Tau
Voice, music and audio · 10 lessons

How to use AI voice generators and music tools

Turn text into natural speech, transcribe recordings, remove background noise and make short pieces of music. You'll also learn when copying someone's voice is fine and when it isn't.

Who it’s for: Podcasters, video makers, teachers and hobby musicians.

You no longer need a studio to make good audio. AI can turn text into natural-sounding speech, transcribe an hour of recording and clean up a noisy room. It can even write a short piece of music from a description. This course shows you how to use these tools well, and fairly.

You'll start by turning a paragraph into narration and fixing names the voice mispronounces. Then you'll transcribe a recording with speaker labels and correct the names and numbers it gets wrong. You'll rescue a noisy recording without making the voice sound underwater. You'll also edit a podcast by deleting words from its transcript. You'll direct an AI voice's pace and emphasis for an ad or a story. Then you'll make a 30-second jingle with your own lyrics.

The later lessons are about rights and consent. You'll check what a music tool's terms let you do with a song, and why imitating a named artist is risky. You'll clone a voice only with consent, starting with your own. You'll write chapters and show notes from a transcript, and dub a video into another language responsibly. It's for podcasters, video makers, teachers and hobby musicians.

By the end you’ll be able to…

  • Generate narration with control over pace, emotion and pronunciation
  • Clean background noise and edit a recording by editing its transcript
  • Create an original track and check what its licence lets you do
  • Apply consent rules before cloning or imitating any voice

Short on time? Start with the essentials

These 4 lessons cover the most useful ideas in this course. Each one takes about a minute.

Essential 1

Turn a paragraph into natural narration

Generate speech from text, choose a voice, and fix mispronounced names with respelling.

Essential 2

Transcribe an hour of audio, then check it

Produce a transcript with speaker labels and correct the names, numbers and jargon it gets wrong.

Essential 3

Rescue a noisy recording

Remove background noise and echo with AI enhancement without making the voice sound underwater.

Essential 4

Clone a voice only with consent

Clone your own voice for narration, document written consent for anyone else's, and explain why platforms verify ownership.

Every lesson in this course

This course has 10 video lessons of about a minute each, from the basics to expert habits. Each one ends with a question, three takeaways and its sources.

Foundations 2 lessons

  1. Turn a paragraph into natural narrationEssential

    Generate speech from text, choose a voice, and fix mispronounced names with respelling.

    “Robotic read? Change the punctuation, not the voice. Hear the difference.”

  2. Transcribe an hour of audio, then check itEssential

    Produce a transcript with speaker labels and correct the names, numbers and jargon it gets wrong.

    “An hour-long interview, typed out in minutes. Now find what it misheard.”

Building skills 4 lessons

  1. Rescue a noisy recordingEssential

    Remove background noise and echo with AI enhancement without making the voice sound underwater.

    “Recorded next to a humming fridge? Here's the fix, and the slider not to max out.”

  2. Edit a podcast by deleting words

    Cut filler words, long pauses and false starts by editing the transcript, then check every cut by ear.

    “Delete 'um' from the text and it disappears from the audio.”

  3. Direct an AI voice's pace, emotion and emphasis

    Use audio tags, punctuation and pacing to control delivery for an ad, a story or a lesson.

    “Same script, three reads: calm, urgent, playful. Only the tags changed.”

  4. Make a 30-second jingle from a prompt

    Write a music prompt with genre, tempo, instruments and structure, and add your own lyrics.

    “Your podcast needs an intro theme by tonight. Here's the prompt that gets it.”

Advanced 3 lessons

  1. Check what you can legally do with an AI song

    Read plan terms for commercial use, understand copyright limits, and avoid imitating named artists.

    “Made it on a free plan and now it's going viral. Can you monetise it?”

  2. Write chapters and show notes from a transcript

    Have AI propose chapter markers and show notes from a timestamped transcript, then check every timestamp, name and claim against the audio.

    “AI put chapter 4 at 31:10. The topic starts at 33:45. Check before you publish.”

