Upward Compression Explained: Bring Up Detail Without Squashing

What is upward compression?

Upward compression makes quiet signals louder once they drop below a threshold, while leaving louder material alone. Ableton's manual puts it simply: regular compression is really downward compression, because it pushes loud signals down, and you can also shrink dynamic range by raising signals below a threshold, which is called upward compression^[1].

Both approaches reduce the distance between loud and quiet parts. They just squeeze from opposite ends. The table shows where upward compression sits among the four basic dynamics processes.

Ableton describes upward compression and upward expansion as much less common than their downward versions^[1]. Waves points out that a transient shaper boosting attack behaves like an upward expander^[2], which is why our transient shaping guide is the natural companion to this one.

Upward vs downward compression: what is the real difference?

The difference is which part of the signal gets processed. Downward compression acts on peaks and transients, so it changes the attack of every hit it touches. Upward compression acts on the quiet material between hits, so transients pass through untouched. Waves lists exactly that as its main advantage: it reduces dynamic range without processing transients^[2].

A downward compressor with makeup gain can look similar on a meter, because quiet parts also end up louder relative to the peaks. Hugh Robjohns in Sound On Sound calls that uplift compression and separates it from true upwards compression: with uplift, the loud elements are still the ones being squashed and the quiet ones are merely raised afterwards^[3]. You hear that difference most on punchy material like drums and plucked instruments.

If the downward side is still new to you, start with audio compression explained.

How do plugins implement upward compression?

Plugins implement upward compression in one of three ways: a ratio below 1:1 on a compressor, a dedicated upward module with a maximum-gain limit, or a separate lower threshold on a multiband dynamics processor. The controls look different, but all of them raise gain when the signal falls below a threshold.

• Ratio below 1:1 iZotope notes that Neutron's compressor switches to upward mode with ratios below 1:1, and recommends very low ratios between 1:1 and 0.67:1 to avoid raising the noise floor 20-30 dB in quiet sections^[4].
• Ratio plus a range limit Neutron 5's Density module presents ratios from 1.1:1 to 30:1 and adds a Range control that caps the maximum added gain^[4]. Waves' F6 works similarly: lower the threshold and set Range above zero^[2].
• Two-knob processors Waves MV2 has a High Level control for downward compression and a Low Level control for upward compression, so you can squeeze from both sides at once^[2].
• Below thresholds in multiband dynamics In Ableton's Multiband Dynamics, each band has Above and Below thresholds; raising the level under the Below threshold applies upward compression^[1].
• OTT Xfer's OTT is a three-band compressor that applies both upward and downward compression and started as a preset for Ableton's Multiband Dynamics^[5]. Xfer lists it as freeware for Mac and Windows^[6].

Attack and release work backwards

On an upward compressor, attack sets how quickly gain rises once the signal drops below the threshold, and release sets how quickly it returns to normal when the signal goes back above it^[4]. A fast attack makes tails swell up almost instantly after a hit. A slower attack lets the natural decay start before the lift arrives, which usually sounds more natural.

Is parallel compression the same as upward compression?

Parallel compression behaves like upward compression, but it is not identical. Blending a heavily compressed copy with the dry signal raises quiet detail while the dry path keeps the peaks. Bob Katz, who coined the term parallel compression, described it as an implementation of upward compression^[7], and iZotope suggests parallel compression as a way to approximate upward compression when you don't have a dedicated upward compressor^[4].

In practice, parallel compression also adds the color and density of the compressor you choose, which is often why people use it. For routing and blend levels, see our parallel compression tutorial.

Where to use upward compression: vocals, ambience, drums and masters

Use upward compression wherever the important information is quiet and the peaks are already fine. Waves highlights room sound, reverb heard in the gaps, ghost notes, breathy vocal qualities and finger noise on guitar as typical targets^[2].

• Vocals Lifts word endings, soft phrases and ad-lib tails without touching the loud syllables. Cap the lift at 3-6 dB and put a de-esser or gate before it, because breaths and S sounds rise too.
• Room mics and ambience iZotope describes using an upward compressor to bring out the space between transients on drum room mics, and even turning an unused tom mic into convincing ambience^[4].
• Drums and percussion Raises ghost notes and snare buzz between backbeats. Set the threshold just under the main hits so only the in-between detail is lifted.
• Reverb and delay returns Makes tails bloom after the dry sound stops. Pair it with a sidechain duck if the lifted tail starts masking the vocal.
• Synths and sound design OTT-style processing brings out texture and noise in leads, pads and risers; blend it with the dry signal to keep some dynamics.
• Masters A gentle upward stage of 1-2 dB can lift quiet passages and fades without touching peaks. Keep it broadband or use wide bands, and compare at matched loudness.

How to set up upward compression step by step

1. Clean the quiet parts first
   Remove noise, hum and unwanted room sounds, or gate them. Anything left under the threshold will be turned up.
2. Find the detail level
   Watch the meter and note where the detail you want sits, for example tails around 15-25 dB below the peaks.
3. Set the threshold
   Place it just above that detail level and below the main hits, so peaks never trigger the lift.
4. Choose a gentle amount
   Start near 0.8:1 on ratio-based compressors, or a moderate ratio with a 3-6 dB Range on plugins that cap the gain.
5. Shape attack and release
   Begin with a medium attack around 10-30 ms and a release of 50-150 ms, then adjust until tails rise smoothly without swelling.
6. Check the quietest section
   Solo the track during a break or intro. If hiss, amp noise or breaths jump out, lower the range or raise the ratio toward 1:1.
7. Level-match the comparison
   Bypass and match loudness before deciding. Upward compression makes things sound bigger mostly because they are louder.

What can go wrong with upward compression?

The main risk is raising what you never wanted to hear. iZotope warns that aggressive ratios can bring the noise floor up 20-30 dB in quiet sections^[4], and Waves notes that most upward tools include a ceiling or range because a very quiet signal should not be pushed all the way up^[2].

The second risk is flattening contrast. A song whose verses and breakdowns are lifted to chorus level loses impact, even if every element sounds more detailed. Use upward compression on specific tracks and buses, and keep multiband versions simple. For choosing crossover points, see our multiband compression guide.