In this guide
Use compression to shape level differences over time and limiting when peaks need firmer control. Check the actual processor and delivery requirements: a sample-peak ceiling is not automatically a true-peak guarantee.
A compressor reduces level differences by turning down louder passages according to a threshold, ratio and timing controls. A limiter is a form of dynamics processing commonly used for firmer peak control. Compressors are often used to shape a voice or instrument; limiters are often placed later in a mix to manage peaks. The boundary is not absolute: individual processors have different detectors, timing, knees and output controls.
Neither operation is the same as making an MP3 file smaller. Saturalabs' audio file compressor encodes a compact MP3; it does not offer the dynamics controls described below.
Compressor vs limiter at a glance
| Question | Compressor | Limiter |
|---|---|---|
| Common use | Make a voice or instrument more consistent, or shape its character. | Control peaks on a track or final mix while preserving an intended output ceiling. |
| Amount of reduction | Ratio and threshold determine the compression curve; timing changes its response. | Usually much stronger near its limiting threshold, with behavior depending on the design. |
| Listen for | Lost transients, pumping, or breaths and noise becoming more apparent after makeup gain. | Flattened attacks, distortion, pumping or an overly dense mix if driven hard. |
| What it cannot do | Recover audio that was clipped before the processor. | Guarantee a clean mix, restore earlier clipping, or meet a true-peak specification without appropriate measurement. |
How threshold, ratio and ceiling differ
A compressor's threshold determines when its gain reduction starts; the ratio describes the compression curve above it. In an idealized hard-knee 2:1 example, a steady signal 6 dB above threshold comes out 3 dB above it before makeup gain. Attack and release affect the behavior of real, changing audio. The compressor guide walks through the controls.
A limiter generally uses a very high effective ratio near its threshold, but its response is not defined by ratio alone. Lookahead, knee, release and the output ceiling matter. In Audacity's current Limiter, the threshold controls where limiting begins and the makeup target controls the maximum output level; its controls do not map identically to every limiter. Earlier versions have different effects and labels.
Choose a workflow for your recording
- Listen and fix individual problems first. If one word is much louder than the rest, a clip-gain edit can address that moment without forcing the whole track through heavy processing. Preserve a copy of the original.
- Try compression for ongoing level differences. Use a modest ratio and adjust the threshold on a passage with both quiet and loud words. Compare at similar perceived loudness, checking consonants and room noise.
- Use limiting when the remaining peaks need control. Put a suitable limiter later in the chain, choose a ceiling appropriate to the destination and increase reduction gradually. Listen to the loudest passage and watch the reduction meter if available. If it sounds flattened or distorted, back off or fix the peaks earlier.
- Measure the delivered file. Export the intended format and recheck loudness and peaks. For a true-peak requirement, use a meter and limiter that explicitly support true peak. Lossy encoding may change decoded peaks, so check the encoded result as well.
For a short local comparison, the free loudness meter reports integrated loudness and decoded sample peak for mono or stereo MP3/WAV files up to five minutes and 50 MiB. It does not measure true peak or certify a complete episode. The podcast loudness guide explains the difference and how to check a delivery.
Can a limiter fix clipping?
It can reduce later peaks, but it cannot reconstruct peaks already lost in an overloaded recording. Lowering the level of that recording makes the distortion quieter, not undone. Compare a clean source or another take first; the audio repair guide explains how to test dedicated reconstruction on light clipping and why it may fail.
Do you need both?
No. A well-balanced recording may need neither. A compressor can be useful on an uneven voice without a limiter, and a limiter can catch occasional peaks without prior compression. Using both is reasonable when they solve separate audible problems, but extra processing is not an end in itself. Keep the version that sounds best at a matched level and meets the actual delivery specification.