You're probably from here: https://www.reddit.com/r/ElevenLabs/comments/1ue09ce/discrepancy_between_voice_design_samples_vs/ --- Voice prompt: Native English (American). Female, late thirties. Perfect audio quality. Persona: sweet, doting, and teasing voice actress. Emotion: sweet, seductive, teasing. Low pitch, extremely raspy and husky, smoky, throaty timbre with a deep vocal fry; warm and rich underneath. Unhurried, flowing cadence. Speaks with a nurturing, doting warmth that turns seductive - a sweet, sing-song, teasing lilt, her intonation rising and falling in a wide expressive range to soothe, tease, and draw him close. --- Text prompt: You want to sink deep, don't you? Mm... I know you do. [giggles] So go on... let go. Deeper now... that's it. Almost there... but not quite. [soft chuckle] Not yet. You'll relax for me, won't you? Of course you will.