Artificial intelligence
🎙️ AI for Voice and Music
Understand how synthetic voices and AI music are made, what makes scripts and prompts come out right the first time, how the results get cleaned and loudness-checked, and what consent, honest labels and the law require as of October 2026.
- 1How Synthetic Speech WorksFollow text through a modern speech model: normalization, the acoustic model, the vocoder or codec, the voice embedding that sets who is speaking, the prosody that sets how, and the speech-to-speech shortcut that borrows your own delivery.17 min
- 2Voice Studio in DepthWork every Voice Studio tool on purpose: Text2Speech settings and models, captions and video voiceover, Speech2Speech, Speech2Text as a proofreader, SoundEffects, AudioIsolation and MusicGeneration, with a credit budget worked out before you start.18 min
- 3Multi-Voice Scripts in Dialogue EditorLearn the script format Dialogue Editor reads, how voices are cast per speaker, how audio tags and stability shape delivery, how long scripts are split and joined, and how the history and log files keep every version.16 min
- 4How Synthetic Voices Read a ScriptWhy synthetic voices misread scripts and what they read reliably: the symbols normalization gets wrong, respelled names, sentences built for breath, punctuation and tags, and timing a script to a slot before a credit is spent.17 min
- 5How AI Music Generation WorksWhy music is harder than speech for a model, how research systems like Jukebox, MusicLM and MusicGen turn text into songs, what Suno, Udio, Mureka and ElevenLabs Music offer as of October 2026, and what the fight over training data is about.18 min
- 6Prompting Songs in MusicStudio and Cleaning Them in SoundStudioWhat makes a style prompt work, how section tags structure lyrics, what MusicStudio's modes, models, studio tools and stem splitting do, and SoundStudio's workflow for cleaning the result.19 min
- 7Voice Cloning Consent and the Law on Voice and LikenessWhat real consent for a voice clone covers, and the rules as of October 2026: Tennessee's ELVIS Act, California and New York digital replica and synthetic performer laws, the federal NO FAKES Act's status, the FCC's AI robocall ruling, and Idaho's election deepfake law.19 min
- 8Copyright, Platform Rules and Disclosure for AI AudioWhat you can and cannot own in an AI-assisted song or narration, where the label lawsuits against Suno and Udio stand as of October 2026, what Spotify and YouTube require, and what disclosure that travels with the audio looks like.18 min
- 9Worked Example: Inside a 60-Second Podcast IntroFollow a 60-second podcast intro from a timed script to a loudness-checked export: a synthetic narrator in Voice Studio, an AI music bed, ducking and Delivery Check in SoundStudio, disclosure in the audio and the show notes, and a project log.24 min