Built an app that provides various audio tweak tools and runs entirely in the browser

Nico — Mon, 08 Dec 2025 16:22:26 +0000

🎧 Why I Built This

I often needed quick & dirty audio utilities. Add a burst of noise to a track, retag WAVs with metadata from another file, extract album covers, convert formats, trim silence, or generate clean waveform PNGs.

Desktop tools (like Kid3 or FFmpeg-GUI) exist… but they were to overkill and not good adjusted for my use cases, so I had to switch apps constantly.

So I made Browser Audio Tools: A Bun + TypeScript + React app that runs ffmpeg locally in a Web Worker. Everything stays on your machine.

🔥 What It Can Do

1. Add Noise + Concat

Useful when you need to slightly alter a track’s fingerprint (e.g., avoiding auto-takedowns on platforms when you already own the rights).

Pink or white noise
Duration & volume controls
Concat automatically
Output bitrate selection

2. Extract Cover Art

Pulls the embedded album image from any supported file. Saves as JPEG.

3. Retag WAV into MP3

Have a pristine WAV render but metadata only in an MP3?
Load both → the app copies all tags + artwork → exports a clean 320 kbps MP3.

Great for tidying metadata from downloaded sources (yt-dlp, SoundCloud, etc.).

4. Convert Between Formats

Supported input: WAV, FLAC, AIFF, MP3, OGG, M4A (AAC), …
Supported output:

Lossy: MP3, OGG, M4A (AAC)
Lossless: WAV, FLAC, AIFF

With controls for:

Bitrate
Sample rate
Channels

5. Retag MP3 (No Re-Encoding)

Fix ID3v2 tags instantly, Title, Artist, Album, Artwork, and more

6. Audio Trimming (Visual UI)

Select the region on a waveform or type times manually.
Options include:

Silence removal using ffmpeg’s silenceremove
Export to any supported format
Keep only the selected region

7. Waveform Generator

Create clean, customizable PNG waveform renders for your track:

Bar width/gap/height
Opacity & background color
Preset color themes
Export transparent PNG for overlays

8. Multi-File Processing

Batch convert or batch extract covers. Results come back as a ZIP.
Uses the multi-threaded ffmpeg core for faster per-file execution.

🛠 Under the Hood

Bun + React + TypeScript
@ffmpeg/ffmpeg with the multi-threaded wasm core (@ffmpeg/core-mt)
All processing inside a Web Worker
~30 MB ffmpeg bundle cached after first load
Zero back-end; the whole site is static

If you open DevTools, you can see the ffmpeg worker doing all the heavy lifting.

🌐 Try It Out

👉 Live app: https://browser-audio-tools.d3sox.me
👉 GitHub repo: https://github.com/D3SOX/browser-audio-tools

💻 Contributing

It’s MIT-licensed and set up to be easy to run locally:

bun install
bun run dev
bun run build

If you want to contribute tools, presets, UI improvements, or advanced filters, PRs are welcome!

🙏 Thanks & What’s Next

This project aims to be a focused, audio-specific toolkit, not a catch-all converter.
For general media conversions, check out the excellent open-source VERT project (vert.sh / github.com/VERT-sh/VERT).

If you have any ideas for additional tools, open an issue or just leave a comment here.

Give it a try, bookmark it, or star the repo, every bit helps!
👉 https://github.com/D3SOX/browser-audio-tools

Happy processing! 🎶

My first big open source project with a friend

Nico — Tue, 01 Jun 2021 22:29:59 +0000

Today I wanted to make a little blog post (also my first one) about my first big open source project.

The beginning

It started with me switching full time to Linux. (arch btw)
I was wondering if there was a soundboard for Linux and found that there were no good alternatives to the Windows counterparts.

Often when I realize something like this I think something like: "How can it be that no one really created this?".

That combined with the fact that I had been planning to get into C++ for a while, then led me to start the project.

First functional version

After a little research I found out that it's fairly easy to implement a soundboard on Linux with PulseAudio modules.

Here is a screenshot from the first version:

I wrote it with C++ and Qt5 as widget toolkit. I had a little help from a friend who was still using Windows at the time and is more proficient in C++ than I am.

It used system calls to the PulseAudio CLI and another CLI program called mpg123 to play the sounds.

Collaboration

Then I collaborated with him and made a new version which was a major rewrite and looked like this:

I did most of the UI work with QML and he did the backend stuff. This was also the first version to support Windows with VB-Cable.

Up until that point the logo looked like this. This was my first attempt on creating a logo using Inkscape

Next-gen version

At some time we both agreed that we want to change the UI framework again because we were not so happy with QtWidgets and QtQuick.

When I created the first version I was more into coding with languages like Java, Kotlin and despised JavaScript a bit back then.
However, by then I was a fan of Web development and JavaScript/TypeScript frontend frameworks and we decided to use a web based UI.

He coded a webview library which uses WebView2 on Windows and Webkit2GTK on Linux. I created the new UI using Vue.js and Vuetify.

This is how the current version looks like:

(We've come a long way 😮)

We also created a new logo using Canva:

This version also introduces major new features like an integrated downloader and pass-through which can be used as a workaround to share an applications' sound on Linux.

Closing words

Thank you for reading my post! I've never written a blog article before, so criticism is very much welcome.

Forem: Nico