7.7 Modes of a rectangular cavity

A rectangular box of dimensions Lx×Ly×LzL_x \times L_y \times L_z with rigid walls supports a discrete set of standing-wave modes. Each mode is a 3-D standing wave labeled by three integers (nx,ny,nz)(n_x, n_y, n_z) specifying how many half-wavelengths fit along each axis.

The modes

Separation of variables in the wave equation, with pressure antinodes at each rigid wall, gives the mode shapes

p(nx,ny,nz)(x,y,z)  =  cos ⁣(nxπxLx)cos ⁣(nyπyLy)cos ⁣(nzπzLz),p_{(n_x, n_y, n_z)}(x, y, z) \;=\; \cos\!\left(\frac{n_x \pi x}{L_x}\right) \cos\!\left(\frac{n_y \pi y}{L_y}\right) \cos\!\left(\frac{n_z \pi z}{L_z}\right),

with nx,ny,nz=0,1,2,n_x, n_y, n_z = 0, 1, 2, \ldots Not all zero — the all-zero mode is just a uniform pressure shift, not an acoustic oscillation. The allowed frequencies are

f(nx,ny,nz)  =  c2nx2Lx2+ny2Ly2+nz2Lz2.f_{(n_x, n_y, n_z)} \;=\; \frac{c}{2}\, \sqrt{\frac{n_x^2}{L_x^2} + \frac{n_y^2}{L_y^2} + \frac{n_z^2}{L_z^2}}.

This is the eigenfrequency equation for the rectangular cavity. Each mode has its own frequency; each frequency is determined by the geometry.

Mode types

Three types of modes, distinguished by how many of the indices are zero:

In rooms, axial modes are the strongest and easiest to identify; tangential and oblique modes are weaker but more numerous at higher frequencies.

Example: a small box

A rectangular room of 5×4×35 \times 4 \times 3 m. Some low modes:

ModeFrequency (Hz)Type
(1,0,0)(1, 0, 0)343/10=34.3343/10 = 34.3axial LxL_x
(0,1,0)(0, 1, 0)343/8=42.9343/8 = 42.9axial LyL_y
(1,1,0)(1, 1, 0)54.954.9tangential
(0,0,1)(0, 0, 1)343/6=57.2343/6 = 57.2axial LzL_z
(2,0,0)(2, 0, 0)68.668.6axial
(1,0,1)(1, 0, 1)66.766.7tangential
(1,1,1)(1, 1, 1)73.073.0oblique
(2,1,0)(2, 1, 0)80.980.9tangential

The modes get dense quickly. By 200 Hz there are hundreds of modes; by 1 kHz tens of thousands.

Why these matter

In a real room, sound from a source excites all these modes. The room colours the sound: at frequencies near a mode peak, sound builds up; at frequencies between modes, sound is weaker. The boomy bass of small rooms — a corner where a 35 Hz hum from a refrigerator is uncomfortably loud — is a low-axial-mode antinode.

At low frequencies, modes are widely separated and individually identifiable. At high frequencies, they overlap into a quasi-continuous reverberant field and the discrete-mode picture stops being useful. The transition between these regimes happens at the Schroeder frequency, where the modal density (modes per Hz of bandwidth) becomes large enough that adjacent modes overlap within their damping bandwidth.

Mode density

The number of modes below frequency ff in a 3-D rectangular cavity of volume V=LxLyLzV = L_x L_y L_z, surface area SS, and total edge length LeL_e is approximately

N(f)    4πVf33c3  +  πSf24c2  +  Lef8c.N(f) \;\approx\; \frac{4\pi V f^3}{3 c^3} \;+\; \frac{\pi S f^2}{4 c^2} \;+\; \frac{L_e f}{8 c}.

The volume term dominates at high frequencies: number of modes f3\propto f^3. The mode density dN/df=4πVf2/c3+dN/df = 4\pi V f^2 / c^3 + \ldots grows as f2f^2.

For our 5×4×35 \times 4 \times 3 m room (V=60V = 60 m³):

At low frequencies, room modes dominate the perception. At high frequencies, the room is approximately a continuous diffuse field — the regime treated by statistical room acoustics. Next lesson.