Terms: the words help pages use for each step

Each term page explains one word as the tools' own pages use it, with the pages it appears on. As of 2026-09-25.

Fig. 1 Every page in it links back here.

Many terms are one tool's name for a step another tool calls something else.

Knowing the name a tool uses is often the fastest way into its help centre.

Each term page quotes two to four tools' pages that use the word, so the meanings can be compared side by side.

Each term page names the tools that use the word and links to their pages on the map.

1Pages here

Agent — In an AI video tool, an agent or copilot is an assistant that takes instructions in plain language and carries them out inside the tool.
Avatar — In avatar video tools, an avatar is a digital person, stock or made from someone's photo or footage
Elements — Elements are saved subjects, a character, a location or a prop, with reference images, that a tool applies whenever the element is named in a prompt.
Finishing formats — Finishing formats are file types made for editing and colour work rather than for posting: ProRes, image sequences such as PNG, and EXR.
Multi-shot — A multi-shot generation produces one clip that contains several connected shots, with the cuts planned by the model, from a single prompt.
Review link — A review link lets someone watch work in progress inside the tool, usually with comments, without it being published or downloaded.
Scene breakdown — A scene breakdown divides a script or story idea into numbered scenes, each described, before any image or video is generated.
Text to speech — Text to speech turns a written script into spoken audio in a chosen voice, stock or cloned, often with controls for tone, pace and pronunciation.
Text to video — Text to video is generating a clip directly from a written description, with no image or footage as input
Version history — Version history keeps earlier states of a project or video, so a change can be compared with what came before and undone.
Voiceover — A voiceover is speech recorded or generated separately and laid over the picture, as distinct from dialogue generated together with it.
Web link — A web link publishes a finished video to a page hosted by the tool, so it can be watched in a browser without the file being sent.

2The other indexes

3The same step, documented elsewhere

Set beside “Terms”, two passages from the guides SceneMixer publishes about its own pipeline:

The script is written from the photos and the one-line description. In step-by-step mode you can rewrite it before anything is shot.
SceneMixer, How to Make an Ad from One Photo — “Do I need a script?” (read 2026-10-13)

Every character comes out of the script read with a voice described in words — age, weight, pace, where it sits.
SceneMixer, How Character Voices Work (read 2026-10-13)