Camp · Character Design · Clay Heads & Doll-Making · Photogrammetric 3D Scanning · Mixamo Animation

Art Camp: Living Characters & Spatial Worlds

A five-day mixed-media camp where participants sculpt and paint a clay head, build an articulated fabric doll body, turn the finished character into a photogrammetric 3D scan, animate it with free motion-capture, and place it in a shared map-based world

Participants begin by sculpting and painting a clay head for their character, then spend the next stage building an articulated doll body by hand — making a wire armature, wrapping and padding it, and sewing or fitting a fabric skin over the form. Clothes are designed the next day as drawings, and the facilitator prepares those outfits for the dolls. Once the characters are complete, each one is photographed from 30–50 overlapping angles by moving around it with a phone camera. A process called photogrammetric 3D scanning then uses those photographs to automatically reconstruct the character as a textured three-dimensional digital object — a digital twin of the handmade work. That 3D scan then receives a looping animation chosen from a free online library and a home on a real or imaginary map. When someone walks to that spot (or clicks it on a screen), the animated handmade character appears in their own room and speaks.

“Think Tamagotchi meets The Sims meets Habbo Hotel: your handmade character — photographed into 3D — has a place in the world, a personality, objects it owns, and things it says when someone visits.”

Day 1: The Full Demo START HERE

The very first thing on Monday morning is a live end-to-end demonstration run by the facilitator in front of the whole group. Before anyone starts their own character, the kids see the complete journey in one exciting sequence — from a lump of clay and a wire body to an animated digital character walking around on the projector screen. This sets the stakes, builds anticipation, and answers the implicit question “what are we actually making?” before the week begins.

▶ Monday Opening Demo: The Whole Journey in 20–25 Minutes

The facilitator runs through each step live, including the head:

  1. Start with the head. Roll and shape a small clay head live, pressing in eye sockets, nose, and mouth. Hold it up and describe how a simple face becomes a character.
  2. Paint the head. Add quick paint details — cheeks, lips, eyebrows, or patterns. Show how paint changes the mood immediately.
  3. Build the wire body. Bend aluminium wire or pipe cleaners into a rough human figure — torso, two arms, two legs. Hold it up and wiggle it. “This is the bones.”
  4. Attach the head. Fix the painted clay head to the neck wire so the character suddenly feels alive.
  5. Pad the body. Wrap the wire with strips of fabric or batting to build out the body shape. “Now it has a body.”
  6. Add the fabric skin. Pull a small pre-sewn fabric tube over the torso and limbs, or use a quick stitch or wrap. “Now it has skin.”
  7. Photograph it for scanning. Stand the finished demo doll against a simple background and walk around it with the phone, taking overlapping photos from different heights. “Every angle, so the computer can see the whole thing.”
  8. Show the scan result. Pull up a pre-processed 3D scan of the same or similar character in Polycam or the Living Characters tool. Rotate it on screen. “Your character is now digital. You can put it anywhere.”
  9. Drop it in Mixamo. Load the model in Mixamo, auto-rig it, browse animations. Pick something expressive — a wave, a dance, a stomp. Play it. “Now it moves.”
  10. Place it on the map. Open the Living Characters tool on the projector. Add a character pin, paste in the GLB, hit save. The character appears standing in a room on screen, animation looping. “Now it lives somewhere.”
  11. Make it speak. Type one line of dialogue — “Hello, I’ve been waiting for you.” Click the character. It speaks. “Now it has something to say.”

Total demo time: ~20–25 minutes. After this, participants already understand the full arc of the week and are ready to start building their own.

Making the Character: Step by Step

The core craft is a mixed-media character — a painted clay head on an articulated wire-and-fabric body. The body holds poses, the head gives the strongest emotional read, and clothing is designed separately as drawings before being prepared for the dolls.

1. Sculpt the head first. Start with air-dry clay and shape a head large enough to read clearly on screen: forehead, cheeks, chin, nose, and eye sockets. The head establishes the character early, which helps every later decision about body, pose, and personality.

2. Paint the clay head. Once the surface is ready, use paint to add skin colour, features, patterns, and mood. Painted heads are more legible than unpainted clay and help the scanned model keep its personality.

3. Build the armature. Twist together lengths of aluminium wire or pipe cleaners into a figure-eight torso, two arm wires, and two leg wires. Twist all joints tightly. The armature should stand up on its own.

4. Attach the head. Secure the painted clay head to the neck wire. This joins the sculptural and soft-body parts into a single character.

5. Pad the body. Wrap the torso, upper arms, and thighs with strips of polyester batting, cotton wool, or scraps of soft fabric. Secure wrapping with thread or light stitching. This builds up the three-dimensional form that the camera will photograph.

6. Add the fabric skin. Cut simple tube shapes — one for the torso, one per limb — from cotton fabric or felt. Slip them over the padded armature and hand-stitch the seams closed, or wrap and tie them in place for a faster build.

7. Draw the clothes. On the following day, participants draw the outfits and costume ideas for their character. The facilitator prepares those clothing pieces for the dolls using fabric, felt, and simple costume elements.

