Back to catalog
Generative mediaIntermediate

AI Video and Music Production

Making watchable clips and usable soundtracks with generative tools — and knowing what they still cannot do

AI Video and Music Production

About this course

A production course, not a tour of services. You will learn to write a prompt that describes a shot rather than a mood, to choose between text-to-video and image-to-video and understand why the second one behaves, to keep a character recognisable across cuts, to generate voice-over and music that survive being placed under picture, and to cut all of it together with ordinary footage.

Equal time goes to the limits. Clips are short, characters drift, text in frame comes out as broken letters, hands still betray the machine. Most generations are discarded, and the course says plainly how many attempts and how much money a usable minute tends to cost. Artifacts are fixed by rewriting the request, not by rescuing the frame in post.

A separate lesson deals with rights: what the terms of a service actually grant you, why machine output may not be protectable at all, and what happens when you generate a living person's face or voice, or a track in the recognisable style of a named artist. Risks are described as risks, not as permissions.

Tool names appear for the sake of concreteness and no version numbers are given. Services come and go; shot language, storyboarding, consistency discipline and licence hygiene stay.

Course curriculum

  1. What generative video actually does

    Why the toolset ages faster than the craft, where the hard limits of generated video are, and what a usable shot really costs in attempts, hours and money.

    • 1.Tools change, the craft does not10 min
    • 2.Hard limits: length, consistency, text, hands14 min

  2. The anatomy of a video prompt

    What a request for a shot consists of, why image-to-video behaves better than text-to-video, and how to turn an idea into a shot list that can actually be cut together.

    • 1.Shot, camera, light, style, pace15 min
    • 2.Text-to-video against image-to-video12 min
    • 3.Storyboard and assembly from short takes13 min

  3. Consistency and the price of a good take

    Why a character changes between shots and what actually holds it together, how to read an artifact and fix it by rewriting the request, and what a usable minute costs in attempts, hours and money.

    • 1.Keeping a character and a scene the same14 min
    • 2.Artifacts, retakes and what a minute really costs15 min

  4. Voice, music and sound

    Writing and directing a synthetic voice-over, the honest state of lipsync, prompting music that has a structure, and the sound design layer that makes generated picture stop feeling generated.

    • 1.Voice-over, speech synthesis and lipsync13 min
    • 2.Generating music: structure, genre, mood, vocals13 min
    • 3.Sound design and the mix11 min

  5. Assembly, delivery and rights

    Cutting generated material together with footage that was actually filmed, and a serious look at who owns the result, what a service licence does and does not give you, and where real faces, real voices and recognisable styles create risk.

    • 1.Cutting generated material with real footage12 min
    • 2.Rights: whose result is this16 min