What Is Dynamic Range Compression?

Dynamic range compression (DRC) is an audio processing technique that reduces the volume gap between the loudest and quietest parts of a signal. In any recording, the dynamic range is the difference between the peak level and the noise floor. When that range is too wide, quiet sections may be buried in ambient noise while loud sections risk distortion. By applying compression, you raise the quiet passages and lower the loud ones, producing a more consistent level that makes background noise less intrusive without destroying the natural character of the audio.

Compression is not a noise removal tool in the traditional sense—it does not subtract noise like a gate or spectral editor. Instead, it changes the relationship between the signal and the noise floor. When you bring up quiet sections, you also bring up any noise that was previously masked. That sounds counterproductive, but when used correctly with threshold and ratio settings, compression can actually make noise less noticeable by preventing the loud parts from drowning out the quiet details and by keeping the overall level steady enough that the listener’s ear does not constantly adjust.

Key Parameters of a Compressor

To use DRC effectively for noise reduction, you must understand its core controls. Every compressor—whether hardware or software—offers these basic parameters:

Threshold

The threshold is the level above which compression begins. Any signal below the threshold passes through unchanged. To reduce noise, set the threshold just above the noise floor but below the average level of your desired audio. This way, the compressor only activates when the signal is strong, leaving quiet noise untouched, which prevents the noise from being amplified in the gaps between words or notes.

Ratio

The ratio determines how much compression is applied once the signal exceeds the threshold. A ratio of 2:1 means that for every 2 dB of input above the threshold, only 1 dB passes through. Higher ratios (e.g., 8:1, 10:1) are more aggressive and can clamp down on loud peaks, but they also risk making the audio sound unnatural. For noise reduction, a moderate ratio (3:1 to 6:1) is often a good starting point.

Attack and Release

Attack time controls how quickly the compressor responds once the signal crosses the threshold. Fast attacks (1–10 ms) catch transients like drum hits, but can also squash the initial punch of a sound. For noise reduction, a fast enough attack to catch the noise burst (like a hiss that appears with a loud sibilance) is helpful, but not so fast that it distorts the waveform. Release time determines how quickly the compressor stops compressing after the signal falls below the threshold. A release that is too short can cause pumping and breathing; one that is too long can leave the gain reduced for too long, making quieter sections sound unnaturally suppressed. Typical release settings range from 50 ms to 300 ms for speech or moderate music.

Knee

The knee controls how gradually compression is applied around the threshold. A hard knee (0 dB) means compression starts abruptly at the threshold. A soft knee (6 dB or more) smooths the transition, which can sound more natural and less obvious when compressing subtle noise. For gentle noise reduction, a soft knee is often preferable.

Make‑Up Gain

After compression reduces the overall level, you apply make‑up gain to bring the signal back up. This is where the noise can become more audible if you have brought the average level too high. Use make‑up gain sparingly and always A/B compare with the uncompressed signal.

How DRC Helps Reduce Noise

When you compress audio, the quiet parts are raised relative to the loud parts. If you have a constant background hum or hiss, it will be amplified during the silent gaps that compression creates. That sounds like a problem, but here is the key: if you set the threshold appropriately, the compressor will only activate when the main signal is present. During silent moments, the compressor releases and the gain returns to unity or even less. This means the noise is actually lowered during those gaps. Combined with a noise gate set just above the noise floor, you can effectively remove noise in the pauses while using compression to maintain a consistent level during speech or music.

Another mechanism is peak limiting. By reducing the loudest transients, you can increase the overall average level without clipping. This makes the audio sound louder and more present, so the listener perceives less noise even if the absolute noise floor is unchanged.

Practical Workflow for Noise Reduction with Compression

To put theory into practice, follow this step‑by‑step workflow in any DAW (Digital Audio Workstation):

  1. Identify the noise floor. Use a spectral analyzer or listen to a silent section of the track. Note the RMS level of the noise (e.g., −60 dBFS).
  2. Insert a compressor on the track or bus. Start with a ratio of 4:1, a fast attack (10 ms), a medium release (100 ms), and a soft knee.
  3. Set the threshold. Adjust it so that the compressor only engages on peaks above about −20 dBFS (assuming your main signal peaks around −6 dBFS). This leaves the noise untouched during pauses.
  4. Adjust the ratio. Increase the ratio if the loud parts are still too dynamic; decrease if the compression sounds too heavy. For noise reduction, a ratio between 3:1 and 6:1 usually works well.
  5. Fine‑tune attack and release. Listen to a section with consonants or transients. If they sound dull, lengthen the attack. If you hear pumping, adjust the release (try 50–150 ms for speech, 200–400 ms for musical phrases).
  6. Apply make‑up gain. Bring the output level up until the loudest moments match the original peak level. Then reduce by 1–2 dB to ensure you are not amplifying noise excessively.
  7. Add a noise gate after the compressor. Set the gate threshold a few decibels above the noise floor (e.g., −55 dBFS) with a fast attack and a release that matches the compressor’s release. This silences the gaps completely.
  8. A/B compare. Toggle the compressor and gate on and off to hear the difference. Adjust as needed.

