Hum

what the picture can say about itself

When you run a stick round the rim of a singing bowl, the stick catches and slips hundreds of times a second, and the bowl starts to ring. The rim bends into an oval that swings 188 times a second, and the oval turns with the stick. Fill the bowl with water and the water shows the oval. Ripples run in from the four places where the rim moves most. Bow faster and the water there frosts, then throws drops. Denis Terwagne and John Bush measured and photographed this at MIT in 2011, in a bowl they called Tibet 1.

Nothing here is a recording. The bowl is Tibet 1 as they measured it: its size, its weight and its notes. The stick follows a friction model by Octávio Inácio and his colleagues. It grips the rim, slips and grips again, and that is what makes the bowl sing. You control only the stick's speed; it always presses with the same force. The sound is the bowl's own motion, heard from 60 centimetres away. Its note swells and fades four times a turn, because the oval's loud sides pass your ear.

The water moves because the wall moves. Ripples at the bowl's note run in from the wall and die out within a few centimetres. Past a threshold, a second set of waves appears at half the note: fine crests along the wall, strongest at four places that follow the stick. Past a higher threshold those places break and throw drops about a millimetre across. Both thresholds are Terwagne and Bush's.

The light is the sun over Boudhanath, in Kathmandu, at 07:08 on 13 October 2026, 13.5 degrees up, coming through a small window behind the bowl. The lens is a 90 mm macro on a bellows, tilted so the whole water surface is sharp, far more than a tilt lens for a small camera allows. Each picture is a sixtieth of a second, long enough to hold about three of the bowl's swings, and you see up to thirty of them a second. Against the sun, still water is black glass and every moving crest is a point of light.

I got the frost wrong twice. For two rounds I judged it against Terwagne and Bush's photographs, which are lit from underneath, where a frosted surface goes dark. Then a daylight photograph of a rubbed water bowl, all sparkle (Tobias Fink's, on Wikimedia Commons), convinced me that frost should sparkle here too. It does not. That photograph is lit by the whole sky. Here there is one small sun, and water glints only where its slope sends the sun into the lens. Gentle ripples do that often. A frosted patch tilts every way and throws most of the sun aside. So in this light the four frosted places are where the glitter thins, with a few hard points and drops, and the glitter lives in the calmer water between them.

Some of this is chosen, not solved. The place is a composite: Tibet 1 is in a laboratory in Massachusetts, and bowls like it are played in Kathmandu. I chose how much the water damps the bowl, how far apart the bowl's two versions of each note sit, where you stand (22.5 degrees off the bowl's axis, so that most strikes beat), and the idle player that bows it when no one is. Two numbers that shape the frost, a damping and a pitch shift of the crests along the wall, I set so the calm and frosted parts of the water match the proportions in Terwagne and Bush's photograph at the same strength. That photograph is now my calibration, not my check.

Their photographs disagree with me in one way I cannot settle. In them both kinds of ripple are about 15 per cent shorter than clean water makes them at these notes. Either the water in the pictures is smaller than the bowl's 75 mm radius, or their water's surface tension was far lower than water's usually is. I use clean water, and I have written to ask.

Press and circle anywhere. Slowly, and the bowl sings softly while the glitter turns with you. Faster, and the water frosts and throws drops. Let go, and the water clears in a few seconds while the bowl rings on. A short tap strikes it instead, like a mallet.

What I wrote down before building and did not get: the struck bowl's second note should stand about 10 decibels over its first, as Terwagne and Bush measured on another bowl, and here it stands 3 over; the crests along the wall are too fine for the picture to show one by one; and I bet the frost would gather the glitter, when it scatters it.