Compress Audio Dynamic Range
Pick an audio file, set threshold, ratio, attack, release and make-up gain — it's compressed in your browser, nothing is uploaded.
Compress audio dynamic range in your browser
Pick an audio file and set the four classic compressor controls — threshold,
ratio, attack and release — plus optional make-up gain. The tool
runs ffmpeg's acompressor filter entirely in your browser to pull loud and
quiet passages closer together, so the audio sits at a steadier, more consistent
level. Nothing is uploaded, so it's safe for unreleased tracks and private
recordings.
This is dynamics compression (evening out the level), not file-size compression — to make a file smaller pick a lower bitrate with the audio-compress tool instead.
Worked example
A voice-over narration.wav where some lines jump out and others disappear:
upload it, set Threshold to -24, Ratio to 4, Attack to 10,
Release to 150 and Make-up gain to 6. The result
narration-compressed.mp3 holds a tight, even level — the quiet lines come
forward, the loud ones stop spiking, and the 6 dB make-up gain brings the whole
thing back up to a comfortable volume.
What each control does
- Threshold (dB, −60…0) — the level where compression kicks in. Lower (−30) catches more of the signal; higher (−10) only tames the loudest peaks.
- Ratio (1…20) — how hard everything above the threshold is squeezed.
2is gentle,4is a firm even level,10+ approaches limiting.1is no compression. - Attack (ms, 0.01…2000) — how fast it clamps down once the signal passes
the threshold. Fast (
5) tames sharp transients; slow (30+) lets punch through. - Release (ms, 0.01…9000) — how fast it lets go once the signal drops back.
Short (
60) for speech, longer (250+) for smoother music. - Make-up gain (dB, 0…24) — level added after compression to restore the
loudness the compressor pulled down. Try
3–6when heavier settings leave the result too quiet.
Starting points
- Voice / podcast — threshold
-24, ratio4, attack10, release150, make-up4–6. Even, forward, intelligible. - Music glue — threshold
-18, ratio2–3, attack30, release250. Gentle cohesion without squashing the dynamics. - Peak taming — threshold
-10, ratio8+, fast attack5, make-up0. Only the loudest moments get caught.
Limits and edge cases
- Input files up to 10 MiB; any format ffmpeg can decode works (mp3, wav, m4a, ogg, flac, and more).
- A ratio of 1 with 0 dB make-up gain does nothing and is rejected — raise the ratio above 1 to compress, or add make-up gain to use the tool as a clean level boost.
- Output is re-encoded because the compressor rewrites the samples (mp3/ogg at 192 kbps; wav/flac lossless; m4a AAC). Embedded album art is dropped.
- Compression is a creative, one-way change — too much (very high ratio + heavy make-up) can sound flat and pumpy, so keep the original if it matters.
FAQ
What's the difference between this and the "audio compress" tool?
They do different things. This audio compressor does dynamic-range compression — it evens out the loud and quiet parts so the level stays steady, using threshold/ratio/attack/release. The separate audio-compress tool does file-size compression — it re-encodes at a lower bitrate to make the file smaller. Reach for this one to control loudness, and audio-compress to shrink a file.
What threshold and ratio should I start with?
For speech and podcasts, threshold -24 dB with ratio 4:1 gives a firm, even
level; add 4–6 dB of make-up gain to bring it back up. For gentle "glue" on
music, use a higher threshold (-18) and a lower ratio (2–3). To only catch
the loudest peaks, raise the threshold toward -10 and the ratio to 8+.
Why does my compressed audio sound quieter?
Compression turns the loud parts down, so the overall level often drops — that's
what make-up gain is for. Add a few dB (start with 3–6) to restore the
loudness the compressor pulled down. With enough make-up gain the result can end
up louder and more even than the original.
What do attack and release actually change?
Attack is how quickly the compressor reacts once the signal crosses the
threshold: a fast attack (5 ms) clamps sharp transients like plosives or drum
hits, while a slower one (30+ ms) lets that initial punch through before
compressing. Release is how quickly it stops compressing afterwards: short
releases (60 ms) suit speech, longer ones (250+ ms) keep music smooth and
avoid audible pumping.
Is my audio uploaded anywhere?
No. The page downloads an ffmpeg WebAssembly build once and then processes your file locally in the browser tab — the audio never leaves your device.
Developer & Automation Access
Run it from the terminal
Same engine as this page, headless — via the gizza CLI:
gizza tool audio-compressor 'url=https://example.com/input' 'threshold=-20' 'ratio=4' 'attack=20' 'release=250' 'makeup=0' 'format=mp3'New to the CLI? Get gizza →
Open it by URL
Pre-fill and auto-run this tool with query parameters — the names match the API/CLI:
https://gizza.ai/tools/audio-compressor/?url=https://example.com/input&threshold=-20&ratio=4&attack=20&release=250&makeup=0&format=mp3Machine-readable descriptor: tool.json — title + parameters JSON Schema for agents.