Introduction: Why Dynamic Range Matters in Dialogue

Clean, intelligible dialogue is the bedrock of any audio production—whether it’s a feature film, a television drama, a podcast interview, or a corporate voiceover. Listeners can forgive a slightly muddied music mix or a noisy background ambience, but if the dialogue is hard to understand, the narrative falls apart. Maintaining a natural dynamic range—the difference between the quietest and loudest parts of a recording—is key to preserving the emotional nuance and clarity of speech. However, unwanted background noise (air conditioning hum, traffic rumble, microphone self-noise, rustling clothing) often creeps into quiet passages, forcing engineers to make tough decisions. Noise gates and expanders are two essential tools that allow you to clean up these distractions while keeping the dialogue’s dynamic contour intact. When used correctly, they can transform a noisy, fatiguing track into a polished, professional one.

The Fundamentals of Noise Gates and Expanders

What Is a Noise Gate?

A noise gate is essentially an automatic volume control that mutes or attenuates any audio signal that falls below a defined threshold. When the signal is louder than the threshold, the gate is open, allowing sound to pass through unchanged. When the signal drops below the threshold, the gate closes, reducing the output to silence or to a level set by the gate’s range parameter. This makes gates very effective for eliminating steady background noise during pauses in speech. However, a poorly tuned gate can sound unnatural—cutting off the tail of a word, for example, or chopping the room ambience too abruptly.

What Is an Expander?

An expander works on a similar principle but is gentler and more musical. Instead of fully closing the gate, an expander reduces the gain of signals below the threshold by a specified ratio. For instance, a 2:1 expansion ratio means that for every 1 dB the input drops below the threshold, the output drops by 2 dB. This results in a smoother, more gradual reduction of low-level sounds. Expanders are ideal for situations where you want to suppress background noise without completely killing the room tone, preserving a sense of space and realism.

Key Differences and When to Use Each

The main difference lies in the aggressiveness of the processing. A gate is a binary on/off (or near-binary) tool; an expander is a variable gain reducer. Use a gate when you have a loud, steady noise source (like a computer fan) that you want to completely silence during pauses. Use an expander when the noise is softer or when you want to retain some of the room’s natural ambience while still reducing the noise floor. In practice, many engineers combine both: a subtle expander to reduce low-level noise, followed by a gate to catch any remaining leakage.

Setting Up a Noise Gate for Dialogue

Identifying the Noise Floor

Before touching any controls, listen critically to your dialogue track in a quiet monitoring environment. Identify the loudest source of consistent background noise—this is your noise floor. Measure its level on a VU or Peak meter (typically around -50 to -40 dBFS in a clean recording, but can be higher in field recordings). You’ll want the gate threshold to sit just above this noise floor, typically 2–6 dB higher, so that the gate stays open during speech but closes during gaps.

Threshold Adjustment

Set the threshold so that the gate opens reliably whenever the actor speaks, even on the softest syllables. A common mistake is setting the threshold too high, causing the gate to close mid-word, which sounds like glitching or choking. Conversely, a threshold too low will fail to gate the noise. Start with the threshold at about -6 dB below the average speaking level, then fine-tune by listening to the quietest lines in the performance. For dialogue, the threshold range is often between -30 dBFS and -20 dBFS depending on the recording’s level.

Attack, Hold, and Release Parameters

These timing controls are critical for natural-sounding gating. Attack determines how quickly the gate opens after the signal exceeds the threshold. For dialogue, a fast attack (0.1–1 ms) is usually desirable to avoid missing any transient consonants (t, k, p). Hold keeps the gate open for a short time after the signal drops below threshold, preventing the gate from chattering during short pauses. A hold of 50–200 ms works well for speech. Release sets how fast the gate closes once the hold time expires. Too fast and you’ll hear the ambience abruptly cut off; too slow and background noise will creep in during gaps. A release around 100–300 ms is a good starting point, but adjust by ear to match the natural decay of the room or performer’s breaths.

The Range Control

Many gates offer a range (or depth) control that determines how much the signal is attenuated when the gate is closed. Instead of going to full silence, you can set the range to -20 dB or -30 dB, so that a subtle amount of room tone remains. This is especially useful when you want to avoid the “digital hole” sound that occurs when the gate goes to complete silence. For dialogue in a film, leaving -10 to -20 dB of ambience often sounds more natural.

Sidechain Gating

In some post-production workflows, you may use a sidechain to trigger the gate from a different track. For example, if a boom mic and a lavalier have different noise floors, you can sidechain the gate on the boom track from the lavalier’s voice signal, ensuring the boom gate opens only when the talent speaks. This technique helps maintain a consistent ambience across multiple microphones. Many plugins (such as those from FabFilter, Waves, or iZotope) offer internal sidechain options.

Using Expansion for Subtle Dynamic Control

Expansion Ratio and Knee

An expander’s ratio defines how aggressively low-level signals are reduced. For dialogue, a ratio between 1.5:1 and 3:1 is typical. Higher ratios (4:1 and above) start to behave like a gate. The knee control (if available) smooths the transition around the threshold, making the expansion sound less obvious. A soft knee (6 dB or more) is often preferable for dialogue, as it reduces the pumping effect that can occur with hard-knee processing.

