# AI use cases for Earshot (Leonxlnx/earshot) Repo: [Leonxlnx/earshot](https://github.com/Leonxlnx/earshot) | Use cases: 4 | Page: https://dedupedai.com/use-cases/leonxlnx-earshot/ earshot is an open source, CPU-only toolkit that gives coding agents the ability to describe, judge, search and generate sound through a CLI, a Python API and an MCP server. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Pick the best sound file from a set of downloads: https://dedupedai.com/use-cases/leonxlnx-earshot/pick-best-sound-download/ - Find the exact moment of a sound in a long recording: https://dedupedai.com/use-cases/leonxlnx-earshot/find-best-moment-in-recording/ - Check a launch film's sound and sync with an agent: https://dedupedai.com/use-cases/leonxlnx-earshot/check-launch-film-sound-sync/ - Generate a sound effect in code and listen to it: https://dedupedai.com/use-cases/leonxlnx-earshot/generate-sound-effect-in-code/ ## How to pick the best sound file from a set of downloads with Earshot Repo: [Leonxlnx/earshot](https://github.com/Leonxlnx/earshot) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Rank candidate sound files against a text description and keep the closest match. **Who it is for:** A developer or agent that needs a specific sound, such as a key press or a click, and has several candidate files from a sound library. **The problem:** Choosing among many similar sound files by ear takes time, and a file name does not say how the sound really is. **The outcome:** A sorted list of candidate files by fit to the description, plus a verdict on the top choice. ### Steps 1. Run `earshot rank "a single mechanical key press" candidates/*.ogg` to sort the files by fit. (inferred, check the README) 2. Run `earshot match candidates/top.ogg "a deep thocky mechanical keyboard press" --alts "a clicky blue switch keyboard" "a mouse click"` to get the verdict for the best file. (inferred, check the README) 3. Run `earshot describe candidates/top.ogg` to read what the file sounds like in words and its measurements. (inferred, check the README) ### Prompt to paste into your agent ```text Use the earshot repo at . Run `earshot rank` over the files in for the description "", then run `earshot match` on the top file. Report the ranked list and the match verdict (strong, plausible, weak or no match) with its z score. Verify that the reported file exists before you use it. ``` ### Caveat The first call to each CLI command loads models for about 20 to 40 seconds, so use the MCP server or one Python script when you run many checks. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/leonxlnx-earshot/pick-best-sound-download/ ## How to find the exact moment of a sound in a long recording with Earshot Repo: [Leonxlnx/earshot](https://github.com/Leonxlnx/earshot) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Locate the best matching moments inside a long file and get sample trims for each. **Who it is for:** A video editor or agent cutting a film or game that has a long ambient or field recording and needs one specific sound from it. **The problem:** Scrubbing a long recording to find one clean hit or key press is slow and error prone. **The outcome:** A short list of start and end times for the best matches, ready to trim. ### Steps 1. Run `earshot find keys.ogg "a single mechanical key press" --top 4` to list the best moments. 2. Note the start and end seconds of the strongest match. 3. Run `earshot describe keys.ogg` to check the file as a whole before you trim. (inferred, check the README) ### Prompt to paste into your agent ```text Use the earshot repo at . Run `earshot find "" --top 4` and return each moment as start and end seconds with its strength. Verify that each returned interval lies inside the file length before you trim. ``` ### Caveat Results depend on the description wording, so try two or three phrasings if the top moments do not match what you hear. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/leonxlnx-earshot/find-best-moment-in-recording/ ## How to check a launch film's sound and sync with an agent with Earshot Repo: [Leonxlnx/earshot](https://github.com/Leonxlnx/earshot) | Difficulty: Intermediate | Time: 45 minutes | Basis: From the README Score a launch video's sound, find where it misses, and get timestamped fixes. **Who it is for:** A developer or agent that generated a product demo or launch film in code and needs to check whether its sound lands on the picture. **The problem:** An agent can write the code that renders a film but cannot hear whether a hit lands on its cut or whether the music swells in the right place. **The outcome:** A scorecard with a verdict, a list of fixes with timestamps, and an annotated picture of the sound. ### Steps 1. Run `earshot critique film.mp4` to get the verdict and the fixes with timestamps. 2. Run `earshot timeline film.mp4` to read the listening transcript with events, sections and sync. (inferred, check the README) 3. Run `earshot sheet film.mp4 -o sheet.png` to get an annotated picture the agent can read. (inferred, check the README) ### Prompt to paste into your agent ```text Use the earshot repo at . Run `earshot critique ` and `earshot sheet -o sheet.png`, then list the top fixes with their timestamps and the verdict. Verify that each fix time falls inside the film duration before you edit the render code. ``` ### Caveat The critique compares against a profile of 7 professional launch films, so its scores describe closeness to those films, not an absolute quality grade. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/leonxlnx-earshot/check-launch-film-sound-sync/ ## How to generate a sound effect in code and listen to it with Earshot Repo: [Leonxlnx/earshot](https://github.com/Leonxlnx/earshot) | Difficulty: Advanced | Time: 1 hour | Basis: From the README Render a sound from a recipe, score it by ear and iterate until it matches the references. **Who it is for:** A game or UI developer whose agent needs a custom key press, click or whoosh and has no sound file to start from. **The problem:** Sound effects usually come from a library, so an agent cannot tune one to the exact feel of the product. **The outcome:** A set of WAV variants with a report, a spectrogram comparison and an audition file to listen to in pairs. ### Steps 1. Run `earshot forge --list` to see the available recipes and reference sets. 2. Run `earshot forge keypress --refs path/to/real/keypresses/ --budget 144 --variants 4 --dir made/keypress` with your own reference folder. 3. Open `made/keypress/report.md` and `made/keypress/audition.wav` to review the picks against the references. ### Prompt to paste into your agent ```text Use the earshot repo at . Run `earshot forge --list`, then run `earshot forge --refs --dir `. Report the chosen WAV files and the critic's notes from report.md. Verify that the output folder contains report.json and the WAV files before you use them. ``` ### Caveat The reference recordings are licensed and not in the repo, so you must supply your own with --refs or a local registry, and the README says the model checks can be pleased without the sound getting better, so judge the results by ear and by the model-free measurements. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/leonxlnx-earshot/generate-sound-effect-in-code/