Dynamic vocal recordings are a common challenge in music production. A singer might whisper a soft verse before belting a chorus, creating a wide discrepancy in loudness. Traditional single-band compression can help, but it often affects the entire frequency spectrum, dulling transients or exaggerating tonal imbalances. Multiband compression offers a more surgical solution: it divides the vocal signal into separate frequency bands and applies independent compression settings to each. This allows engineers to tame specific problems like harsh sibilance, plosive boom, or resonance buildup without harming the natural character of the voice. When used correctly, multiband compression turns an uneven vocal performance into a polished, professional recording that sits seamlessly in a mix.

What Is Multiband Compression?

Multiband compression is a dynamic processing technique that splits the audio spectrum into two or more frequency regions using crossover filters. Each band then passes through its own compressor with independently adjustable parameters: threshold, ratio, attack, release, and makeup gain. The compressed bands are summed back together at the output. This differs from a standard wideband compressor, which processes the entire signal with a single set of controls. By isolating frequencies, multiband compression can apply heavy gain reduction to a problematic range (e.g., 5–8 kHz sibilance) while leaving other frequencies untouched. Common configurations are three-band (low, mid, high) or four-band, though some plugins allow arbitrarily many bands. The frequency bands are defined by crossover points: for example, low band below 200 Hz, mid band from 200 Hz to 4 kHz, and high band above 4 kHz. Each band’s compression acts independently, making it possible to compress a hard "sss" sound aggressively without squashing the warm body of the vocal.

Why Use Multiband Compression on Vocals?

Vocals are one of the most dynamic audio sources. Singers naturally vary volume, pitch, and timbre throughout a performance. Furthermore, the proximity effect (increased bass when close to the microphone), breath noise, room resonances, and plosives create inconsistencies that are difficult to fix with EQ alone. A standard compressor might reduce the overall dynamic range but can’t differentiate between a loud low-frequency rumble and a loud high-frequency sibilance. Multiband compression addresses each issue in its own frequency zone. For example, you can compress the low mids (200–400 Hz) to reduce boxiness from a singer moving closer to the mic, while simultaneously compressing the high frequencies (6–10 kHz) to tame harshness from belted notes. The result is a more even, polished vocal tone that maintains clarity and presence.

Key Problems It Solves

  • Sibilance and harshness: High-frequency "sss," "sh," and "t" sounds can be sharp and irritating. Multiband compression allows aggressive gain reduction in the 5–10 kHz range without dulling the rest of the vocal.
  • Plosives: "P" and "B" sounds create low-frequency pops. Compressing the low band (below 100 Hz) can reduce the thump without affecting the vocal body.
  • Inconsistent projection: A singer may go from soft to loud in one phrase. Multiband compression can even out the energy across frequencies so that the vocal remains balanced.
  • Resonant frequencies: Room modes or microphone peaks cause certain notes to sound overly loud. A narrow band around that frequency can be compressed to smooth the response.
  • Proximity effect: When a singer moves close to the mic, the low end swells. The low band compressor can react to that extra bass without affecting the mids and highs.

How Multiband Compression Works on Vocals

To use multiband compression effectively, you need to understand the three core elements: crossover points, per-band compression parameters, and how they interact. Most modern multiband compressors have a visual interface showing the frequency spectrum with adjustable crossover lines. You set these crossovers based on where the vocal’s problematic frequencies reside. Common ranges for vocals:

  • Low band: 20–200 Hz. Handles rumble, proximity effect, and plosives.
  • Low-mid band: 200–800 Hz. Controls boxiness, muddiness, and body inconsistencies.
  • Mid band: 800 Hz–4 kHz. Affects presence, nasality, and overall loudness.
  • High band: 4–20 kHz. Targets sibilance, air, and harshness.

Once crossovers are set, adjust each band’s threshold so that the compressor only activates when the level exceeds the threshold. The ratio determines how much gain reduction is applied: lower ratios (2:1 to 3:1) are transparent for vocals; higher ratios (5:1 or more) can squash a band unnaturally. Attack and release times are critical: faster attack (0.5–5 ms) helps catch transients like sibilance; slower attack (10–30 ms) allows the natural transient through but compresses the sustain. Release times should be set to follow the vocal phrasing, typically 20–80 ms for highs and 50–150 ms for lows. Makeup gain is applied to each band if needed, but it’s often better to let the bands sum naturally and use global output gain control.

