mixing

Compression, part 2

more about compression

What Is Sidechain in a Compressor?

How a Compressor Normally Works

A compressor listens to the same signal it is affecting. When the volume of that signal goes above a certain threshold, the compressor reduces it. This helps control loud parts and smooth out the dynamics.

Splitting the Signal

Inside the compressor, the signal is split in two:

graph LR Input[Input] --> Split Split -->|Audio Path| GR{Gain Reduction} Split --> TD[Threshold Detection] TD -->|controls the| GR GR --> GMU{Gain Make-Up} GMU --> Output[Output]

This flowchart shows how an external signal is processed through an EQ and used to control the compression applied to a different signal — a common technique in music production and sound design.

One part goes through the compressor and gets processed.

The other part is only used to control when and how much the compression kicks in. This second part is called the sidechain.

Why This Is Useful

Because the sidechain is separate, we can do cool things like:

EQ the sidechain signal: For example, make the compressor react more to bass or less to harsh high frequencies.

graph LR Input[Input] --> Split Split -->|Audio Path| GR{Gain Reduction} Split --> EQ[EQ in Sidechain] EQ --> TD[Threshold Detection] TD -->|controls the| GR GR --> GMU{Gain Make-Up} GMU --> Output[Output]

Use an external signal as the sidechain: This means one sound can control the compression of another sound.

graph LR subgraph Inputs Input[Audio to Compress] External[External Sidechain Input] end External --> EQ[EQ in Sidechain] EQ --> TD[Threshold Detection] TD -->|controls the| GR{Gain Reduction} Input --> |Audiopath|GR GR --> GMU{Gain Make-Up} GMU --> Output[Output]

Examples of Sidechain Compression in Action

Kick Ducking a Bassline (Classic EDM Trick)

In electronic music, the kick drum is often used as a sidechain input to compress the bassline. Each time the kick hits, the bass volume ducks (gets lower) so the kick stands out. This creates that “pumping” sound you hear in a lot of dance music.

Taming Harsh Vocals with EQ Sidechain

You can insert an EQ in the sidechain path to make the compressor react more strongly when certain harsh frequencies (like around 3–5 kHz) get too loud. This makes the vocal smoother without affecting the whole signal.

Ducking Background Music for Voiceover

In podcasts or videos, you might want the background music to lower in volume automatically when someone starts talking. You can send the voice track into the sidechain input of the music’s compressor. When the voice comes in, the music ducks under it.

Knee

How It Works

The knee controls how gradually or abruptly compression starts once the signal crosses the threshold. In other word, once the threshold is exceeded, the ratio increases gradually from 1:1 to the set amount.

  • Hard Knee: Compression begins immediately and aggressively at the threshold.
  • Soft Knee: Compression begins more gradually, with the ratio increasing smoothly around the threshold rather than at a fixed point.

Visually, on a graph of input vs. output level, the “knee” refers to the bend in the curve where compression starts — which is where the term comes from.

Where It Comes From

The name “knee” originates from analog compressor circuits. When plotting input/output level graphs, the point where compression begins looks like a bent “knee” — either sharp or smooth, depending on how suddenly the compression is applied.

Why It’s Important
  • Hard knee is good for fast, aggressive compression — useful on drums or transient-rich material.
  • Soft knee is useful for transparent compression, such as vocals or mastering, where you want smooth, inaudible control.

In short:

  • Use hard knee for precision and punch.
  • Use soft knee for smoothness and subtlety.

Wet/Dry

The Dry/Wet knob allows you to blend the original (dry) signal with the compressed (wet) signal. This is useful for parallel compression, where you retain transients and natural dynamics while adding the body and tone of the compressed signal.

Examples:

  • Dry = 100%, Wet = 0% → No compression applied.
  • Dry = 0%, Wet = 100% → Fully compressed sound.
  • Dry = 50%, Wet = 50% → Balanced mix of clean and compressed signals.
Where It Comes From

Before plugins, engineers used two separate tracks:

  • One clean (dry)
  • One heavily compressed (wet)

They would then manually blend the two on a mixing desk. The Dry/Wet knob brings this parallel processing approach into a single tool.

Why It’s Important
  • Preserves transients while still adding compression.
  • Perfect for drums, vocals, or any sound needing more thickness without losing impact.
  • Allows extreme compression to be used musically, without sounding squashed.
  • Use Dry/Wet for a blend of control and natural tone.
  • Ideal for parallel compression setups without routing complexity.