Creepy Robotic Piper TTS Female Voice
Budget / Salary$10–30
TypeFreelance project
LocationRemote
Posted1 hour ago
I need a custom female voice for Piper TTS that sounds deliberately creepy—calm, faintly metallic, almost as if an AI has gone slightly wrong. You may start from fresh recordings, an existing reference, or your own synthesis workflow; the key is that the finished voice installs cleanly in Piper and produces that unsettling robotic tone on any input text.
Please manage the whole pipeline: audio preparation, transcript alignment, model training, and basic parameter tuning. Once done, I want to drop the files into my Piper setup and instantly generate dialogue for a horror-themed project.
Deliverables
• Trained voice model compatible with Piper (e.g., .onnx + config)
• At least five short sample clips demonstrating the intended tone
• A brief README covering installation, recommended inference settings, and any licensing notes
I will test by running the model on several custom scripts; the output must match the eerie, robotic style with clear intelligibility. Let me know what source material you plan to use, your estimated training time, and any questions you have about the target vibe.
Please manage the whole pipeline: audio preparation, transcript alignment, model training, and basic parameter tuning. Once done, I want to drop the files into my Piper setup and instantly generate dialogue for a horror-themed project.
Deliverables
• Trained voice model compatible with Piper (e.g., .onnx + config)
• At least five short sample clips demonstrating the intended tone
• A brief README covering installation, recommended inference settings, and any licensing notes
I will test by running the model on several custom scripts; the output must match the eerie, robotic style with clear intelligibility. Let me know what source material you plan to use, your estimated training time, and any questions you have about the target vibe.
Apply on Freelancer →
Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.