On the DOTV canvas, everything is a node: an image, a clip, a line of narration, a shot script — each one gets its own card. Wire a line between two cards and you have a production pipeline that takes an AI video from idea to finished cut. This section covers which kinds of node there are, how to create them, and what the pieces of a card are — the chapters after it are all about what one kind of node can do on its own.
| Node | What it holds | More |
|---|---|---|
| Image | Generating and editing images, the most common kind on the canvas | Chapter 3 |
| Video | Generated clips, or ones you upload | Chapter 4 |
| Audio | Voiceover and sound effects | Chapter 5 |
| Text | A piece of writing: copy, narration, a prompt | 2.2 |
| Script | A shot script that can lay out a whole storyboard in one go | Chapter 6 |
| Video editor | A multi-track timeline that joins clips into a finished cut | Chapter 7 |
| Previz | A 3D scene: place the camera and the mannequins, then take a screenshot | Chapter 8 |
| Frame study | Breaks a finished clip into four sets of references: shots, motion, music and script | 4.8 |
| Style / effect sample | The sample you picked out of the style or effect library; whatever it is wired to is what it dresses | 3.2 |
| Group | Wraps a few nodes into one group | 2.3 |
While the canvas is still empty, a row of five cards sits in the middle: Image Generation, Video Generation, Audio Generation, Script Generation and Smart edit. Click one and its node is created in the middle of the canvas — the same nodes the four routes above build, just a faster start. Script Generation goes the script route: it lands both a screenplay and a script generator node.
The first eight you can create directly, and the four ways in all do the same thing:
The last entry in the upper group, Library, is a submenu: Style library and Effect library. Pick one and a sample node lands where you opened the menu.
The lower group in the menu is for bringing in what you already have: Upload picks a file from your machine, and Pick from history lists the images, videos and speech you have generated on this machine — whatever you pick lands right where you opened the menu.
You cannot create a group here — a group is drawn around nodes you have already selected (see 2.3).
There is an easier route too: drag from the dot on the side of a node out to empty canvas, and when you let go it asks what you want to build there. The new node comes out already wired to the old one.
The "Try:" suggestions on a draft node are a shortcut, not a step you have to take — writing a prompt in the composer and submitting works just the same.
A node goes through these looks between being created and having a result:
| State | What the card looks like |
|---|---|
| Draft | A gray type icon in the middle, with a few "Try:" suggestions under it. It holds nothing yet — select it and the composer floats up below the picture; write a prompt and hit generate. |
| Generating | The picture becomes a sweeping placeholder with a pill floating in the middle: a progress bar, Generating 62%…, and Cancel. When a whole group is running, a Running badge lights up next to the title as well. |
| Done | Shows the result. Select it now and the toolbar for this kind of node floats up above the picture. |
| Failed | A warning icon, which kind of failure it was, and one line explaining it, with a Retry button under that which runs it again unchanged. Rewriting the prompt and submitting works too — no need to delete the node and build a new one. |
A line between two nodes means the first one's output is the second one's input: wire an image into a video node and that image is the first frame; wire an image into another image and it becomes a reference. How to wire, and how to run a whole set at once, is in Chapter 9.