Compression Basics

Compression is one of the most essential—and most misunderstood—tools in music production. Used correctly, it can tighten your mix, add punch to drums, control vocal dynamics, and shape the character of any sound. Used incorrectly, it can squeeze the life out of your music, creating a flat, lifeless result. This comprehensive guide demystifies compression, walking you through every parameter, practical applications, and real-world scenarios that professional producers encounter daily.

Overview

Compression is a dynamic process that reduces the volume of audio when it exceeds a certain threshold level. Think of it as an automatic volume controller that turns sounds down when they get too loud, then returns them to normal when they drop back below that threshold. The amount of reduction, how quickly it responds, and how it releases all affect the character of your mix. Understanding these parameters is fundamental to professional-sounding music production.

Key Points

Compression controls dynamic range by reducing volume when audio exceeds a threshold

Key parameters (threshold, ratio, attack, release) determine how a compressor sounds and responds

Different compressor types (FET, optical, VCA, tube) impart different characters to audio

Proper compression tightens mixes, adds punch, controls problematic peaks, and shapes tone

Over-compression is a common mistake that results in dull, lifeless audio

Genre and source material dramatically affect optimal compression settings

Subtle compression is often more effective than dramatic settings

Detailed Guide

What Compression Does: The Fundamental Concept

Before diving into parameters, understand what compression actually accomplishes. Imagine a vocal track where some words are loud and others are quiet. Without compression, you'd need to manually ride the fader, turning up quiet parts and turning down loud parts constantly. Compression automates this process. When audio exceeds your set threshold (say, -10 dB), the compressor reduces its volume. If you set a 4:1 ratio, for every 4 dB that audio exceeds the threshold, only 1 dB comes out. So audio 4 dB over the threshold gets reduced to 1 dB over. Audio 8 dB over gets reduced to 2 dB over, and so on. This maintains consistency without the manual labor of riding faders. The power of compression comes from its ability to make dynamic performances more consistent while maintaining musicality. Professional vocals, drums, and bass all use compression extensively.

The Five Essential Compression Parameters

Every compressor, from the simplest to the most complex, is built around five core parameters: #### 1. Threshold The threshold determines what level audio must reach before compression begins. Measured in dB, typical values range from -30 dB to 0 dB. How it works: Audio below the threshold is unaffected. Audio that exceeds the threshold gets compressed according to your ratio setting. Setting it: In your DAW, play your audio at normal volume. Adjust the threshold until you hear compression engaging intermittently—not constantly, but regularly. A good starting point is setting it just below the average level of your audio. If your vocal is averaging -6 dB, set threshold to -8 or -10 dB. Practical impact: Lower threshold = more audio gets compressed = tighter control but potentially more obvious compression effect. Higher threshold = only the peaks get compressed = more natural sound but less control. #### 2. Ratio The ratio determines how much the compressor reduces audio when it exceeds the threshold. Common ratios are 2:1, 4:1, 8:1, and 16:1 (or higher). Understanding ratios:

2:1 ratio: Soft compression. For every 2 dB above threshold, 1 dB comes out. Transparent, natural sound.

4:1 ratio: Moderate compression. For every 4 dB above threshold, 1 dB comes out. Audible but not obvious.

8:1 ratio: Strong compression. For every 8 dB above threshold, 1 dB comes out. Clearly audible character.

∞:1 ratio (limiting): Any audio above the threshold is held at exactly the threshold. Used for protecting against peaks. Hard brick wall effect.

Practical example: A vocal sits at -12 dB on average with peaks hitting -3 dB (9 dB above average). Setting threshold to -6 dB means peaks are 3 dB above threshold.

