lamella — observatory of soap films

the practice

we point serious instruments at things that refuse to last.

lamella is an observatory devoted entirely to soap films. Each membrane raised on our wires is a numbered observation: its birth is logged, its colors are read as thickness maps¹, its lifespan is measured to the millisecond, and its pop is attended by at least one staff member, as a matter of policy. The film above is real in every way that matters here — its color is computed from the same interference physics we use on the bench, and when it dies, we will write it down.

protocol 01

every film is numbered. There are no anonymous membranes at lamella. The register currently runs past twelve thousand; each entry records duration, terminal state, and any visitors present at the pop.

protocol 02

color is data. A film's iridescence is not decoration — it is a thickness map accurate to tens of nanometers, readable by eye once you learn the sequence. Our observers can estimate thickness to within one fringe at a glance.

protocol 03

the black film is sacred. When drainage thins a membrane below the shortest visible wavelength, reflection cancels and the film goes almost invisible². This is its final state. We lower our voices.

protocol 04

every pop is mourned. Briefly. Proportionally. A film that lived 27,342 milliseconds receives 27,342 milliseconds of silence, rounded down. We are precise about grief, which is the only way we know to be sincere about it.

the register

notable observations, selected from the ledger.

Twelve thousand films, each one different, each one gone. These are the entries our staff re-read on difficult days.

no.	designation & note	lifespan	terminal state
09,214	the long filmSurvived four hallway drafts, a dropped clipboard, and the entire evening shift change. Staff began speaking to it around minute five. It did not answer, which we logged as consistent with expectations.	401,208 ms	black film, then pop
10,566	the patient oneHeld the black-film state for forty-one seconds — nearly invisible, structurally immaculate, refusing to end. The night observer described it as “a held breath with a perimeter.”	88,913 ms	silent collapse
07,772	the unnamed colorFor 0.3 seconds near terminal drainage, the western quadrant reflected a hue our spectral rig logged as valid and our observers could not name. Three staff saw it. None agree on what it resembled. The register simply says: see appendix, unresolved.	31,004 ms	pop, mid-sentence
11,031	the perfect sphereMaximum deviation from ideal sphericity: 0.4 micrometers, our instrument floor. Mathematically, we cannot prove it wasn't perfect. The geometry desk still brings this up at meetings.	12,847 ms	pop, unprovoked
12,001	the loud onePopped with a report audible in two adjoining rooms. Acoustic desk estimates the film had drained into a rare high-tension configuration. Its entry in the register is written slightly larger than the others. This was not authorized, but it was not corrected either.	7,331 ms	audible event

the optics

how a film chooses its colors.

A soap film is two skins of soap holding a sheet of water a few hundred nanometers thick — thinner than a wavelength of light. Light reflects twice: once off the front face, once off the back. The two reflections recombine, and whether they agree or cancel depends on the extra distance the second one traveled.

Δ = 2 n d cos θ_t + λ/2

That path difference Δ is compared against each wavelength separately. Where Δ lands on a whole number of a wavelength, that color survives; a half number, and it is erased³. Red might cancel exactly where green constructs — so a single thickness produces a single, specific color, and a draining film becomes a moving map of its own thinness.

The membrane above runs this arithmetic honestly: six wavelengths sampled across the visible band, per pixel, per frame, with the film draining downward the whole time. Nothing is painted on. The color is the measurement.

fig. 1 — two reflections, one argument. drawn to conceptual scale; an honest scale would make the film invisible, which is rather the point.

0 nm — black film≈250 nm≈500 nm≈750 nm1000 nm

fig. 2 — the drainage sequence, computed live from the same shader arithmetic as the membrane above: reflected color versus film thickness at normal incidence. Read right to left and you are watching a film die.

the observance

we are precise about grief.

Nothing we study survives the studying. The register is therefore also an obituary, and the observatory keeps its accounts the only way it knows how — exactly.

12,408

films observed

4d 07h

silence owed & paid

401,208

longest life, ms

films that lasted

every one of them ended. attendance remains mandatory. — director of observances

footnotes

1 Thickness is inferred from fringe order. A film reads straw-yellow near 300 nm, magenta near 480 nm, teal near 550 nm; past a micrometer the fringes crowd into pastel wash and the eye gives up before the film does.
2 Below roughly 30 nm the front and back reflections cancel across the whole visible band at once. The film is still there — holding its shape, doing everything a film does — reflecting almost nothing. Visitors routinely believe it has already popped. It hears this, presumably.
3 The half-wavelength term exists because the front reflection flips its phase and the back one doesn't. Two surfaces, one rulebook, applied unevenly — most of the color in this building comes from that asymmetry.