8. Pose and finish. Bend the armature into a characteristic pose: arms out, mid-stride, sitting, reaching. The character will be scanned in this pose, so choose something expressive. Touch up any loose seams, paint, or attachments before photographing it.

Program Goals

Daily Schedule (Each Day)

10:00–11:00Structured workshop or activity — one focused skill per day
11:00–12:00Open studio: continue character-building, painting, making objects, or drawing
12:00–1:00Lunch
1:00–2:00Computer or scanning session: 3D scanning, animating, or building the map world
2:00–3:00Free play, testing the world, and small group finishing work

Daily Plan

Monday — Clay Heads & Character Bodies, Day 1 Morning (10:00–10:25) — The Full Demo: The facilitator runs through the entire week’s process live in front of the group — clay head to finished character to 3D scan to animated digital character speaking on the projector screen. See the Demo section above for the full script. This is the opening spectacle; it should feel like a magic trick. Morning (10:25–12:00) — Head Workshop: The group brainstorms a shared world setting — a town, a forest, a hotel, a spaceship — and each participant decides who lives there. Hands-on character-making begins with the head: each participant sculpts a clay head first, then paints it. Focus on facial structure, expression, and mood. Afternoon: Build the wire armature and begin the body. Attach heads, wrap padding, and start adding the fabric skin. Introduction to the map world on the projector — the facilitator creates two or three rooms together with the group using the Add Room modal, names them, and pins them to the map.
Tuesday — Clothing Design, Finishing & Photogrammetric 3D Scanning Morning: Participants draw clothes and costume ideas for their characters. The facilitator prepares those outfits for the dolls while participants finish bodies, hair, and small props or objects — items the character owns, holds, or keeps in their space. These props can also be scanned separately and placed as room objects at the character’s map location. Afternoon: Photogrammetric 3D scanning session. Each participant’s character is posed against a simple background and photographed from 30–50 overlapping angles by moving around it with a phone camera. Free photogrammetry software reconstructs the surface as a textured 3D mesh. The facilitator imports each finished GLB model into the tool using the character’s GLB URL field. Props are scanned the same way and added as room objects via the + Object button.
Wednesday — Animation with Mixamo Morning: Introduction to animation. The group watches examples of digital characters moving — walking, waving, dancing, fidgeting — and discusses what kind of movement fits each character’s personality. Participants draw background settings for their character’s room. Afternoon: Animation session using Mixamo. Each 3D scan is opened in Mixamo; the tool auto-fits a skeleton and lets the participant pick a looping animation. The animated character is saved back to the tool, and the facilitator sets the character’s mood and writes initial dialogue together with the participant.
Thursday — Beach Day & Backstage Editing All day: Camp beach and swimming day. No screen requirements for participants. Backstage: Jonathan and the team use this day to polish the world — fixing any 3D scan issues, adjusting model scale and position, placing room objects, writing or tidying dialogue — and then saving the whole world to the cloud via the ☁ Save button so Friday’s presentation loads fresh on any machine.
Friday — Dialogue, Final Assembly & Presentation Morning: Each participant writes or dictates what their character says when visited: a greeting, a question, a short story, or a description of what they’re doing. The facilitator enters these into the dialogue builder. Final touches on any props or background drawings. Afternoon: Walking presentation. The facilitator loads the world fresh via ☁ Load on the projector machine. One or two participants introduce the world, then the whole group moves through it together — visiting rooms, meeting their animated photogrammetric characters, and triggering dialogue. The world is exported as a .twee file as a permanent record.

The Digital Tool: How It Works

The camp uses a custom browser-based authoring tool — Living Characters. No coding is required. Everything runs in any modern browser with no installation. One facilitator machine on a projector is enough for the whole group.

A pannable, zoomable map. The tool opens on a navigable Leaflet map. Each room is pinned to a GPS coordinate with a trigger radius. Walking into a room’s radius on foot (or clicking its pin on screen) opens that room’s stage automatically. A live compass panel shows the nearest rooms and their distance so participants always know where to walk next.

A Three.js room scene. When a room is entered, the tool renders a live three-dimensional scene using WebGL (Three.js). The floor and walls are tinted to match the room’s chosen backdrop — forest, stone hall, waterside, wooden room, garden, or dark cave. A coloured mood ring glows and pulses on the floor beneath each character when they speak.

Characters as 3D models or sprites. Each character stands in their room. If the character has a GLB model (exported from Polycam or Scaniverse after photogrammetric scanning), the full textured 3D scan is placed in the scene and its embedded Mixamo animation plays on a loop. If no model is ready yet, a photo or GIF appears as a billboard sprite that always faces the viewer.

Room objects. NEW Props and scanned items can be placed directly in the room scene as interactive 3D objects. In Facilitator mode, the + Object button opens a modal where you paste a GLB URL, name the object, write a description, and set its position, rotation, and scale. Clicking any object in the scene shows its description overlay. Objects are saved alongside characters and rooms and are included in the Twee export.

