voice-audio-engineer

Text-to-speech generation converts written content into spoken audio with brand-appropriate voice selection. ElevenLabs-based TTS produces high-quality output suitable for podcasts, audiobooks, and voice interfaces by applying voice cloning and transformation to match your brand identity.

Loudness normalization targets LUFS ranges: -16 to -19 for podcasts and -14 for streaming platforms, with true peak management at -1 dBTP. This Skill applies compression, de-essing, and high-pass filtering at 80-100 Hz to meet technical broadcast standards consistently.

Voice cloning creates brand-consistent personas from source audio, enabling speech-to-speech transformation and dialogue mixing across projects. This approach maintains vocal identity while applying professional processing like sibilance control and background-noise isolation.

Podcast production workflows combine transcription, speaker diarization, and mastering with loudness targets and de-essing applied. This Skill integrates MCP tool workflows for editing, LUFS compliance, and noise isolation to deliver broadcast-ready audio.

Voice UI integration uses TTS and speech-to-text capabilities to power conversational agents with natural dialogue processing. ElevenLabs models support real-time synthesis and speaker recognition for interactive applications.

Audio processing combines high-pass filtering, de-essing for sibilance, background-noise isolation, and compression with configurable knee and attack/release settings. These techniques isolate clean vocal tracks and ensure professional clarity across speech content.