Soap film colours
A Socratic walk-through of soap film colours — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why does a colourless soap film show bands of colour that shift as it drains?
Hold a wire loop dipped in soapy water up to a window. The liquid is colourless, the light is white, and yet the film shows bands — magenta, gold, cyan — that slide slowly downward and then, at the very top, give way to a patch that is not merely colourless but black, seconds before the whole thing bursts. Nothing has been added. So where do the colours come from, and why does the film go dark just as it becomes thinnest?
Reasoning it through
REASONING #Begin with what the film actually does to light. It is not one surface but two: the front face where air meets water, and the back face a fraction of a micron behind it. Each face reflects a little of the incoming light — a few per cent — and lets the rest through. So every point on the film sends two reflections back at your eye, not one.
That is the whole apparatus. Two copies of the same wave, separated by a small extra journey. Ask what happens when they meet again. If the second copy has fallen behind by a whole number of wavelengths, the crests line up and that colour is reinforced. If it has fallen behind by half a wavelength, crest meets trough and that colour is cancelled. The extra journey is twice the thickness, travelled inside water, so what matters is the optical path: 2nt, with n about 1.33.
Now the crucial extra ingredient, and the one the tidy story usually omits. Reflection is not always faithful. When a wave bounces off a medium with a higher refractive index than the one it is travelling in — air into water, at the front face — it comes back flipped, half a wavelength out of step. At the back face, water into air, high to low, there is no flip. So one of your two copies has been inverted and the other has not.
Follow that to the limit and the black patch explains itself. As the film drains and the top thins to far less than a wavelength, the extra path 2nt goes to nearly nothing — but the flip remains. Two copies, identical in every respect except that one is upside down, cancel. Every colour cancels, so the film reflects essentially nothing and reads as black against the room. The blackness is not the film disappearing; it is the film becoming too thin for the path difference to compete with the phase flip.
Between those extremes the bands follow. Because of the flip, a colour is brightly reflected when 2nt equals an odd number of half-wavelengths — that is, when the thickness is about a quarter wavelength inside the water, roughly 100 nm for green. Thicker, and longer wavelengths take their turn. And since white light is a mixture, at any given thickness some wavelengths are reinforced and others suppressed at once, so what you see is white minus a bite taken out of it — which is why the bands look like magentas and golds rather than pure spectral colours.
Why do they move? Gravity drains the liquid downward, so the film is thinnest at the top and thickest at the bottom, and it keeps thinning. Each band marks a particular thickness; as the whole film thins, the thickness that produces "gold" migrates downward, so the bands drift down while the black grows from the top. And a thick film — a puddle, a windowpane — shows nothing, because once many wavelengths satisfy the condition at once the reinforced and cancelled sets interleave finer than the eye resolves and the colour washes back to white. Iridescence belongs only to films comparable to a wavelength, which is why it is rare.
The analogy
THE ANALOGY #Imagine two identical drummers marching, one a few paces behind the other, and one of them holding his drum upside down so his beat is inverted. Where the gap between them happens to be an exact number of strides, the inverted beat lands against the other's and you hear almost nothing. Change the gap slightly and one rhythm partly fills the other's silences. Bring them to zero separation and the inversion alone cancels them completely.
The drummers are two sources, whereas the film's two reflections are the same wave split and rejoined, which is why they stay in lockstep at all; and drumbeats are one rhythm while white light carries hundreds of wavelengths simultaneously, each with its own verdict at the same thickness.
Clarifying the model
THE MODEL #The most common misreading is that the film splits white light into a spectrum the way a prism does. It does not — there is no dispersion doing the work. Every wavelength travels the same geometry; they simply disagree about how many of their own wavelengths fit into it, and that disagreement is what sorts them.
The second refinement: the observed colour also depends on viewing angle, because a slanted ray takes a longer path through the film, so the true condition carries a cos of the internal angle. Tilt a bubble and the bands shift, which is a decent test that interference rather than pigment is at work.
Two honest limits. Real soap films are not simply thinning water: surfactants form the surfaces, and the very thinnest states — the black films — are stabilised structures a few tens of nanometres thick or less, where surface forces rather than gravity govern the thickness, and the details of that regime are their own research field. And the film's drainage is turbulent and uneven, which is why the bands writhe rather than sit in neat horizontal stripes.
A boundary worth marking: elsewhere in this collection, a night window turning into a mirror is explained by the same partial reflection at an interface, and colour mixing by which wavelengths are absorbed or added. Here nothing is absorbed and nothing is added — the colours are made purely by two reflections of the same wave arguing with each other.
A picture of it
THE PICTURE #How to readFollow the numbered messages downward for a single point on the film. The first dashed reply is copy A, reflected straight off the front face and inverted; the solid arrows track copy B going in, bouncing off the back face without inversion, and emerging later by the extra optical path 2nt. Both reach the eye together, and the three notes at the bottom are the only verdicts available — reinforced, cancelled, or, as thickness shrinks toward nothing, cancelled for every wavelength at once.
What became clearer
WHAT CLEARED #The colour is not in the soap. A film thinner than a wavelength returns two copies of the same wave, one inverted at the front face and one delayed by twice the thickness, and the eye receives their sum. Which wavelengths survive that sum depends on thickness alone — so a draining film, thinnest at the top and thickening downward, displays its own thickness as a map of colour. The black patch is the honest signature of the mechanism: it appears exactly where the delay vanishes and the inversion is left with nothing to cancel it out.
Where to go next
ONWARD #- Why a peacock feather or a beetle's shell keeps its iridescence permanently, using stacked layers rather than a draining film.
- How the same two-path cancellation is exploited deliberately in the anti-reflection coating on a camera lens.
Key terms
TERMS #| Term | What it means |
|---|---|
| Thin-film interference | colour produced by two reflections from the faces of a film recombining in or out of step. |
| Refractive index | how much a medium slows light; about 1.33 for water, which is why the internal path counts for more than its geometric length. |
| Phase inversion on reflection | the half-wavelength flip a wave acquires when it reflects off a medium of higher refractive index. |
| Black film | a soap film thinned to far less than a wavelength, which reflects almost nothing because the two reflections cancel at every colour. |
Every term the collection defines is gathered in the glossary.