Door sound effects
Name the material, the action and the space. Door sounds show clearly how much each one changes the result.
Listen: 10 samples from the door 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.
Sliding glass door
via audiocpp-stable-audio-3-sfxsliding glass patio door rolling open along its track and stopping, close mic, domestic room
Automatic door
via audiocpp-stable-audio-3-sfxautomatic supermarket door, sensor chime then motorised glass panels sliding apart, entrance reverb
Vault door
via audiocpp-stable-audio-3-sfxheavy steel vault door, mechanism unlocking with deep bolts retracting and a slow massive swing
Screen door
via audiocpp-stable-audio-3-sfxwooden screen door pushed open on a spring and slapping shut, porch acoustics, outdoor
Heavy knock
via audiocpp-stable-audio-3-sfxthree heavy knocks on a solid wooden door, dull resonant thuds, hallway reverb
Doorbell
via audiocpp-stable-audio-3-sfxtwo-tone doorbell chime ringing inside a house, warm metallic ding dong with a decaying tail
Key unlock
via audiocpp-stable-audio-3-sfxkey inserted into a deadbolt, turning with mechanical resistance, then the door opening, close mic
Elevator doors
via audiocpp-stable-audio-3-sfxelevator doors sliding open with a soft ding, pausing, then closing with a rubber seal thump
Cell door
via audiocpp-stable-audio-3-sfxiron jail cell door sliding shut and locking with a hard metallic clang, long corridor echo
Slow squeak
via audiocpp-stable-audio-3-sfxsmall cupboard door opening very slowly on a dry hinge, sustained high squeak, quiet close mic
Prompts that work
These models respond to physical description. Name the material, the action and the space.
Wooden door open
interior wooden door opening, single slow hinge creak rising in pitch, close mic, small room reflection
"Rising in pitch" describes what a hinge does under increasing torque. Without it the model defaults to a flat creak that sounds looped.
Wooden door close
interior wooden door closing, firm push, solid low thud on latch, short room decay, dry
Name the latch as the sound source to get a percussive thud. Naming only the door gives a vague wood-on-wood knock.
Metal slam
heavy metal door slamming shut, sharp clang with a low boom, hard reverb tail, industrial space
"Clang" with "low boom" covers both the thin metal surface and the heavy mass behind it. Either word alone sounds like a cymbal or a drum.
Horror hinge
long rusty hinge creak, slow uneven pitch wobble, dry attic-sized room, unsettling
"Uneven pitch wobble" separates a worn hinge from a clean synthetic creak. Horror editors listen for it.
Latch click
small door latch click, single metallic snap, close mic, no reverb, under 200 ms
A duration cap stops the model adding a decay tail. A latch click is a single instant.
Making door sounds that fit
Separate the hinge from the impact
A door sound is usually two separate events: a continuous hinge creak while it moves, and an impact (a latch click, a slam, a thud) when it stops. Decide which one your prompt needs. "Opening" prompts should describe the creak's pitch behaviour (rising, wobbling, steady). "Closing" prompts should describe the impact's material and force (firm push or slam, wood thud or metal clang). Asking for both in one short generation tends to blur into a single unconvincing scrape.
Name the room
Doors are almost never heard dry in real life. The reflections from the space tell the listener whether they are in a hallway, an attic or outdoors. Name the room ("small room reflection", "industrial space", "outdoor open space"). Without one, the model returns an ambiguous mid-sized reflection. If you need it dry for later processing, say "dry".
Lead with the material
Wood, metal and glass doors differ in timbre and also in what they suggest about a scene. A metal door slam sounds institutional or dangerous; a wooden door closing sounds domestic; a car door thunk suggests arrival or departure. Put the material first in the prompt and the action after it.
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.
claude mcp add --transport http soundbox http://127.0.0.1:17494/mcpcurl -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
- 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 Stable Audio 3 models use the Stability AI Community License: commercial use is free below USD $1M annual revenue, and above that Stability requires an enterprise licence. ACE-Step 1.5 is MIT. MiniMax's models allow commercial use under a revenue cap and require an attribution notice. Some voice and stem-separation models are non-commercial only. 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 run offline?
- Yes, once the weights are downloaded. Soundbox runs the models locally and stores everything in one SQLite file and one folder of WAVs under ~/.soundbox. There is no account, API key or upload step. Only the initial model download needs the network.
- What hardware do I need?
- An NVIDIA GPU with 8 GB or more. CPU-only is supported, but each generation then takes minutes. That is fine for a few short clips and slow for a large batch.
- Can my coding agent generate these for me?
- Yes. While Soundbox is running it serves MCP at http://127.0.0.1:17494/mcp, with tools for generating and listing sounds. An agent can generate a sound as part of the task it is already doing and gets back an absolute file path to use in your project.
- How is this different from downloading a sample pack?
- A pack gives you someone else's front door, recorded in someone else's hallway, and its room tone may not match your scene. Generating gives you a door for your space, with the room size, material and force you specify. The open and the close come from one prompt family, so they sound like the same door. It costs time and a GPU.