Week 2 Tutorial: Projection mapping studio

Hands-on time turning the lecture's technique onto your own scanned object

1 March 2027

Before the session#

Bring the 3D model you scanned in week 1, exported in whatever format the digitizing station gave you.

In the session#

1. Attend the lecture on projection mapping, then move into tutorial time. The lecture explains why mapping a straight rectangle onto a curved object goes wrong and how calibration corrects for it; the tutorial is where you do that correction yourself, on the object you actually built.

2. Set up a projector in the studio, aimed at a stand or table where your week 1 object can sit still for the rest of the session. “Still” is doing real work in that sentence — a calibration is only valid for one fixed relationship between projector and object, so if the object gets bumped or the projector nudged partway through, you’re calibrating again from scratch. Settle the object on a stand you’re confident won’t move before you start step 3, not after.

3. Calibrate it against your own scanned object from week 1 rather than a flat test surface. A flat calibration grid only ever gets you a rectangle sitting flat on a rectangle — it can’t tell you anything about what happens where your object actually has depth. Work through it in two passes:

   • Corners first: match the projected image’s corners to the object’s outer edges, which gets the overall position and scale roughly right.
   • Curves and angles second: adjust for whatever a flat grid can’t account for — a curved face will smear or double an image projected straight at it, and an angled face will stretch it, so this pass is about nudging the projection, or reshaping the source image, until it sits correctly on the geometry you actually have.

   A tutor is on hand through this step because calibration goes wrong in the same couple of places for almost everyone — don’t spend twenty minutes stuck on a problem they can spot in twenty seconds.

4. Map computed imagery onto the object’s irregular geometry. This is the step that makes the calibration worth having done: once the projection is aligned to the object’s real edges and curves, whatever you project — a pattern, a texture, something that responds to code — reads as though it belongs to the object’s surface, rather than sitting on top of it as a flat, misaligned rectangle of light.

5. Capture at least one still image or clip of the mapped result. A live projection only exists in the room while it’s running, and the spec and later crits need evidence of it — treat this capture as part of the deliverable, not an afterthought, and frame it so the mapping’s alignment to the object is actually visible in the shot.

Afterwards#

You leave with a mapped projection running once on your own object — the digitizing pillar’s second technique, and one of the two pillars Studio Brief 1 (due week 4) asks you to combine with the Programming I on-ramp that starts next week.