Voice change for the seek of privacy

Hi I’ve got 2 questions:

  1. Can I get true “anonymous privacy” by changing my voice through audacity? And by that I mean making my voice sound as something else (maybe synthetic) with no possible recovery of the original.


  1. If yes, which would be the steps to follow for that?

I hope I’m in the right place to ask this, and if I am, thanks in advance.

No, because even if we made you sound like Yogi Bear (and we can’t), you’d get killed the first time somebody analyzed your word use and acting.

Three sentences in, your mum would say: “Oh, hi Loc. I didn’t recognize your voice. How’s the new car holding up?”


All I want is to delete any blueprint of my voice’s biometric, is that really impossible?

Please read Koz’s reply. You can reverse your words, but someone else can reverse that to get your words back. You can use voice changing plugins (I don’t know good ones for Linux) but people who know you will still recognise your delivery and cadences.

If this is for a legal experiment, I recommend consulting your lawyer.


It isn’t for legal porpuses, I won’t go to jail if someone could get my voice back, but if you would have to say the hardest change to recover, which would be it?

That depends on the skill level of the users who will be trying to defeat your attempts to disguise your voice.

The best solution is to hire someone else to speak your words for you, or type your words into a tool like http://www.acapela-group.com/voices/demo/ then record the voice you hear.


It isn’t. It’ll only require lots and lots of work and a deeper understanding of statistics. I surely couldn’t do it.

And then, which level of obscurity do you need? And, why?

I’d advice running your text multiple times through Google Translator, then using a speech synthesizer.
Can’t guarantee you that it’ll work, but that shold seriously mess up the way you composed the text, probably partialy distorting it beyond intelligibility.