At 2:1 ratio: 3 dB over becomes 1.5 dB over = gentle smoothing

At 4:1 ratio: 3 dB over becomes 0.75 dB over = noticeable tightening

At 8:1 ratio: 3 dB over becomes 0.375 dB over = strong glue effect

Setting it: Start at 4:1 for most applications. Increase ratio if you need more control; decrease it if compression sounds obvious and unmusical. #### 3. Attack Time Attack determines how quickly the compressor reacts when audio exceeds the threshold. Measured in milliseconds, typical values range from 0.1 ms to 1000+ ms. Fast attack (0.1-5 ms):

Compressor engages almost immediately

Catches fast transients (the initial spike of a drum hit, vocal sibilance)

Reduces punchiness and impacts

Use on: vocals (to tame sibilance), snare drums (to add control without losing punch)

Medium attack (5-30 ms):

Lets initial transient through, then compresses the sustain

Maintains punch while controlling sustained levels

Sweet spot for most applications

Use on: kick drums (to keep punch), bass (to control sustain), most instruments

Slow attack (30-100+ ms):

The full transient passes through uncompressed; only the sustain gets compressed

Maintains natural dynamics and impact

Use on: drums (to preserve attack), transient-heavy instruments, when you want compression to be invisible

Quick example: Compress a snare drum with a 4:1 ratio and threshold at -8 dB.

Fast attack (1 ms): The initial "crack" is compressed. Snare sounds controlled, less punchy.

Medium attack (10 ms): The crack comes through; the tail is compressed. Snare sounds punchy with controlled sustain.

Slow attack (50 ms): The crack is fully uncompressed. Snare sounds natural, with subtle body control.

Setting it: For most applications, start at 10-20 ms. Adjust based on how much transient you want to preserve. #### 4. Release Time Release determines how quickly the compressor stops working once audio falls below the threshold. Measured in milliseconds, typical values range from 10 ms to several seconds. Fast release (10-100 ms):

Compressor stops almost immediately when audio drops below threshold

Creates pumping effect (audible compression moving with the rhythm)

Use on: drums (creates rhythmic tightness), bass (creates groove-oriented compression)

Medium release (100-300 ms):

Balanced approach between natural and controlled

Maintains smoothness without obvious pumping

Sweet spot for most applications

Use on: vocals, guitars, pianos, most instruments

Slow release (300+ ms):

Compressor holds compression for extended time even after audio drops below threshold

Creates smooth, glued sound where compression sustains

Use on: vocals (for smooth, consistent level), bass (for transparent control)

Practical example: Compress a vocal track with threshold -8 dB and 4:1 ratio.

Fast release (50 ms): As soon as the vocal phrase ends, compression stops. Between words, the track gets noticeably louder. Pumping is audible.

Medium release (200 ms): Compression gradually releases over 200 ms. Smooth but natural feel.

Slow release (500 ms): Compression holds even after the phrase ends. Level stays very consistent. Glued sound.

Setting it: Match your release to the tempo and feel of your music. A general starting point is 3-4 times your attack time. Adjust by ear, listening for whether the effect is too obvious or not enough. #### 5. Make-Up Gain (Output Level) Since compression reduces volume, make-up gain restores it. Without make-up gain, heavily compressed tracks sound quieter than uncompressed ones, making comparison difficult. How it works: The compressor measures how much gain reduction is occurring on average and the make-up gain adds back that amount (or whatever you manually set) to restore the level. Setting it: Use your meters. A properly set make-up gain means the output level of compressed audio matches the input level. You can then switch between compressed and uncompressed by toggling the compressor on/off, and the level stays consistent.

Practical Applications: Using Compression Across Your Mix

#### Vocal Compression Vocals benefit enormously from compression because singing involves dynamic range—some words are loud, others soft. Compression makes this consistent. Typical vocal compression settings:

Threshold: -10 to -20 dB (adjust so compression engages on most phrases)

Ratio: 4:1 to 8:1 (moderate to strong control)

Attack: 10-20 ms (lets initial consonants through, compresses sustain)

Release: 100-200 ms (smooth, natural feeling)

Make-up gain: Whatever restores original level

Workflow:

Play your vocal and identify the loudest word/phrase

Set threshold so compression engages noticeably (2-4 dB of gain reduction) on this peak

Set ratio to 4:1; increase if you want tighter control

Adjust attack/release by ear—it should feel natural, not obviously pumping

Set make-up gain to match the input level

A/B between compressed and uncompressed; should sound similar in volume but vocals feel more glued

