Quick Settings Table
A compressor is easier to learn when each control has one job. Start with threshold and ratio, then adjust attack and release only after you can hear what the compressor is doing. Knee and makeup gain refine the result, while input and output gain keep the signal at a workable level before and after processing.
| Parameter | What it controls | Typical beginner range | When to adjust |
|---|
| Threshold | The level where compression begins | -24 to -12 dB for speech | Lower it when peaks are not being controlled; raise it when everything sounds squeezed |
| Ratio | How strongly signal above threshold is reduced | 2:1 to 4:1 | Increase for firmer voice control; decrease for a more natural sound |
| Attack | How fast compression starts after the signal crosses threshold | 5 to 30 ms | Faster catches sharp peaks; slower keeps more transient detail |
| Release | How fast compression stops after the signal falls back | 50 to 200 ms | Adjust when the sound pumps, breathes, or stays flattened |
| Knee | How gradually compression begins around the threshold | Soft or medium for beginners | Use soft knee for smoother voice; hard knee for more obvious control |
| Makeup Gain | Level added after compression | As needed, often 1 to 6 dB | Raise only enough to match loudness after gain reduction |
| Input Gain | Level entering the compressor | Enough to hit threshold on peaks | Use when the compressor is barely reacting or reacting constantly |
| Output Gain | Final level leaving the compressor | Match bypassed loudness | Use to prevent clipping or keep plugin chains balanced |
Threshold
Threshold answers the question: how loud must the signal get before compression starts? If a voice peaks around -12 dB and the threshold is set to -18 dB, the louder words cross the line and get reduced. If the threshold is set too high, the compressor may barely move. If it is too low, the whole recording can sound held down.
For podcast speech, lower the threshold until the gain reduction meter moves mainly on louder words, not during every syllable. Background music often needs a higher threshold or gentler compression because constant gain reduction can make the bed feel dull. The dedicated audio compressor threshold guide goes deeper into this starting point.
Ratio
Ratio decides how much the compressor pushes down the part above threshold. A 2:1 ratio is gentle: 6 dB over the threshold becomes about 3 dB over it. A 4:1 ratio is stronger and is common for voice control. Around 10:1 the result starts to feel limiter-like, especially when paired with a low threshold and fast timing.
Higher ratio is not automatically better. Too much ratio can make speech dense, small, or obviously processed. For a focused explanation with examples, use the audio compressor ratio guide.
Attack and Release
Attack and release decide how quickly the compressor reacts. Fast attack grabs peaks quickly, which can help harsh consonants or sudden loud words, but it can also remove life from drums, guitar, or expressive speech. Slower attack lets the first moment of a sound pass through before compression starts, preserving punch and clarity.
Release controls recovery. Too fast can cause audible pumping because the level rises and falls between words or beats. Too slow can leave the compressor clamped down, making the next phrase feel flat. The attack and release guide explains how to tune timing by ear.
Knee and Makeup Gain
Knee controls how abruptly compression begins near the threshold. A hard knee applies the ratio more suddenly and can sound precise or aggressive. A soft knee eases into compression around the threshold, which is usually more forgiving on speech, podcast voices, and acoustic sources.
Makeup gain compensates for the level lost when peaks are reduced. It does not make the file smaller, and it should not be used just to make everything loud. Add enough gain to compare the processed signal fairly with the unprocessed one. If makeup gain pushes the output into clipping, reduce output gain or use a limiter later in the chain.
Starter Settings for Common Uses
The numbers below are starting points, not rules. Different microphones, rooms, voices, and recordings need different thresholds because the signal level coming into the compressor changes. Set threshold by watching gain reduction, then use these ranges to choose the general behavior.
| Use case | Starter settings | Listen for |
|---|
| Podcast voice | Ratio 3:1, attack 10-20 ms, release 80-150 ms, soft knee | Even phrases without obvious pumping |
| Streaming microphone | Ratio 3:1 to 4:1, attack 5-15 ms, release 80-120 ms | Sudden laughs controlled without sounding pinned |
| Voice-over | Ratio 2:1 to 3:1, attack 10-25 ms, release 100-200 ms | Natural delivery and stable narration |
| Acoustic guitar | Ratio 2:1, attack 20-40 ms, release 100-250 ms | Preserved pick detail and less spiky strumming |
| Background music | Ratio 1.5:1 to 2:1, slower attack, medium release | Gentle leveling without obvious movement |
| General speech recording | Ratio 2:1 to 4:1, threshold for 3-6 dB reduction on peaks | Clarity, not loudness for its own sake |
Settings That Affect File Size Instead
If your goal is smaller MB, compressor plugin controls are the wrong starting point. File size is mainly controlled by bitrate, format, duration, sample rate, and channels. A 30 minute WAV can remain huge even after perfect voice compression. A 30 minute MP3 can become smaller when exported at a lower bitrate or as mono speech.
Use the Audio Bitrate Calculator to estimate a bitrate from duration and target size, or the Audio File Size Calculator to compare settings before export. Then use the browser Audio Compressor when you need the actual smaller file.
- Lower bitrate reduces data per second.
- More efficient formats can reduce size at similar perceived quality.
- Mono can save space for centered speech.
- Lower sample rate can help speech but should be tested carefully for music.