Shrink an audio file with bitrate, sample-rate, and channel controls, and see the result before you commit. You pick the tradeoff, not a preset.
The audio compressor shrinks file size, with a live estimate of the result before you commit. Three levers do the work: bitrate, which is how much data each second gets; sample rate, which is how often the audio was measured; and channels, where stereo to mono halves the data outright. You choose the tradeoff instead of a preset choosing it for you.
Use it to get a file under an email or upload limit, to fit more onto a device, or to make something quicker to send. Worth being clear about the name: this is file compression — making the file smaller — not dynamic range compression, which is the mixing process that evens out loud and quiet. For that, the audio master is what you want. How to compress audio files covers the levers.
It re-encodes your audio as an MP3 with settings you control: a lower bitrate (the biggest lever), a lower sample rate, and optionally mixing down to mono. Bitrate is how many kilobits per second are used to store the sound — halving it roughly halves the file. The before/after panel updates live as you adjust the settings so you can see the projected size before committing, and the exact final size is shown once compression finishes.
The estimate is based on your chosen bitrate multiplied by the track length, which is how constant-bitrate MP3 size works, so it's usually very close. The actual encoded file can differ slightly because of container overhead and encoder framing, so once you compress, the panel switches from the estimate to the real measured size. The percentage-smaller figure is calculated against your original file.
For music, 192 kbps is close to transparent for most listeners and 128 kbps is a solid balance of size and quality; below that you'll start to hear artifacts, especially in cymbals and reverb tails. For spoken word — podcasts, audiobooks, voice memos — you can go much lower (96 or even 64 kbps) and switch to mono, since speech doesn't need stereo or high bitrates. Use the preview player to check the result before downloading.
It can, but bitrate is the main driver of MP3 size. Lowering the sample rate (for example from 44.1 kHz to 32 or 22.05 kHz) reduces the highest frequencies the file can represent, which helps most for voice content where those top frequencies aren't important. For music, it's usually better to leave the sample rate alone and reduce bitrate instead, since a low sample rate audibly dulls the high end.
No. Like every Vocal Cut tool, compression runs entirely in your browser — decoding uses the browser's own audio decoder and re-encoding runs locally, so the file never leaves your device and there's nothing to download first. You can prove it: load a file, disconnect from the internet, and compress — it works exactly the same. Note that Safari can't decode OGG/Opus input files (a Safari limitation); MP3, WAV, FLAC, and M4A work in every modern browser.