Setting Crossover Frequencies for Vocals

Choosing crossover frequencies is the most creative part of multiband compression. Every voice is different, but a good starting point for a lead vocal is three bands with crossovers around 200 Hz and 4 kHz. This divides the signal into low, mid, and high. You can then fine-tune based on the specific recording. For example, if a vocal has excessive nasal energy around 1–2 kHz, you might split the mid band into two by adding a fourth crossover at 1.5 kHz. Always listen and use spectrum analyzers to identify problem areas. Be cautious about placing crossovers too close together – this can create phase cancellation and comb filtering at the crossover region. Typical crossover slopes (12–24 dB/octave) minimize phase issues, but overlapping bands can still cause muddiness if not tuned carefully. When in doubt, use broader bands and rely on the compressor’s threshold to target specific frequencies.

Adjusting Compression Parameters per Band

Each band requires different compression behaviors. For the low band, use a slower attack (10–30 ms) to let the initial transient of a plosive through – compressing the attack can make the pop sound worse. A ratio of 2:1 to 4:1 is sufficient to smooth out proximity effect. For the mid band, where the vocal’s core resides, a gentle ratio (2:1 or 3:1) with a medium attack (5–15 ms) and release (40–80 ms) helps maintain consistent loudness without killing expression. The high band is where aggressive compression is often needed: fast attack (0.5–3 ms), higher ratio (4:1 to 6:1), and fast release (10–30 ms) to clamp down on sibilance. Some engineers even use a limiter on the high band for complete control. Always listen to each band in solo mode to understand what it’s doing, but make final decisions in the context of the full mix.

Practical Workflow: Using Multiband Compression in a Mix

Here is a step-by-step workflow for applying multiband compression to a vocal track. This assumes you have already done basic editing (gain riding, noise removal, EQ, and de-essing if needed).

  1. Insert a multiband compressor on the vocal track. Popular plugins include FabFilter Pro-MB, iZotope Neutron, Waves C4/C6, and the stock compressor in your DAW.
  2. Identify problem areas. Play the vocal and listen for specific issues: sibilance, boxiness, resonant peaks, loudness variations. Use a spectrum analyzer to visualize frequency imbalances.
  3. Set the number of bands and crossover frequencies. Start with three bands: low (<200 Hz), mid (200–4 kHz), high (>4 kHz). Adjust crossovers based on what you hear. For example, if sibilance is extreme, set the high crossover lower (e.g., 5 kHz) so that band only contains the sibilant region.
  4. Set thresholds. Reduce the threshold of each band until you see 2–6 dB of gain reduction on the loudest parts. Avoid heavy compression (more than 8–10 dB) on any single band unless the problem is severe.
  5. Adjust ratios. Use 2:1 to 3:1 for low and mid bands; 4:1 to 6:1 for the high band if sibilance is aggressive. Lower ratio for more natural sound.
  6. Tweak attack and release. For high frequencies, use fast attack (1 ms) and fast release (15–20 ms). For low frequencies, slower attack (20 ms), slower release (80–100 ms). Mid band: medium attack (10 ms), medium release (50 ms).
  7. Apply makeup gain per band if necessary – some plugins automatically compensate. Then use the plugin’s output gain to set the vocal level back to its original perceived loudness.
  8. A/B compare with the original. Toggle the plugin bypass. The compressed version should sound smoother and more consistent without obvious pumping or tonal changes. If the vocal sounds dull, you may have overcompressed the high band; if it sounds thin, the low band may be over-attenuated.

This workflow is a starting point. Every vocal is unique, so trust your ears and make incremental adjustments.

Advanced Techniques

Sidechaining with Multiband Compression

Multiband compressors often support sidechain input. You can use another track (like a synth pad or guitar) to trigger compression on specific frequencies of the vocal. For example, sidechain the vocal’s low band from a bass guitar to automatically duck the vocal’s low end when the bass plays, avoiding frequency masking. This is more transparent than a wideband sidechain ducking.