Expert 1 lessons

  1. Dub your video into another language responsibly

    Translate and dub content while keeping voice and timing, have a native speaker review it, and disclose the AI dubbing.

    “Your video, now in Spanish, in your own voice. Who checks the translation?”

Apps we recommend

These are the tools this course uses, and what each one is best for. We have no partnership with any of them.

ElevenLabs

Best for: Natural narration, dubbing and voices

Expressive text-to-speech, dubbing, sound effects and music; professional voice clones require verifying it's your voice.

Freemium

Suno

Best for: Full songs with vocals from a prompt

Generates complete songs from a description or your own lyrics; paid plans grant commercial use, the free plan doesn't.

Freemium

Descript

Best for: Editing podcasts by editing text

Transcribes recordings so you cut audio by deleting words; Studio Sound removes noise and room echo.

Freemium
Hands-on project

Produce a short audio piece from recording to release

Record a 5 to 10 minute piece: a podcast segment, an audio guide for a class or a narrated story. Clean and edit it, add an intro you generated or licensed, write chapters and show notes, and confirm every voice and music track is used with consent and a licence that covers your use.

What you’ll make: A finished 5 to 10 minute audio file with chapter times, show notes and a credits note listing each voice, track and its licence.

Check your own work

  • No hiss, hum or underwater sound from over-processing is audible on headphones
  • Every cut sounds natural when played back, with no clipped words
  • Each chapter time lands within 5 seconds of the topic change
  • Every voice is your own or used with written consent from its owner
  • The music licence covers your intended use on your plan, noted with the date checked

Further reading and viewing

We’ve chosen these articles, research papers, videos and guides from other trusted sources.

Guide · ElevenLabs Documentation · 8 min

Voice Cloning overview

How instant and professional cloning differ, and the verification that protects other people's voices.

Article · Federal Trade Commission · 4 min

Scammers use AI to enhance their family emergency schemes

Why a familiar voice on the phone isn't proof, and how to verify a call from family.

Guide · Suno Help Center · 4 min

Does Suno own the music I make?

Plain-English rules on who owns a generated song and when you can use it commercially.

Guide · Descript Help Center · 5 min

Studio Sound

How one-click enhancement removes noise and echo, and when to dial it back.

Article · Spotify Newsroom · 6 min

Spotify Strengthens AI Protections for Artists, Songwriters, and Producers

How a major streaming platform handles voice impersonation, AI spam and AI credits.

Questions about this course

What is the best AI voice generator?

There isn't one best tool for everyone. ElevenLabs is known for expressive text-to-speech and dubbing. Descript is popular for editing recordings by editing the transcript. Suno makes full songs from a description. The course teaches skills that work across tools, such as fixing pronunciation, directing pace and emotion, and checking what each tool's terms allow.

Is AI voice cloning legal?

Cloning your own voice is generally fine, and some tools ask you to prove the voice is yours. Cloning someone else's voice needs their clear consent, ideally in writing. Using a cloned voice to deceive people can break the law and platform rules. Laws differ between countries, so check the rules where you live before you publish.

Can I use AI-generated music commercially?

It depends on the tool and your plan. Some music generators only allow commercial use on paid plans, and free plans may not. Copyright in purely AI-made music is also uncertain in many countries. Read the terms before using a song in an ad or a video you'll earn from, and avoid prompts that imitate a named artist.

How do I remove background noise from a recording?

Many audio tools now include an AI clean-up feature. Run it on your recording, then listen carefully with headphones. If it's set too strongly, voices can sound thin or underwater, so try a lighter setting. For the best results, record in a quiet room with soft furnishings to begin with. AI can rescue a lot, but not everything.

Do I need any audio experience?

No. The first lessons start from scratch, with turning text into speech and making a transcript. Later lessons cover editing, music rights and dubbing for people who make audio regularly. Each lesson is a short video of about a minute, followed by one question and three takeaways. Your first five lessons in Tau are free.

Get early access to this course

Tau opens soon. Join the waitlist and we’ll email you as soon as you can start this course.