The voice of me.

April 21, 2009 3:07 PM

I want to produce a continuous synth tone from a sample of my own voice.

Spelled out more:
I want to take a sample of my own voice and make it so that I can play a continuous tone at whatever note I wish. Ultimately, I will have several entire audio CDs, each with one continuous tone that takes up the whole CD.

So 1. How to make my voice continuous. Can this be done with software, or should I manually stretch or copy the longest, most constant section of my voice (humming a tone I guess)?
2. How to modulate it to specific frequencies?

I have OSX and Garage Band, but pretty much all I can do is record stuff with it. Otherwise, I mess with settings aimlessly. The tools for this would preferably be free, if I don't have them already.

Also, I see another AskMe that's sort of similar, and mkight help, but this is more specific.

posted by cmoj to media & arts (3 comments total)

This would be relatively trivial to do with a microphone, recording software, and CSound, which has a steep learning curve but wouldn't be too hard to do what you want.

The traditional way is to buy a hardware synthesizer. I did something exactly like this with a MIDI controller keyboard and a synthesizer module in college.
posted by muddgirl at 3:12 PM on April 21

If you know the tones you want in advance, you could use a wave editor (Audacity is free) to record humming, find a loop point by looking at the waveform, and paste that until you've got the length you want. If you know the frequency of the initial humming, you can just pitch-shift the waveform as needed to get other tones.
posted by substars at 3:34 PM on April 21

You could also create a soundfont. Soundfonts are supported in GarageBand, and you can add loop points to a soundfont to allow you to loop your voice without having to record a super long tone.

This would also allow you to load a single "set" of sounds to one track, instead of loading several separate audio files.
posted by baxter_ilion at 7:19 PM on April 21