**This is an old revision of the document!**

The note in string instruments is generated by exciting a vibration and promoting a particular vibration in the string.

String instruments also use the body of the instrument to amplify the sound. We can see the standing wave patterns of objects with Chladni Patterns. Some examples on a violin and a guitar. And now with lasers.

For a great page on music acoustics: http://www.phys.unsw.edu.au/music/

Ruben's tube, invented by Heinrich Rubens in 1905, like the Shive Wave Machine, is really only useful at demonstrating wave concepts. But it's very good at that!

To understand why the flames are lower where the pressure varies more we need to realize that the velocity at which gas flows out is proportional to the square root of the pressure difference. This comes from Bernoulli's principle.

It's worth noting that as the speed of sound is $v=\sqrt{\frac{B}{\rho}}$ and the density $\rho$ can be approximated as the propane pressure the standing wave frequencies depend on the propane pressure and will not be the same as the frequencies when the tube is just full of air at external pressure.

Sound is generated by an oscillation and propagated as a longitudinal wave or pressure wave.

We can hear sounds between ~20Hz and ~20kHz. Probably you can hear higher frequency sounds than me. (It seems I can only hear to about 17kHz. In fifteen years time you may also only hear to this frequency..)

Our sensitivity to the loudness of sound is logarithmic, a sound that is ten time as intense sounds only twice as loud to us. The sound level $\beta$ is thus measured on a logarithmic scale in decibels is

$\beta=10\log_{10}\frac{I}{I_{0}}$

$I_{0}$ is the weakest sound intensity we can hear $I_{0}=1.0\times 10^{-12}\mathrm{W/m^{2}}$

Some examples of different sounds loudness in decibels can be found here.

Our hearing is not equally sensitive to all frequencies, you can test your hearing here

Beats occur when two waves with frequencies close to one another interfere.

If the two waves are described by

$D_{1}=A\sin2\pi f_{1}t$

and

$D_{2}=A\sin2\pi f_{2}t$

$D=D_{1}+D_{2}$

Using $\sin\theta_{1}+\sin\theta_{2}=2\sin\frac{1}{2}(\theta_{1}+\theta_{2})\cos\frac{1}{2}(\theta_{1}-\theta_{2})$

$D=2A\cos2\pi(\frac{f_{1}-f_{2}}{2})t\sin2\pi(\frac{f_{1}+f_{2}}{2})t$

A maximum in the amplitude is heard whenever $\cos2\pi(\frac{f_{1}-f_{2}}{2})t$ is equal to 1 or -1. Which gives a beat frequency of $|f_{1}-f_{2}|$.

Typically when two tones are seperated by less than about 30-40Hz we hear beating, if the separation is more than that they tend to sound like to different tones. (You can try a similar experiment at a higher frequency at Beats from Physclips).

The equation for a traveling wave

$D(x,t)=A\sin(\frac{2\pi}{\lambda}x-\frac{2\pi t}{T})=A\sin(kx-\omega t)$

tells us that if we are standing a distance $r_{1}(x)$ from a one dimensional wave the wave displacement at a time $t$ will be

$D_{1}(x,t)=A\sin(kr_{1}(x)-\omega t)$

if a second identical source is a distance $r_{2}(x)$ away

$D_{2}(x,t)=A\sin(kr_{2}(x)-\omega t)$

and the total wave displacement is

$D_{1+2}(x,t)=A\sin(kr_{1}(x)-\omega t)+A\sin(kr_{1}(x)-\omega t)$

Using $\sin\theta_{1}+\sin\theta_{2}=2\sin\frac{1}{2}(\theta_{1}+\theta_{2})\cos\frac{1}{2}(\theta_{1}-\theta_{2})$

$D_{1+2}(x,t)=2A\sin(\frac{k}{2}(r_{1}(x)+r_{2}(x))-\omega t)\cos(\frac{k}{2}(r_{1}(x)-r_{2}(x)))$

We thus hear a wave that has the same wavelength and frequency as that coming from the source, but the amplitude will depend on the distance between the sources, being maximum ($2A$) when $\frac{k}{2}(r_{1}(x)-r_{2}(x))$ is 0 or a whole number multiple of $\pi$ which corresponds to the waves at that point being in phase. The amplitude is minimum ($0$) when $\frac{k}{2}(r_{1}(x)-r_{2}(x))$ is a multiple of $\frac{\pi}{2}$ which corresponds to the waves at that point being out of phase.

