Select a node and a card floats up below the picture — the prompt, the model, the specs for this run, and send, all on it. Almost everything on the DOTV canvas that amounts to "make me something" starts from this bottom input bar; image, video and audio each have their own version, but they look and work the same way. This section covers when it shows up, how to write the prompt, how to set the specs for this run, and what happens once you send.
Each type of node gets its own input bar: the image one has the ratio and the presets, the video one has first and last frames, the audio one has a whole parameter panel underneath. What follows uses the image bar as the example; the other types are in their own chapters.
The middle of it is the input area. Write the shot you want.
Change the Ratio on a draft node and the card above it reshapes on the spot, so you can see at a glance whether this run is landscape or portrait. On a node that has already generated, the ratio doesn't change until it regenerates.
| Control | What it's for |
|---|---|
| Model | Which model this run uses. Change the model and the credit estimate on the right changes with it. |
| Specs | Ratio · Quality · Sharpness · How many all sit behind this one button, and what the button says is the current settings. Its contents follow the model: Midjourney takes the ratio only, and older nodes still using Qwen multi-angle swap in the three camera-position dials (Orbit · Tilt · Shot scale). |
| Preset | A dozen or so ready-made presets (Multi-angle, Relight, Expression tuning…). Pick one and a preset chip appears at the front of the input box, see 3.6. |
| Style | This opens the Style library: the plaza holds published looks, and there are My favorites, Recently used and Published by me alongside it — four tabs in all. Pick one and the chip folds away, replaced by its cover thumbnail on the row above the input box — the word Style sits on it permanently, so it is both the sign that a style is applied and the way back in (click it to swap, or the × in its corner to drop it). Its style description is appended to your prompt when the image runs, and parameters like the model and the ratio switch over with it. Publish style on the image node toolbar is how things get sent here. |
| Camera | A gear setup (Camera · Lens · Focal length · Aperture) on four scroll columns, with the switch in the bottom right turning it on and off. While it is on, a camera chip appears at the front of the input box and the gear description is appended to the prompt on submit. It stacks with Style, but it is mutually exclusive with Preset — with the camera on, that button is greyed out. |
Once it's sent, the input bar stays where it is (the prompt isn't cleared), the node starts running its progress, and the picture takes over when it's done. Want another version? Edit the prompt and send again — the new image replaces the one already on the node, so if you want to keep both, press ⌘D for a copy first (see 3.2).