Advanced vocal compression: Use two compressors in series. The first (2:1 ratio, 20 ms attack) handles overall dynamics gently. The second (8:1 ratio, 5 ms attack) catches sibilance and peaks more aggressively. This approach is common in professional vocal chains. #### Drum Compression Drums are percussive, so compression handling depends on whether you want to preserve or shape their transients. Kick Drum:

Threshold: -8 to -12 dB

Ratio: 4:1 to 8:1

Attack: 20-30 ms (preserves the initial impact/beater sound)

Release: 50-150 ms (quick enough to let the kick breathe between hits)

This creates a tight, controlled kick with preserved punch

Snare Drum:

Threshold: -8 to -15 dB

Ratio: 2:1 to 4:1 (use lower ratio for light glue effect)

Attack: 5-10 ms (lets crack through, compresses ring)

Release: 100-200 ms (controls the sustain without killing it)

The snare should sound punchy but not wild

Drum Bus Compression (compressing all drums together):

Threshold: -10 to -15 dB (so drums are always gently engaged)

Ratio: 2:1 (transparent, glue effect)

Attack: 3-5 ms (catches all transients together)

Release: 30-50 ms (fast, creates cohesion)

This glues drums together, making them feel like a unit rather than individual elements

#### Bass Compression Bass guitar and synth bass both benefit from compression to control dynamics and maintain consistent levels, especially important in the low-end of your mix. Bass Guitar:

Threshold: -10 to -15 dB

Ratio: 4:1 to 6:1

Attack: 20-30 ms (lets attack through, compresses sustain)

Release: 100-150 ms

Consistent level prevents certain notes from disappearing and others from jumping out

Synth Bass:

Threshold: -12 to -18 dB (adjust based on synth output level)

Ratio: 3:1 to 5:1

Attack: 10-20 ms

Release: 50-100 ms (faster release than bass guitar, creates groove feel)

Compression at the synth level shapes the tone before it even hits the mixer

#### Parallel Compression (New York Style) One of the most professional techniques is parallel compression: compressing heavily on a separate track while blending it with the dry signal. This allows you to get thick, compressed tone while maintaining dynamics. Setup:

Create a new auxiliary/return track

Send your vocal (or other source) to this track with pre-fader send

Put a compressor on the auxiliary with aggressive settings: 4:1 ratio, fast attack (5 ms), medium release (100 ms), threshold set so you get 6-10 dB of gain reduction

Blend the heavily compressed auxiliary with the dry vocal using the send amount

Now you have both the original dynamics and the compressed character

At 50% blend, you get a smooth, thick vocal with maintained dynamics. At 100%, you get full compression. The dry signal is preserved, so you never lose musicality.

Understanding Compressor Types

Different compression hardware designs have influenced DSP compressor emulations. Each has a sonic character: FET (Field Effect Transistor):

Very fast attack, aggressive character

Used for: vocals (sibilance control), drums (punchy glue), bass (aggression)

Examples: Universal Audio 1176, Empirical Labs Distressor

Optical:

Smooth, musical compression that's less obvious

Used for: vocals (natural glue), strings (smooth), any source needing subtlety

Examples: Universal Audio LA-2A, Softube Tube-Tech CL 1B

VCA (Voltage Controlled Amplifier):

Clean, precise control; can be subtle or aggressive

Used for: mixing bus (glue), drums (control), general purpose

Examples: SSL 4000E, API 2500

Variable-Mu (Tube):

Soft-knee compression, smooth and transparent

Used for: vocals, overall mix bus (sweetness), any source for subtle enhancement

Examples: Universal Audio Manley Variable Mu, Warm Audio WA-76

PWM (Pulse Width Modulation):

Modern, clean approach with transparent control

Used for: precise control without character, digital sources, transparent mastering

For starting out, focus on the compressor type less and more on the parameters. Any compressor can work well if parameters are set correctly.