If the waves propagate from the source in all 3 dimensions then we need to take in to account that as we showed in lecture 27, $A\propto\frac{1}{r}$. To determine perceived loudness we need to remember that it depends logarithmically on intensity. I have factored these considerations in the following calculations I performed in Maple. The patterns are for two speakers separated by 1m.

When a source of sound moves a stationary observer hears an apparent shift in the frequency. The origin of this effect can be seen nicely in this animation from Wikipedia.

For a source moving **towards** an observer

$\lambda'=\lambda-d=\lambda-v_{source}T=\lambda-v_{source}\frac{\lambda}{v_{sound}}=\lambda(1-\frac{v_{source}}{v_{sound}})$

$f'=\frac{v_{sound}}{\lambda'}=\frac{v_{sound}}{\lambda(1-\frac{v_{source}}{v_{sound}})}=\frac{f}{(1-\frac{v_{source}}{v_{sound}})}$

If the source is moving **away** from the observer

$\lambda'=\lambda(1+\frac{v_{source}}{v_{sound}})$

$f'=\frac{f}{(1+\frac{v_{source}}{v_{sound}})}$

A doppler effect also occurs when an observer moves towards a source, but here the wavelength does not change, instead it is the effective velocity that changes and leads to an apparent change in the frequency of the sound. When the observer moves **towards** the source of the sound

$T'=\frac{\lambda}{v_{sound}+v_{obs}}$

$f'=\frac{v_{sound}+v_{obs}}{\lambda}=\frac{v_{sound}+v_{obs}}{v_{sound}}f$

When the observer moves **away** from the source of the sound

$f'=\frac{v_{sound}-v_{obs}}{\lambda}=\frac{v_{sound}-v_{obs}}{v_{sound}}f$

For an observer moving towards a source which is also moving towards it

$f'=\frac{v_{sound}+v_{obs}}{\lambda'}=\frac{v_{sound}+v_{obs}}{\lambda(1-\frac{v_{source}}{v_{sound}})}=\frac{(v_{sound}+v_{obs})v_{sound}}{\lambda(v_{sound}-v_{source})}=f\frac{(v_{sound}+v_{obs})}{(v_{sound}-v_{source})}$

A general formula for the doppler effect is

$f'=f\frac{(v_{sound}\pm v_{obs})}{(v_{sound}\mp v_{source})}$

Top part of the $\pm$ or $\mp$ sign is for a source or observer moving towards each other, the bottom part is for motion away from each other.

If we project sound at a moving object and wait for it to come back it's frequency is shifted by two doppler effects. In the first part of the process the source is stationary and the observer is moving, the moving object is hit by sound of frequency $f'$

$f'=f\frac{(v_{sound}\pm v_{obs})}{(v_{sound})}$

Now when the sound is returning to us it is a case of a moving source and stationary observer, we apply the formula for that on to the already shifted frequency $f'$ to get $f''$ the frequency we would hear

$f''=f'\frac{(v_{sound})}{(v_{sound}\mp v_{source})}=f\frac{(v_{sound})}{(v_{sound}\mp v_{source})}\frac{(v_{sound}\pm v_{obs})}{(v_{sound})}$

Of course in this case $v_{source}=v_{obs}=v_{object}$ so

$f''=f\frac{(v_{sound}\pm v_{object})}{(v_{sound}\mp v_{object})}$

This double doppler effect, used with either sound (sonar) or radio waves (radar) can be used to detect the speed of an object from the frequency of the reflected waves.

When a moving object moves faster than the speed of sound the wavefronts pile up, creating a shock wave, which is heard as a sonic boom.

A video of a sonic boom.

More on sonic booms on wikipedia.