I hope I’m explaining myself right.

I’m thinking of modulating my voice.

Thoughts? How would I go about this?

  • @darkcalling@lemmygrad.ml
    13•1 year ago

    Text to speech is the only way. However your manner of speaking even in writing could be used to fingerprint you. If you’re just making communist videos on youtube it likely doesn’t matter as the government isn’t going to go to that much effort to unmask you and unless you write or speak for a living matching your cadence, structured way of speaking, mannerisms, etc back to you as a real person will be difficult.

    If you’re not a high profile person, don’t work for the government or government cut-outs like think tanks, etc it becomes much harder for them to zero you without additional clues. Of course at that point we’re talking about other opsec like preventing them from seeing your real IP address when sharing content which is a direct line back to who you really are.

    So vary the way you write your text to speech, change it from your everyday, intentionally insert pauses, breaks, phrases you never use which suggest you’re from a different demographic, region, etc than you really are. Editing and refining your text to go into the text to speech generator can help immensely with this versus pure stream of consciousness/off-the-cuff speaking.

      • @darkcalling@lemmygrad.ml
        2•1 year ago

        To do what? Text to speech? There are a variety of tools but unfortunately I don’t have any specific recommendations. Not something I personally dabble in.

        I’d prefer something not from a big top 10 evil tech companies where possible. Definitely something local, ideally without internet access allowed to it, ideally in a sandbox/virtual machine that has no internet access period so it can’t possibly report back to daddy big tech you’re saying no-no things. In fact take the precaution of denying it internet access and using something local and it doesn’t matter if it’s from a big evil tech company.

        As to speech pattern changes. That I don’t know. Wouldn’t be surprised if there were concept tools by university students for it but personally I do it the old fashioned way with discipline and effort.

  • @pancake
    13•1 year ago

    Theoretically, yes. Speech-to-text + text-to-speech does exactly what you need, so at least it is possible. There are probably more sophisticated ways to do it, but of course, even your language, pauses, etc. could be fingerprinted in theory, so…

  • Muad'DibberM
    11•1 year ago

    I don’t think so unfortunately. If the goal is just for your voice to not be instantly recognizable tho, @Aru@lemmygrad.ml is right, just change the pitch. But if you’re trying to hide from state-level actors, they could probably easily undo whatever transformations you apply to it.

  • Rania Rudhan 🇩🇿🏳️‍⚧️
    9•1 year ago

    Ok so first you want editing software and now this? Are you going to make a youtube channel where you talk about communism but you don’t want irl people to know?

    Anyhow, just change the pitch no one is going to bother about remodulating.

  • Star Wars Enjoyer
    8•1 year ago

    There are a few means of distorting your voice for anonymity, but a sophisticated enough algorithm can reverse anything you try. The most secure thing you can do is to blend things, pitch your voice, add audio distortion, use multiple samples, shift accents / cadences / ways of speaking, change modulation at random points in the audio, etc. If you’re trying to evade the government they will eventually figure it all out, but your average schmuck will probably never get a fully clean audio out.