Skip to the content

Websites we've worked on:
Sound tourism map link
sound around you link
sound101 link
Dr Andy Moorhouse on piano ...

BBC radio docs on sound:

- presented by Prof Trevor Cox

Breaking glass sound?

An opera singer hits a high note and this shatters a glass in the drawing room. This may have appeared in the movies, but is it really possible to shatter glass with sound? Below are some results from an experiment carried out at the University of Salford that shows it is possible to shatter glass using sound, but it is very difficult.


glass breakingHigh speed footage of glass breaking (from our schools' site)

See Video of breaking glass using apparatus on this page (1.5M, real player needed) or (1.4M, windows media player needed

How is the glass broken?

A glass has a natural resonance, a frequency at which it will vibrate easily. A recent example of resonance was the millennium walkway in London, which oscillated alarmingly when a large number of people walked on it (this has now be cured). Blow across a beer bottle and you might get a note. This is another example of a resonance; in this case the air in the bottle neck is resonating against the spring provided by the air in the main body of the bottle. In the case of the glass we are trying to shatter, the body of the glass vibrates at resonance. If the force making the glass vibrate is big enough, the size of the vibration will become so large that the glass breaks. The most dramatic example of this was the Tacoma Narrows bridge which was oscillated by cross-winds so strongly that it broke.

Two loudspeakers, a sound level meter

Two loudspeakers, a sound level meter and the glass. Note the paper on the lip of the glass being used to indicate resonance

What type of glass should be used?

The glass should have a very strong resonance. When you ping it with a finger it should ring easily. This indicates a glass with little damping. Damping provides a force that acts to stop objects vibrating. It acts against the force provided by the sound, so damping is not wanted for this experiment as it makes the glass harder to break. A common example of damping is found in car suspension. Without damping, the car would bounce up and down. The damping in the shock absorbers stops this oscillation. In the case of the glass, however, we want a strong resonance, because then it is possible to force it to vibrate with a big displacement and then break. Although expensive crystal glass is meant to be best, because it is thin and very pure, the added lead in crystal may make it harder to break. We used quite cheap glasses because that is what we had to hand; thin is best. Cheap glasses might also have faults in them, which is very useful - see later. They should also have a very simple shape and not be embossed, as the embossing would act as cross braces and affect the resonance. We fixed the bottom of the glass to the floor, because otherwise it moved around. Provided the stem is quite heavy, this shouldn't affect the damping of the vibration of the mode at the top of the glass.

We suspect that glasses only break at the sound levels produced in this first experiment if they have faults e.g. cracks. It is from these cracks that the fracture grows and breaks the glass. The cracks can be small enough that they are not visible to the naked eye, but without these, the glasses don't break. We suspect the glass shown in the pictures and video on this page had invisible faults.

The glass breaks

How loud does the sound have to be?

This depends on the glass. But for the glass we used the sound level was in the region of 135-140dB. (We didn't have the meter exactly where the glass was so we don't have an exact measure). To generate this needs very powerful loudspeakers - we used a hefty public address system. But as we shall show on other pages, this isn't the best approach. One loudspeaker wasn't enough, so we crowded two around the glass. We also placed the glass very close to the drivers to ensure the sound was as loud as possible. We also placed the glass and loudspeakers in a Perspex box. This enclosure increases the sound level generated, and also keeps the glass from going everywhere. The sound levels generated are loud enough to be a risk to hearing. Although the box did reduce the level in our case, hearing defenders or ear plugs might be needs for operators close to the loudspeakers.

Is it safe?

Before trying this, the risks to hearing and to operators from broken glass need to be considered. Salford University accepts no responsibility for people trying this experiment, it should be left to the professionals.

Second Experiment

Click here for details of the second experiment which works better than this one

Compression-driver Experiment

This is the most reliable way of breaking glass

So can an opera singer do it?

If amplified, yes it is possible. Unamplified it is very difficult, although it was done on Myth Busters. However, you need to get the glass very close to your mouth, and having spent time picking glass out of a loudspeaker, I wouldn't risk doing it myself. I can, however, break the glass with a clarinet.

Email the team

Re-use of material permitted provided it is clearly labelled with an appropriate credit such as: "(c) University of Salford,"