Using Multiband Compression for De-Essing

While dedicated de-essers work well, multiband compression can serve as a more flexible de-esser. Set a band with crossover around 5–8 kHz, use a very fast attack (0.3 ms), a high ratio (6:1 or higher), and fast release. This will clamp down on sibilant sounds without the need for a separate plugin. You can also apply it only when needed by adjusting the threshold high enough to only catch the loud sibilants.

Dynamic EQ vs Multiband Compression

Both techniques can address frequency-specific problems, but they work differently. Dynamic EQ adjusts gain at a specific frequency based on the level, while multiband compression compresses an entire band. Multiband compression is better for smoothing wide frequency regions, such as controlling the overall low-mid energy. Dynamic EQ is more precise for catching a single resonant note. Combining both can yield excellent results: use dynamic EQ to fix a honky 800 Hz note, then multiband compression to control the general low-mid consistency.

Parallel Multiband Compression

Send the vocal to an auxiliary channel with a multiband compressor set aggressively (e.g., high ratio, low threshold). Blend the compressed signal with the dry vocal. This retains the original dynamics while adding a controlled, denser version underneath – similar to parallel compression but frequency-sculpted. This technique is powerful for adding perceived loudness without obvious artifacts.

Common Mistakes to Avoid

  • Over-compressing a single band: Reducing gain by 10+ dB on the high band can cause the vocal to sound dull and lifeless. Aim for subtle reduction – 3–6 dB is usually enough.
  • Placing crossovers too close: This can create phase cancellation and a "swooshing" sound. Keep a gap of at least an octave between crossover frequencies.
  • Forgetting to set proper attack/release: A fast attack on the low band can destroy the punch of a vocal. Always match attack/release to the energy of the frequency band.
  • Ignoring makeup gain: After compressing, the overall level drops. Without makeup gain, the vocal may sound weaker than the original, leading you to push the compressor harder.
  • Using multiband compression as a substitute for good recording: It cannot fix distorted recordings, poor mic technique, or excessive room bleed. Proper source capture remains essential.
  • Not checking in mono: Some plugins can introduce phase issues between bands. Listen in mono to ensure the vocal doesn’t lose body or sound hollow.

Genre-Specific Considerations

Pop and Rock

In pop and rock, vocals need to be consistently loud and present. Multiband compression is often used aggressively to reduce dynamic range, especially in the high band to control belted notes. Apply moderate compression (3:1 ratio) on the mid band to keep the vocal front and center. Often a final multiband compressor on the vocal bus can glue all layers together.

Jazz and Classical

These genres value natural dynamics and tonal fidelity. Use multiband compression sparingly, if at all. A low ratio (1.5:1) on the high band only to tame occasional harshness, and maybe a gentle low band to control proximity effect. The goal is subtle correction, not smoothing everything flat.

Hip-Hop and R&B

Hip-hop vocals often have extreme dynamic contrast between verses and hooks. Multiband compression can help even out the level while preserving the character. Use a faster attack on the high band to control sibilance and a slower attack on the low band to keep the weight of the voice. Some producers like to use a multiband compressor to emphasize certain frequencies by heavily compressing others – for example, compress the mids to make the highs and lows stand out.

Metal and Hard Rock

Aggressive vocals (screams, growls) need extreme control. High bands often require heavy compression to avoid ear-fatiguing frequencies. Low bands can be compressed to remove rumble and focus the aggression. Use higher ratios (5:1) and fast attack times. Be careful not to squash the expression – leave some dynamic contrast for impact.

Conclusion

Multiband compression is a powerful tool in any audio engineer’s arsenal, especially for taming unruly vocal recordings. By dividing the frequency spectrum and applying independent compression, you can address specific problems like sibilance, plosives, and boxiness without compromising the natural tone of the voice. The key is to listen critically, set crossovers and parameters carefully, and apply compression subtly. When blended with good recording practices, EQ, and dynamic processing, multiband compression elevates vocals from amateur to professional. Experiment with different settings, plugins, and techniques – the more you practice, the more intuitive this process becomes. For further reading, check out Sound On Sound’s multiband compression guide and explore reviews of the best multiband compressor plugins to find the right tool for your workflow.