🖼️

Diffuser Design Calculator

Design a quadratic-residue diffuser (QRD): pick a prime number of wells and a design frequency to get the exact well-depth sequence, the high-frequency limit from the well width, the period size, and a depth diagram — or switch to a 2D Skyline block array.

ℹ These are the textbook Schroeder equations (dₙ = (n² mod N)·λ/2N), which give a sound starting design — but real diffusers are approximate: a finite panel diffuses less at the low end than the formula implies, periodic tiling causes lobing (optimised/modulated diffusers reduce it), and well dividers (fins) and build accuracy matter. A diffuser scatters sound to keep a room lively — it does not absorb, so use it where you want energy preserved (rear wall, or first reflections if you prefer scattering to absorption). Metric; everything runs in your browser.

How It Works

A quadratic-residue diffuser (invented by Manfred Schroeder) is a row of wells of different depths, separated by thin fins. The depths follow a number sequence that gives the reflected sound an almost-flat scattering pattern across a wide range of angles, so instead of a hard echo you get even, diffuse reflection. For a prime number of wells N, well n gets the depth dₙ = (n² mod N) · λₗₒₓ / (2N), where λₗₒₓ = c / fₗₒₓ is the wavelength at your chosen design frequency. The deepest well sets how low the diffuser still works; the well width sets the top limit, fₕₒₕₕ = c / (2·width), because once the wavelength gets close to a well’s width the wells stop behaving as designed. A 2D Skyline diffuser extends the idea to a grid of square blocks whose heights come from a 2D residue array ((x²+y²) mod N), scattering in both directions instead of just one.

The equations are exact, but a built diffuser is an approximation of the ideal. A panel of finite size loses low-frequency diffusion (you need enough periods), and tiling identical periods side by side creates lobing — concentrations of energy at certain angles — which modulated or optimised sequences are designed to avoid. Fin thickness, accurate well depths, and a rigid build all affect the result. Most importantly, a diffuser scatters energy rather than removing it: use it to keep a space sounding live and spacious (classically on the rear wall behind the listener), and reach for absorption when you actually need to reduce energy.

Frequently Asked Questions

Why must the number of wells be prime?
The quadratic-residue sequence (n² mod N) only has the flat-scattering property that makes a QRD work when N is prime. Common choices are 7, 11, 13, 17, 19 and 23.
What sets the low and high frequency limits?
The deepest well sets the low limit (your design frequency) — deeper wells reach lower. The well width sets the high limit, f_high = c/(2·width): narrower wells diffuse higher. Between those two it scatters as designed.
Diffusion or absorption — which do I need?
Absorption removes energy (shortens reverb, tames reflections); diffusion spreads energy out while keeping the room live. Treat first reflections and bass with absorption; use diffusion on the rear wall or where you want a spacious, non-dead sound.
What is lobing?
Tiling identical diffuser periods side by side makes the scattered energy bunch up at particular angles (grating lobes). Using a larger prime, fewer repeats, or a modulated/optimised sequence reduces it.
QRD or Skyline?
A 1D QRD scatters in one plane (mount the wells vertical to spread horizontally). A 2D Skyline scatters in both planes at once, but it’s harder to build and divides its scattering between both directions (so per-plane it’s less focused than a 1D QRD of the same depth). Pick 1D for a wall, 2D for a more omnidirectional effect.