26 Jun 2025, 10:47
(This post was last modified: 10 Aug 2025, 16:28 by AI_addict_hypnofan.)
Elevenlabs yesterday added the ability to create new A.I. voices from text prompts.
What's exciting is you can reverse-engineer a voice using Gemini Pro 2.5 to perfectly match any script you have.
First, feed your preferred AI with the Elevenlabs guidance ("PRODUCT GUIDES - Voices - Voice design - A guide on how to craft voices from a text prompt.")
Then upload your hypnotic/femdom text.
Then ask the AI to write a voice prompt for that script, incorporating any strong preferences you have for age and accent, etc.
You may need to revise it a few times, for example to stop it being too literal (e.g. interpreting dominant as shouty, rather than seductive.)
You'll also want to make sure the voice prompt knows how to vary intonation, so the output isn't monotonous.
It looks like Elevenlabs have strong filters at the voice generation stage and it won't generate if your sample text is at all adult/suggestive. So it shouldn't be anything at all graphic.
However I think the sample text is combined with the character details to create the voice, so they do need to fit together. What I found works is using a standard hypnotic induction.
Prompt:
A 22 year old upper middle class southern English female voice with a low fundamental pitch and a consistent, soft, breathy quality. The voice has a clear, forward resonance with no vocal fry or harshness. Her pacing is generally slow and calming, but it is not monotonous; she instinctively varies her rhythm, using thoughtful, deliberate pauses to build anticipation and allow her words to settle. Similarly, her tone possesses a rich variety; it can shift seamlessly from a sincere, nurturing warmth that builds trust, to a deeper, more resonant and possessive tone when asserting control. Her inflection is fluid, but used with purpose and precision to guide emotion, rather than being performatively melodic.
Text to read:
"Focus on this. A point of light in the center of your vision. And as you watch that point, it begins to move. Not away from you, but a slow, lazy circling that begins to trace a line. A simple, perfect spiral. And the spiral begins to turn. Slowly at first. Following the easy, gentle rhythm of your own breath. Each turn pulling your focus a little deeper. Each rotation letting the outside world grow a little dimmer. A little quieter. It’s so easy to watch. So simple to follow that bright, circling line as it spins for you. And as it spins, you can feel yourself beginning to drift along with it. A gentle, pleasant pull down into the center of the spiral. Deeper and deeper. Round and around. Down and down. All the tension just unwinding. All the thoughts just spinning away. Leaving only the spiral. Only the quiet. Only this deep, soft, easy descent. Down. Down. Down…"
What's exciting is you can reverse-engineer a voice using Gemini Pro 2.5 to perfectly match any script you have.
First, feed your preferred AI with the Elevenlabs guidance ("PRODUCT GUIDES - Voices - Voice design - A guide on how to craft voices from a text prompt.")
Then upload your hypnotic/femdom text.
Then ask the AI to write a voice prompt for that script, incorporating any strong preferences you have for age and accent, etc.
You may need to revise it a few times, for example to stop it being too literal (e.g. interpreting dominant as shouty, rather than seductive.)
You'll also want to make sure the voice prompt knows how to vary intonation, so the output isn't monotonous.
It looks like Elevenlabs have strong filters at the voice generation stage and it won't generate if your sample text is at all adult/suggestive. So it shouldn't be anything at all graphic.
However I think the sample text is combined with the character details to create the voice, so they do need to fit together. What I found works is using a standard hypnotic induction.
Prompt:
A 22 year old upper middle class southern English female voice with a low fundamental pitch and a consistent, soft, breathy quality. The voice has a clear, forward resonance with no vocal fry or harshness. Her pacing is generally slow and calming, but it is not monotonous; she instinctively varies her rhythm, using thoughtful, deliberate pauses to build anticipation and allow her words to settle. Similarly, her tone possesses a rich variety; it can shift seamlessly from a sincere, nurturing warmth that builds trust, to a deeper, more resonant and possessive tone when asserting control. Her inflection is fluid, but used with purpose and precision to guide emotion, rather than being performatively melodic.
Text to read:
"Focus on this. A point of light in the center of your vision. And as you watch that point, it begins to move. Not away from you, but a slow, lazy circling that begins to trace a line. A simple, perfect spiral. And the spiral begins to turn. Slowly at first. Following the easy, gentle rhythm of your own breath. Each turn pulling your focus a little deeper. Each rotation letting the outside world grow a little dimmer. A little quieter. It’s so easy to watch. So simple to follow that bright, circling line as it spins for you. And as it spins, you can feel yourself beginning to drift along with it. A gentle, pleasant pull down into the center of the spiral. Deeper and deeper. Round and around. Down and down. All the tension just unwinding. All the thoughts just spinning away. Leaving only the spiral. Only the quiet. Only this deep, soft, easy descent. Down. Down. Down…"