Combining Compression with Other Tools

Compression alone is rarely enough for serious noise problems. Pair it with these tools for best results:

Noise Gate

A noise gate mutes the signal when it falls below a set threshold. Placing a gate after a compressor works well because the compressor has already narrowed the dynamic range, so the gate can cleanly cut the noise without interfering with the main audio. For dialogue, set the gate threshold about 3–6 dB above the noise floor with a fast attack (1–5 ms) and a release of 50–100 ms.

Equalizer (EQ)

Use a parametric EQ to cut frequencies where the noise is prominent—for example, a high‑pass filter at 80 Hz to remove rumble, or a notch at 60 Hz to eliminate hum. Compression can then act more cleanly on the remaining content. Always EQ before compression to avoid amplifying unwanted frequencies.

Multiband Compression

Multiband compressors split the frequency spectrum into several bands and compress each independently. This allows you to target noise in a specific range (like hiss in the 5–10 kHz region) without affecting low frequencies. For instance, you can compress only the high band with a 4:1 ratio, leaving the mid and low bands untouched. This is more precise than wideband compression for noise reduction.

Expanders

An expander is the opposite of a compressor: it makes quiet sounds even quieter. A downward expander can be used to reduce noise in the gaps without the hard cutoff of a gate. It works well on material where you want to preserve some ambient sound but still reduce noise.

Common Mistakes When Using DRC for Noise Reduction

  • Over‑compression. Applying too much ratio or too low a threshold leads to a lifeless, “squashed” sound. Always aim for the minimum compression that achieves your goal.
  • Ignoring the release time. A release that is too fast causes the compressor to “pump” on every syllable; too slow and the noise becomes audible again during pauses.
  • Boosting make‑up gain too high. This raises the noise floor along with the signal. Keep make‑up gain conservative—you want the perceived loudness to be consistent, not louder.
  • Not using a gate. Relying solely on compression to remove noise often fails because compression amplifies noise in quiet moments. A gate or expander is essential for clean results.
  • Compressing before EQ. If you have a low‑frequency rumble, compression will amplify it. Always remove problematic frequencies first with EQ, then compress.

Real‑World Examples

Podcast Dialogue with Background Hum

You have a podcast recorded in a room with a 50 Hz hum and some computer fan noise. Start with a high‑pass filter at 80 Hz (12 dB/octave) to reduce the hum. Then insert a compressor with a ratio of 3:1, a threshold around −30 dBFS, a fast attack (5 ms), and a release of 100 ms. Add a noise gate set to −50 dBFS with a release of 80 ms. The result: the hum is reduced, the fan noise is inaudible during pauses, and the voice maintains a consistent level.

Live Music Recording with Audience Noise

For a live acoustic guitar performance with audience rustling, use a multiband compressor. In the 2–5 kHz band (where rustling is prominent), apply a 4:1 ratio with a low threshold. On the full signal, use a light compressor (2:1, slow release) to even out the dynamic range. Finally, a gate on the guitar track with a high enough threshold to close during quiet audience noise.

Advanced Tips for Professional Results

  • Use sidechain filtering. Many compressors allow you to filter the signal that triggers compression. Apply a high‑pass filter to the sidechain so that only high frequencies (where hiss lives) trigger compression. This reduces noise without affecting bass content.
  • Automate threshold. In DAWs, you can automate the compressor’s threshold to change over time. For example, raise the threshold during loud sections to prevent over‑compression, and lower it during quiet sections to keep the signal consistent.
  • Compress in stages. Apply gentle compression at the source, then a bit more on the bus, and a final limiter on the master. This gradual approach sounds more natural than one heavy‑handed compressor.
  • Use parallel compression. Blend a heavily compressed version of the track with the dry signal. This preserves transients and dynamics while adding density, which can mask noise without bringing it up.

Choosing the Right Compressor

Different compressor types have distinct characteristics that affect noise reduction:

  • VCA compressors (like the dbx 160 or SSL bus comp) offer precise, clean compression with fast attack and release. Ideal for dialogue and modern music where transparency is important.
  • Optical compressors (like the LA‑2A) have a smooth, musical response that can soften harshness. They are slower but great for vocals and acoustic instruments.
  • FET compressors (like the 1176) add color and can catch transients aggressively. Useful for drums and rock vocals, but may emphasise noise if not set carefully.
  • Digital/plugin compressors offer the most control, often with multiband and sidechain options. For noise reduction, start with a digital compressor like FabFilter Pro‑C 2 or Waves C6.

External Resources

For further reading on dynamic range compression and noise reduction techniques, see these authoritative sources:

Dynamic range compression is not a magic bullet for noise, but when combined with thoughtful threshold settings, complementary tools like gates and EQ, and careful listening, it can dramatically improve the clarity and professionalism of your audio tracks. Practice on real recordings, trust your ears, and always aim for the simplest solution that cleans the noise while preserving the natural dynamics of your source material.