6 min read
How to Use the selfielab Canvas: A Beginner's Guide
The selfielab canvas is an infinite board paired with a chat agent. Instead of engineering prompts from scratch, you drop your ingredients — a character, an outfit, a location — onto the board as draggable tiles, then tell the agent what you want in plain words. Every shot lands on the canvas next to your last one, so a session looks less like a chat log and more like a contact sheet from a real photoshoot. This guide walks you through your first five minutes, from uploading a face to downloading your first finished photo.
step by step
Upload your character
Open the Characters panel in the left rail and upload a clear, well-lit photo where the face is easy to see. That face becomes the reference for every shot you make — the studio keeps it consistent automatically. Your first shot is free, so there's nothing to set up before trying it.
Drop them on the canvas
Drag your character from the rail onto the board. It appears as a chip — a small tile that tells the agent who is in the next shot. Chips exist for outfits, locations, props and poses too, but a character alone is enough to start.
Describe the shot in chat
Type something like "rooftop at golden hour, candid laugh" and hit Enter. The agent composes the scene around your character's face and the photo lands on the canvas in seconds, placed near your viewport so it never gets lost.
Iterate on the result
Hover any image for its toolbar: regenerate it, copy its prompt back into chat, edit one detail, animate it into video, download it, or delete it. Select an image before your next message and the agent uses it as a reference for the next generation.
What's actually on the canvas
Two kinds of objects live on the board, and telling them apart makes everything click. Chips are inputs — tiles for your character, outfit, location, props, pose, plus icon chips for shot settings and camera motion. Result nodes are outputs — the generated images and videos, plus text labels and dialogue lines you can place anywhere. Chips shape the next generation; results are what you keep, edit and download.
Chips compose the shot
Every active chip on the canvas is folded into your next generation. A character chip supplies the face, an outfit chip the clothes, a location chip the scene. Toggle a chip off to leave it out without deleting it.
Results remember their lineage
Edits appear beside their source image, videos beside the photo they animate, screenshots beside their video. The board becomes a visual history of how each shot evolved.
The agent does the prompt work
Enhance mode is on by default: the agent refines your short description into a full prompt before generating. Turn it off when you want your exact words sent unedited.
Moving around the board
The canvas is infinite, so navigation matters. Hold Space and drag (or drag with the middle mouse button) to pan. Hold ⌘ or Ctrl and scroll to zoom toward your cursor. Press Shift+1 to fit everything in view when you lose your bearings, and ⌘0 to snap back to 100%. Type / in the chat box to open the command menu — every major action, from /image to /plan to /keybindings, is one slash away.
Where your shots go next
A finished image is a starting point, not an endpoint. Download it full-resolution from the node toolbar, animate it into a short video, run a quick edit to change the lighting or background, or select it as a reference so your next shot matches its look. Most creators keep one canvas per project — a feed, a campaign, a character — and let each board fill up like a shoot day.
quick tips
- Use a well-lit, front-facing photo for your character — the clearer the face, the more consistent every future shot will be.
- Start with a 9:16 aspect ratio (the default) if you're shooting for social feeds; switch to 1:1 or 16:9 from the chat controls when you need it.
- Press Esc any time to cancel a drag, close a preview, or clear your selection.
- Don't delete near-miss shots immediately — select one as a reference and ask the agent to fix what's wrong instead.
questions, answered
Do I need to know how to write AI prompts?+
No. Short, natural descriptions work — "coffee shop window seat, morning light" is plenty. Enhance mode expands it into a full prompt, and your chips carry the character, outfit and location details automatically.
Is the canvas free to try?+
Yes. New accounts get their first shot free, and you can explore the whole board, rail and chat before subscribing. Free accounts get one canvas; paid tiers unlock more boards, more saved assets and bigger batches.
What happens to a shot while it's generating?+
A placeholder node appears on the canvas immediately and fills in when the image is ready — usually seconds. You can keep chatting or arranging the board while it works.
Can I use the canvas on my phone?+
Yes — the studio runs in the browser, and on mobile the left rail collapses into a Panels tab. The iPhone companion app also syncs with the web studio.
keep learning
consistent characters →
Character consistency is the whole point of an AI photo studio. Here's how selfielab keeps one face across every shot — uploads, personas, styling, and reference images.
building scenes →
Chips are draggable tiles — characters, outfits, locations, props and poses — that compose your next shot. Learn how to mix, link and toggle them to build any scene.
shortcuts & workflow →
Every canvas keyboard shortcut and slash command in one place: pan, zoom, multi-select, grid arrange, and the / commands that drive images, videos, edits and plans.