a singing machine. What would it take incorporate this kind of capability into Musescore so that you can build an arrangement, use the lyrics tool to add lyrics and then have the program literally sing your arrangement back to you instead of playing it back with pianos. I would imagine there are several things that would need to be integrated to do this:
- Text to Speech processor to change the lyrics into a set of phonetic sounds strung together (this would become the modulator wave)
- Vocoder to combine the modulator wave with some instrument like a sine wave
How hard would it be to integrate these technologies into MuseScore? Are there freeware text to speech processors that one could integrate with Musescore?
I'm interested to hear community thoughts on this sort of thing, particularly from some of the engineers, or people who have messed with the development of either of these technologies that have thoughts on how one might go about accomplishing this. Of course if anyone wants to take this on as a plugin, that would be even better.
Comments
A quick list of existing software has been made here http://musescore.org/en/developers-handbook/google-summer-code/ideas-20…
Once MuseScore has VSTi support you will be able to use software such as Vocaloid to do this.
Vocaloid is a commercial product, however:
http://en.wikipedia.org/wiki/Vocaloid
I believe there is an open source project working along the same lines, but I'm not sure about details.