Record in a cathedral without leaving the studio: how convolution reverb works
19 Nov 2026
Imagine you want a vocal or a synth to sound as if it were played in a cathedral. Going to a cathedral, setting up microphones, getting permission, carrying gear... You don't actually need any of that. Without leaving the studio you can place your sound inside a cathedral. The secret is a single hand clap. In this article we explain convolution reverb, the subject of our Studio Note video narrated by Execore, together with the idea behind it and how to use it in the studio.
What is the sound of a space?
When you speak or play an instrument in a room, what reaches your ear is not only the sound coming straight from the source. Sound spreads in every direction; it hits the walls, the ceiling, the floor, the columns, the seats, and comes back. Some of these reflections arrive very early (early reflections), others bounce many times, blend together and form a slowly fading cloud (the reverb tail). In a small studio this tail may last a tenth of a second; in a cathedral, four, five seconds or more.
What we call the "sound" of a space is the timing, strength and colour of those reflections. Stone walls carry the highs for longer; a room with curtains swallows them quickly. A long tunnel produces distinct, repeating echoes. That's why the same sound is heard completely differently in different places.
Why is a single clap enough?
To capture the whole behaviour of a space you don't need to measure every wall and surface one by one. Instead you make a very short, sharp sound in that space and record it: a hand clap, a popping balloon or a starter pistol. This short sound is an almost instantaneous "impulse" containing all frequencies at once.
Because the impulse itself is so short, everything you hear in the recording is the room's answer: the direct sound, the early reflections and the fading tail. All the echoes coming back from the walls, the ceiling and the columns are stored in that one recording. Engineers call it the impulse response, or IR for short.
In professional IR recordings a sine sweep is usually used instead of a clap: a tone gliding from low to high is played and recorded, then converted mathematically into an impulse response. This gives a cleaner, noise-free result. But the idea is the same: capture how the space answers a short sound.
A song = a stream of tiny claps
Here's the beautiful part. A digital recording is made of tens of thousands of tiny measurements per second; 44,100 per second at 44.1 kHz. You can think of each measurement as a tiny clap played at its own strength at that moment. Your song is nothing more than millions of tiny claps in a row.
What would each of those tiny claps do if it were played in the cathedral? Each would produce its own echo in the cathedral. And that echo would have the same shape as the impulse response of the clap, only louder or quieter according to the strength of the clap.
Shift, multiply, add: convolution
Three steps are enough to get the song in the cathedral:
- Shift: move the impulse response to the moment of each tiny clap.
- Multiply: scale that copy by the strength of the clap. A strong moment gives a loud echo copy; a weak moment, a quiet one.
- Add: stack all the copies on top of each other and add them up.
The result is the song as played in that cathedral. In mathematics this operation is called convolution. A computer does it millions of times in seconds. Modern software uses the Fourier transform to speed it up, but the result is the same: shift, multiply, add.
In the video we show this both visually and in the sound: the same melody plays first dry, then convolved with the impulse response. We didn't touch the melody; we only placed it in a space.
Convolution reverb in music software
The convolution reverb effects in music software do exactly this. They contain impulse responses recorded in real spaces: cathedrals, concert halls, tunnels, stairwells, old studios, even car interiors or water tanks. When you choose an IR, you "move" your sound into that space. Many DAWs include a convolution reverb, and there are free and paid IR libraries too.
How it differs from algorithmic reverb
In the studio you'll meet two main types of reverb:
- Algorithmic reverb: synthesises the reflections with mathematical models (delay lines, filters, feedback). It's flexible; you can freely change room size, decay time and density and automate them. It uses less CPU.
- Convolution reverb: uses the measured response of a real space. It's extremely realistic and carries the character of a specific space exactly. On the other hand its parameters are more limited and long IRs need more processing power.
Both are the right tool; the choice depends on what the music needs. Convolution if you want the feeling of a real space, algorithmic if you want a shapeable, "musical" space.
Tips for the studio
- Use send/return: put the reverb on a send rather than on the channel, so several instruments share the same space and the stage feels coherent. Run the reverb 100% wet on the send.
- Gain clarity with pre-delay: 20–60 ms between the dry sound and the start of the reverb keeps the vocal in front.
- Clean up the lows: high-pass the reverb return (e.g. 200–300 Hz) so the low end of the bass and kick doesn't get muddy. In club music this is almost mandatory.
- Match the tail to the tempo: in a tight house groove a four-second cathedral tail blurs everything. Long reverb in breakdowns and short reverb in the drop is a classic approach.
- Try creative IRs: an impulse response can be any sound. Using a synth note or a field recording as an IR creates unusual textures.
- Record your own IR: clap in a stairwell with your phone, trim the silence at the start and load it into your convolution reverb. You've just brought your own space into your music.
Worth knowing in the booth too
Convolution reverb is rarely used in a DJ set, but understanding reflections helps in the booth too. The reverb and echo effects on the mixer are used to widen the floor in a break or to push a track away. And the acoustics of the venue you play in are an "impulse response" too: in a big room with hard surfaces, bass and kick ring longer. That's why the sound you hear during a soundcheck in an empty venue changes when it fills up; people absorb sound and the reverb gets shorter.
At Institute of Sound
You learn reverb, send/return routing and space design in our production course, in the studio, on your own track; listening to the same sound through different impulse responses and hearing the difference with your own ears. Try it with your own ears in the studio.
Contact us for details.
Narration: Execore · Video: Institute of Sound 3D Studio Note




