American English Voices for TTS

I’m building a new American-English text-to-speech engine and need one engaging, crystal-clear voice (male or female) to supply three hours of finished audio. Once I approve your short, unpaid sample, I’ll send the full script—six distinct narrative styles that will ultimately feed the model.

The final sessions must meet the following capture specs so the data can pass straight into the pipeline without extra clean-up:

• WAV, mono, 48 kHz, 16-bit
• Peak no hotter than ‑3 dB FS, average loudness between ‑18 LUFS and ‑24 LUFS
• Background noise below ‑65 dB RMS (device self-noise below ‑85 dB)
• Reverb under 0.4 s and speech intelligibility D50 above 97 %
• Absolutely no clipping

Deliver exactly three hours of polished, style-labeled takes that reflect the six script categories I’ll provide. If you can record to spec from a treated room, deliver consistent file naming, and are comfortable signing a standard usage release for TTS, you’re the voice I want to hear first.