Compression Mistakes to Avoid

Over-compression: Using too low a threshold or too high a ratio, compressing too much of the signal. This creates a dull, lifeless sound where nothing has any dynamics. A good test: disable the compressor and toggle it on/off. Should sound similar in level but with slightly different character, not dramatically different. Wrong attack time: Using too fast an attack on drums removes their punch; using too slow an attack on vocals fails to control peaks. Always listen to how the transient is affected. Ignoring make-up gain: Comparing compressed and uncompressed at different levels is unfair. Always match levels with make-up gain for accurate comparison. Compressing everything: Not every track needs compression. Some elements sound better uncompressed. Compress to serve the song, not for the sake of it. Using the same settings everywhere: A compressor setting perfect for drums is wrong for vocals. Adjust parameters for each source. Missing the release time: A release that's too fast creates pumping; too slow removes all dynamics. This is as important as ratio and threshold.

Visual Reading: Gain Reduction Meters

Most compressors show gain reduction in dB. Learning to read this meter is crucial:

0 dB gain reduction = no compression

1-2 dB gain reduction = subtle compression, barely noticeable

3-5 dB gain reduction = moderate compression, clearly working but not obvious

6+ dB gain reduction = strong compression, obviously affecting the signal

When setting compression, watch the gain reduction meter rather than just listening. It helps you understand what's happening technically. Ideally:

Vocals should show 3-6 dB of gain reduction on peaks

Drums should show 2-4 dB on hits

Bass should show consistent 2-3 dB throughout

Stereo sources (drums bus) should show smooth, consistent reduction

Compression Chains: Using Multiple Compressors

Professional mixes often use multiple compressors in series, each doing a slightly different job: Common vocal chain:

First compressor (gentle): 2:1 ratio, 20 ms attack, 200 ms release (general glue)

Second compressor (aggressive): 8:1 ratio, 5 ms attack, 50 ms release (sibilance and peak control)

The first handles overall dynamics; the second catches problem peaks. Together, they create smooth, professional vocals without any single compressor being obvious. Drum bus chain:

First compressor (fast, glue): 2:1 ratio, 3 ms attack, 50 ms release

Second limiter (safety): ∞:1 ratio, 1 ms attack (stops anything from going over -3 dB)

The first glues drums together; the second acts as insurance, preventing surprises.

When NOT to Use Compression

While compression is essential, some sources sound better uncompressed:

Ambient pads: Often sound best completely uncompressed

Orchestral instruments: Can sound overly controlled with compression

Well-recorded spoken word: Natural delivery often beats compressed clarity

Electronic synthesizers: If already compressed during synthesis, additional compression may be redundant

Use compression because it improves the sound, not as a rule.

A/B Comparison Workflow

Always compare compressed and uncompressed:

Set your compressor with make-up gain matching input level

Record a 30-second section of your audio playing

Play it back and toggle the compressor on/off several times

At equal levels, which sounds better for this mix?

Does compression improve clarity? Reduce dynamics? Add sustain? Is that helpful?

Adjust parameters and repeat

This scientific approach prevents over-processing and ensures you're making conscious choices.

Makeup: Gain Staging for Clarity

Proper gain staging before compression ensures optimal results:

Set your input level so peaks reach approximately -12 to -6 dB (not too hot)

This leaves headroom for compression to work without clipping

Use input gain on your channel to control the signal level hitting the compressor

Then use make-up gain to restore the output level

This workflow ensures clean signal flow and optimal compressor response.

Recommendations for Effective Compression

Start subtle: Most compression should be invisible. If you notice it immediately, it's probably too much.

Listen on multiple systems: What sounds good on headphones might be wrong on monitors. Check your compressed mix on speakers and earbuds.

Learn one compressor deeply: Rather than always switching compressors, master one well. This helps you develop intuition.

Use automation instead sometimes: Occasionally, automation is better than compression for shaping dynamics.

Compress groups, not just individuals: Bus compression on drums, guitars, or instruments often sounds better than individual track compression.

Reference professional mixes: In your genre, compare your compressed signals to professional references. Are your compressors less or more obvious? Adjust accordingly.

Keep it simple: Three well-chosen parameters (threshold, ratio, attack/release) matter more than understanding every feature.

---