Bell and chime sound effects you generate yourself

Bells are pure resonance, which makes them forgiving to generate and easy to get subtly wrong. Most of a bell's identity is in its decay: how long it rings, how the partials beat against each other as it fades, and whether the room joins in. With the decay right, even a rough render is convincing. With it wrong, extra brightness does not help.

Listen: 10 samples from the bells & chimes collection

Every clip shows the exact prompt that produced it. The render script reads its prompts from the same file as this page. Copy one, change a word and generate your own.

  • Church bell

    via audiocpp-stable-audio-3-sfx

    single large church bell tolling, deep bronze strike with a long beating decay, stone tower and open air

  • Full peal

    via audiocpp-stable-audio-3-sfx

    full peal of church bells ringing changes, overlapping bronze tones in a village, open air, distant

  • Service bell

    via audiocpp-stable-audio-3-sfx

    hotel reception desk bell struck once, bright short ding with a clean metallic decay, lobby reverb

  • Bicycle bell

    via audiocpp-stable-audio-3-sfx

    bicycle bell thumbed twice, bright thin metallic trill, close mic, outdoor street

  • Wind chimes

    via audiocpp-stable-audio-3-sfx

    metal wind chimes moving in a light breeze, random overlapping tuned tones, outdoor porch, seamless loop

  • Boxing bell

    via audiocpp-stable-audio-3-sfx

    boxing ring bell struck three times to end a round, harsh bright clang, arena reverb with crowd behind

  • Hand bell

    via audiocpp-stable-audio-3-sfx

    brass hand bell shaken vigorously, rapid clapper strikes with a bright ringing wash, close mic

  • Ship's bell

    via audiocpp-stable-audio-3-sfx

    ship's bell struck twice on deck, bright brass ring with sea and wind behind, open outdoor

  • Singing bowl

    via audiocpp-stable-audio-3-sfx

    Tibetan singing bowl struck once and left to ring, pure sustained tone with slow beating partials, quiet room

  • Tubular chime

    via audiocpp-stable-audio-3-sfx

    orchestral tubular bell struck with a mallet, bright metallic strike with a very long shimmering decay, concert hall

Prompts that work

These models respond to physical description. Name the material, the action and the space.

  • Big bell

    single large church bell tolling, deep bronze strike with a long beating decay, stone tower and open air

    "Beating decay" names the slow interference between partials that only large bells produce. It is the strongest size cue available.

  • Small bell

    hotel reception desk bell struck once, bright short ding with a clean metallic decay, lobby reverb

    "Short" and "clean" do the same job for a small bell: no beating, no wobble, gone in a second.

  • Many bells at once

    full peal of church bells ringing changes, overlapping bronze tones in a village, open air, distant

    "Ringing changes" is the term for the pattern English peals use. It produces overlapping, non-repeating tones instead of a chord.

  • Sustained tone

    Tibetan singing bowl struck once and left to ring, pure sustained tone with slow beating partials, quiet room

    "Left to ring" tells the model not to cut the tail, and the tail carries this sound's character.

Making bells & chimes sounds that fit

Decay length is size

A hand bell rings for under a second and a church bell for ten or more, and nothing else in the sound shows scale as clearly. So give the render enough seconds for the tail to finish. A large bell rendered into a three-second clip loses the part that makes it sound large. Every big-bell prompt here asks for ten seconds or more for that reason.

Ask for beating

Real bells are not perfectly tuned. Their partials sit slightly apart, and as they decay they interfere and produce a slow wobble in the tail, which campanologists call beating. Synthetic bells lack it and sound sterile. "Beating decay" and "slow beating partials" reliably produce it, and it moves a generation from sampled keyboard toward cast bronze faster than anything else.

Place the bell somewhere

Bells are usually heard through their surroundings: a tower, a hall, a village across fields. Distance matters as much as size. The same bell is bright and metallic up close, and at two hundred metres it is a warm low tone with most of the attack gone. If a bell sounds too aggressive, move it instead of filtering it: add "distant" or "heard across fields" and the model handles the air absorption.

Why generate locally

The models run on your machine, which changes the cost, the privacy and what you can automate.

Offline after download

Weights are fetched once. After that, generation, the library and the API all work with the network off. Everything lives in one SQLite file and one folder of WAVs under ~/.soundbox, and nothing is uploaded.

Your agent can use it

Soundbox runs an MCP server in the same process, with tools for generating and listing sounds. Generated files come back as absolute paths, so an agent can copy them into a project.

Licences stated up front

Every model's terms are shown before you download it, and non-commercial weights need an explicit confirmation. Soundbox is MIT. The licence of the model that made a sound governs what you may do with it.

Connect a coding agent
claude mcp add --transport http soundbox http://127.0.0.1:17494/mcp
Generate over HTTP
curl -X POST http://127.0.0.1:17494/api/generate \
  -H 'Content-Type: application/json' \
  -d '{"model":"audiocpp-stable-audio-3-sfx","prompt":"8-bit coin pickup blip","duration_s":1}'

Frequently asked questions

How do I get a bell that is clearly enormous?
Combine a long render, an explicitly large object, and beating in the decay. "Single large church bell tolling, deep bronze strike with a long beating decay, stone tower and open air" at ten or twelve seconds gets there. If it still sounds small, the render is usually too short: the size is in the tail, and a short clip cuts it off.
Can I use these sounds in a commercial video or game?
It depends on the model that made them. Soundbox shows each model's licence before you download its weights. The clips on this page were rendered with Stable Audio 3.0 Small SFX, which uses the Stability AI Community License: commercial use is free below USD $1M annual revenue, and above that Stability requires an enterprise licence. Soundbox itself is MIT, but the audio is governed by the licence of the model that made it. Read that licence before you ship; this summary is not legal advice.
Does this work offline?
Yes, once the weights are on disk. Generation, the library and the HTTP API all run locally against one SQLite file and one folder of WAVs under ~/.soundbox. There is no account and no API key, and nothing is uploaded. The only step that uses the network is the initial model download.
What hardware does it need?
The clips here were rendered on a consumer NVIDIA GPU at roughly ten seconds of wall time per clip, through audio.cpp's quantised GGUF runtime, which needs little VRAM. CPU-only generation is supported, but expect minutes per clip instead of seconds: fine for a handful of sounds, slow for a whole library.
Can an agent generate these for me?
Yes. While the app is open, Soundbox runs an MCP server at http://127.0.0.1:17494/mcp with tools for generating, listing and mixing sounds. Every generated file comes back as an absolute path, so a coding agent can render the sound a scene needs and put it straight into the project. The several hundred clips across these pages were made this way.