circulation, process log

cupping, colour, and collective sensing · SCD Seed Funding 2026

2026-08-07 · week 5

A body of noise, and a circuit with nothing plugged in

Met Khang: the physical prototype arrives next Friday (14 Aug). Two ways to keep working before it does, both without touching a soldering iron.

Shared rendering (the cup panel, the flow layer, the presence/afterglow mark) moved into src/cupView.ts so the idealised system and the noisy one stay visibly the same artwork, disagreeing only about how clean the world is.

2026-07-13 · week 2

The deadline becomes the work

A decision: Circulation will not show at VFCD in September. The whole funded window, now to 30 November, goes into building the machine fully.

A cup is held, not rushed. The manual, the technician brief, the report skeletons, and the paper draft have all been re-planned to the new horizon; the finished work then goes to the venues it deserves, ISEA and its peers, on its own terms.

2026-07-11 · week 1

Circulation made literal

Synchronisation is agreement; circulation is passage. The title's mechanic arrives in the simulator in two parts.

The mark is sampled from the steady moments of the hold, not from the release itself, so the flick of letting go (which reads as Metal) does not decide what lingers. Only held cups circulate; marks receive nothing, they only fade.

2026-07-07 · week 1

The collective layer, and a way into TouchDesigner

With one cup's logic settled, the layer that makes three cups into one is next, and it is built to keep the cups independent of it.

The architecture holds one rule: the cup decides its own colour and glows on its own; TouchDesigner sits above a system that already works, so a dropped connection dims the room but never darkens the cups.

2026-07-07 · week 1

The engine before the hardware

Eleven weeks to VFCD. Before any cup exists, the interaction itself is the thing to get right, so this week starts with the logic, not the electronics.

The one experiment that decides the design: can the pulse sensor read a usable signal from a hand that is actively gripping? That is the first thing to test on the bench. If not, the reading leans on pressure and a brief "hold and be still" phase, which is coherent with the practice anyway.