When to Use Expansion vs. Gating

Use an expander when the background noise is moderate (e.g., air conditioning rumble at -45 dBFS) and you want to preserve the room’s natural decay. In a quiet podcast recorded in a treated room, a 2:1 expander set just below the voice level can reduce the hiss without ever fully muting the silence. In contrast, use a gate when the noise is loud enough to be distracting (e.g., a loud air handler or traffic) and you need to remove it entirely between words. Often, an expander is applied before a gate in the signal chain (expander → gate) to handle the subtle reduction first, leaving only the cleanest residue for the gate to catch.

Combining Expansion with Compression

Many dialogue tracks also benefit from a compressor to tame loud peaks and raise the average level. However, compression can also amplify background noise during quiet passages. By placing an expander before the compressor, you reduce the noise floor early in the chain, so the compressor doesn’t pump or bring up noise during pauses. This combination — expander → compressor → (optional) gate — is a standard arsenal for dialogue cleanup.

Advanced Techniques and Workflow

Serial Processing: Expander Then Gate

As mentioned, a serial chain of expander followed by gate can yield excellent results. The expander handles the bulk of noise reduction with a gentle ratio (2:1, soft knee), lowering the floor by 6–10 dB. Then a gate with a fast attack, short hold, and moderate range (e.g., -20 dB) catches any remaining leakage. The gate’s threshold can be set a few dB higher than the expander’s, ensuring it only activates on the very quietest moments.

Parallel Processing for Transparency

For extremely delicate dialogue (e.g., a quiet scene in a film), try parallel gating. Duplicate the dialogue track, apply a heavy gate that fully mutes the noise, then mix that processed version back with the original at a low level (blend). This adds gated “clean” ambience without ever fully replacing the natural room tone. The technique is similar to parallel compression but applied to the noise floor.

Using a De-Esser Before Gating

Sibilant consonants (s, sh, ch) can inadvertently trigger a gate to open or cause it to close too quickly. Placing a de-esser before the gate reduces these problematic frequencies, resulting in more consistent gating. It also helps prevent the gate from “breathing” on sibilant sounds. Many engineers also apply a dedicated de-esser plugin before any dynamics processing.

Practical Application Scenarios

Film and Television Dialogue

In post-production for narrative content, the goal is invisibility. You never want the audience to hear the processor working. Use an expander with a very low ratio (1.5:1) and a high threshold (just below the softest spoken word). This preserves the dynamic range of the performance while lowering the floor by 3–6 dB. Then apply a gate with a hold time of 100–200 ms and a range of -15 dB to keep room tone. Avoid aggressive gating on multi-microphone scenes; instead, use sidechain linking to ensure consistent volume when switching between shots.

Podcast and Voiceover

Podcasters often work in untreated home studios with significant background noise. Here, a gate can be used more aggressively, but combine it with a gentle expander first. A common setup in PODR (Podcast Dedicated Recording) or DAW templates includes: expander (2:1, threshold -40 dBFS) → compressor (3:1, fast attack) → gate (fast attack, 50 ms hold, range -30 dB). Many podcasters find success using dedicated dialogue processors like the Waves CLA Vocals or stock DAW plugins with built-in expander/gate modules.

Live Sound and Broadcast

In live sound, noise gates are essential for preventing feedback and handling multiple open microphones. For broadcast (e.g., radio, news), a downward expander is often preferred over a gate to avoid cutting off the announcer’s breath or the natural decay of the room. The threshold is set to the microphone’s ambient noise level, and the expander ratio is 2:1 to 4:1. Many broadcast consoles include built-in downward expanders specifically designed for voice processing.

Common Mistakes and How to Avoid Them

  • Setting the threshold too high: The gate cuts off words; you hear the “ch ch ch” of consonants being chopped. Fix: Lower the threshold until the softest syllables open the gate cleanly.
  • Release time too short: The ambience drops out abruptly after each word, creating an unnatural, hollow sound. Fix: Lengthen release to match the room’s natural decay; aim for 150–300 ms as a starting point.
  • Range too aggressive (full muting): Complete silence during pauses sounds clinical — especially in a film where you want a sense of space. Fix: Use a range control to leave 10–20 dB of room tone.
  • Using a gate on sibilant-heavy voice: The gate can chatter or open unpredictably. Fix: De-ess first, or use a sidechain filter so the gate triggers only on the voice fundamental.
  • Processing in isolation: What sounds clean in solo may feel lifeless in the full mix. Fix: Always check gating/expansion in context with music and sound effects.
  • Over-processing: Applying both an expander and a gate with too much reduction creates a “breathing” or “pumping” artifact. Fix: Back off on the ratio or range; subtlety is key.

Conclusion

Noise gates and expanders are not merely “noise reduction tools” — they are dynamic processors that shape the background of your dialogue, preserving its naturality and emotional impact. By understanding the difference between gating and expansion, setting timing parameters judiciously, and employing advanced techniques like serial processing and sidechain triggering, you can clean up unwanted noise without compromising the performance. Remember to always compare your processed audio with the original, listen in the context of the full mix, and trust your ears over the meters. With practice, you’ll be able to maintain a wide, expressive dynamic range while delivering a pristine dialogue track that keeps your audience engaged from first word to last. For further reading, consider in-depth resources from Sound On Sound and the iZotope learning hub.