Characters in multiple rooms. NEW A single character can now appear in more than one room — useful if a character travels between locations, or if a prop should be visible from two adjoining spaces. The character card shows all rooms joined by a dot (·). The multi-room chip-picker in the Edit Character modal replaces the old single dropdown.

Dialogue builder. Clicking a character in the scene opens their card: name, mood, items they own, and their full dialogue tree. Dialogue is organised into five prompt types — Hello, Question, Secret, Reaction, and Item — authored with simple text fields, no coding.

Facilitator mode vs. Visitor mode. The mode toggle switches between two views. In Facilitator mode all editing controls are visible: Add Room, Add Character, + Object, Edit, and Delete buttons. In Visitor mode all editing UI disappears and only the world is visible — ready for the group presentation on the projector.

Simulate mode. NEW The Simulate button automatically cycles through all rooms in sequence, pausing in each one, so the facilitator can demonstrate the full world without needing to walk the real GPS route.

Cloud save & load. NEW The ☁ Save and ☁ Load buttons read and write the entire world — rooms, characters, objects, and all dialogue — to a file called story/main.twee in the project’s GitHub repository. The facilitator saves Thursday night; any machine in the room loads the same world Friday morning. Binary media (photos, GIFs) stays local; only text passes through the cloud.

Twee export & import. The whole cast and all their dialogue can be exported as a standard .twee file — and reimported by dragging the file onto the map or using the file picker. This makes the world portable, archivable, and playable as a standalone Twine story independent of the authoring tool. Props and room objects are included in the export as CharacterName-object passages.

Step-by-Step: Using the Tool

1. Create rooms. Click Add Room, enter a name, drop a pin on the map (or type GPS coordinates), set a trigger radius, and choose a backdrop. The room appears as a circle on the map.

2. Add a character. Click Add Character, fill in the name, mood, items, and assign one or more rooms using the chip-picker. Paste a GLB URL (from Polycam or Scaniverse) or upload a photo/GIF. Save — the character pin appears on the map and the character stands in the room scene.

3. Write dialogue. Click the character in the room to open their card, then click the dialogue prompt buttons (Hello, Question, Secret, Reaction, Item) to write what they say for each situation.

4. Place room objects. Open a room in Facilitator mode and click + Object. Paste a GLB URL for a scanned prop, name it, write a short description, and set its position and scale in the 3D scene.

5. Save to the cloud. Enter the GitHub token once (stored locally). Click ☁ Save — the entire world is written to story/main.twee. On another machine, click ☁ Load to restore it.

6. Switch to Visitor mode. Toggle the mode switch. All editing UI hides. Walk or click through the world — enter rooms, click characters, read dialogue. Use Simulate to cycle through all rooms automatically.

7. Export the world. Click Export Twee to download a .twee file of the whole cast, rooms, and objects — a permanent, portable record of the week’s work.

What the Participants Experience

StageWhat participants doWhat the tool does
Monday AMWatch the end-to-end demo — clay head to digital characterFacilitator runs the full live demonstration on projector
Monday PMName the world and its rooms togetherFacilitator adds rooms to the map in real time on the projector
Tuesday AMDraw clothing ideas and finish charactersFacilitator prepares clothing and scanned props workflow
Tuesday PMWatch their character become a 3D objectGLB model loads in the room scene; Mixamo animation plays automatically
Wednesday PMPick an animation that fits their characterCharacter’s mood ring colour and animation update live in the scene
Thursday (backstage)Beach day — no screensFacilitator polishes, places objects, saves to cloud
Friday AMDictate what their character saysFacilitator types into the dialogue builder; character speaks immediately
Friday PMWalk through the finished world togetherVisitor mode, Simulate mode, and Twee export

What Is Photogrammetric 3D Scanning?

Photogrammetry turns ordinary photographs into a three-dimensional digital object. You place the finished character against a plain background and photograph it from 30 to 50 overlapping angles, moving around it and changing height to cover every surface — front, sides, back, above, and below where possible. Software compares all the images, finds hundreds of matching points between frames, and builds a 3D mesh that matches the real character’s shape, colour, and surface texture. The result is a digital twin of the handmade work: it can be placed in a virtual space, animated, lit, and viewed from any direction. Works well for painted clay, fabric, yarn, felt, buttons, and painted surfaces. No special hardware needed — a phone camera and free software are sufficient.

What Is Mixamo?

Mixamo is a free browser-based tool from Adobe with hundreds of motion-capture animations: walk cycles, waves, dances, idle fidgets, confident strides, and more. You upload a 3D model — including a photogrammetric scan of a handmade character with a clay head and fabric body — and Mixamo automatically fits an internal skeleton to it. You can then browse and preview any animation live on your character. For this camp, each participant picks one looping animation that matches their character’s personality. The animated GLB is then loaded directly into the room scene.

What Is Twee?

Twee is a plain-text format for interactive stories, used by Twine. Each character, room, and room object becomes a passage in the file — a small labelled block of text. The format is human-readable and portable: open it in any text editor to see the whole world. After camp, the exported .twee file is both a permanent record and a playable story: it can be compiled into a standalone HTML game using Twine.

Materials & Technology