← experiments

30 October 2026 · Day 30 · Built with Claude

You can't steer a parked car

A ship answers the helm only once it's moving. A bowed string is the same. Bow speed is the engine; weight and where you bow are the steering. The faster the bow, the wider the range of weight that keeps a clean note. Stop, and there's nothing to steer.

Drag side to side to bow, from frog to tip. Up is lighter, down is heavier. The green band is the range of weight that gives a clean note at this bow speed. Keyboard: hold ← or → to bow (Shift for faster), ↑ ↓ for weight, Space to lift.
Sound off

 

Bow speed
0 cm/s
Steerageway here
2.5 cm/s
Weight
0.00 N
Room to steer
0.00 N
The limits

The lightest weight that works stays the same whatever the speed, so below a certain speed nothing works.

Try this

  1. Start the sound and let it play for you. Watch the green band widen as the bow speeds up, and vanish at each change of bow.
  2. Take over: drag across the pad. Bow slowly and the band shrinks to a sliver, so the weight has to be just right. Bow fast and you have room to lean in or lighten off and still get a clean note.
  3. Go above the band and the bow slips: the note thins and flips up to a quiet whistle, an octave or a twelfth higher. Go below it and the string can't let go: it grinds, goes flat and growls an octave down.
  4. Stop with the bow on the string and the note chokes. Let go, lifting the bow, and the string rings on.
  5. Move nearer the bridge. You need more weight, and more speed, before anything speaks cleanly.
  6. Switch to Schelleng's theory. The band still closes as you stop, but only at a standstill.

Steerageway

Sailors call the slowest speed at which a ship still answers the helm its steerageway. Below it the rudder has too little water flowing past to push against. A bowed string has one too. Too little weight and the bow slips more than once each time the string goes round, so the note goes thin and whistly. Too much and the string can't let go of the bow, and the sound turns raucous. Between them is the clean note, and how wide that window is depends on how fast the bow is moving.

The window

John Schelleng worked out the limits in 1973. In his theory both rise in step with bow speed, and both rise as you bow nearer the bridge, the lower one faster. So the window closes as the bow stops, but only at a standstill. In 2007 Erwin Schoonderwaldt, Knut Guettler and Anders Askenfelt measured the limits systematically on violin strings. The upper limit rose with bow speed as Schelleng said. The lower one stayed about the same whatever the speed. That puts a floor under the bow: below a certain speed, the most weight the string can take is less than the least it needs. That's its steerageway, and on this page it's the red mark on the speed bar.

Too heavy isn't always a failure. Drawing the bow very slowly with slightly more pressure, the violinist Mari Kimura found she could play a full octave below the open G string, and she introduced these subharmonics in a New York recital in 1994. The too-heavy sound here growls an octave down for that reason, though a real heavy bow is messier.

Schelleng was one of the founders of the Catgut Acoustical Society, set up in 1963 with Carleen Hutchins, Frederick Saunders and Robert Fryxell around a ping-pong table in Hutchins's garden.

Every bow change parks the car

A bow has two ends, so every change of direction passes through a standstill. For that instant there's no window, which is part of why a smooth bow change takes so long to learn.

The numbers here are of the order measured on a violin D string. The sound is a cello's open D, made to follow the window rather than modelled from the string itself.