Why EQ and Compression Matter for Voice-Over

Voice-over recordings appear in everything from corporate videos and e-learning modules to audiobooks, podcasts, and commercial ads. Listeners expect a polished, professional sound that feels intimate and clear. Without proper processing, raw recordings often suffer from rumble, muddiness, sibilance, uneven volume, and background noise. Equalization (EQ) and compression are the two most essential tools for transforming a dry take into a broadcast-ready track. Mastering them not only improves intelligibility but also ensures consistency across different playback systems—from studio monitors to smartphone speakers.

Great voice-over sound begins at the source: microphone technique, room treatment, and proper gain staging. Yet even the best raw recording benefits from thoughtful EQ and compression. These processes are not about fixing mistakes but about enhancing the natural qualities of the voice while removing distractions. This guide explains how to use EQ and compression effectively for voice-over, with practical settings, advanced techniques, and common pitfalls to avoid.

Understanding Equalization (EQ) for Voice-Over

EQ allows you to adjust the level of specific frequency ranges in your audio. The human voice spans roughly 80 Hz to 12 kHz, but not all frequencies are equally useful. A voice-over track typically needs to be clear, present, and free from low-end rumble or harsh sibilance. EQ helps you achieve this by boosting or cutting targeted bands.

Key Frequency Ranges in Voice-Over

  • Sub-bass (20–60 Hz): Usually contains mechanical rumble, HVAC noise, and mic handling thumps. Roll off everything below 60–80 Hz with a high-pass filter.
  • Bass (60–250 Hz): Adds body and warmth. Too much can cause muddiness or a boomy, “shirt” sound (the resonance of the chest cavity). A gentle cut around 150–200 Hz often cleans up the low mids.
  • Low mids (250–800 Hz): Can make the voice sound “honky” or boxy if excessive. Cutting narrow bands in this region can reduce a nasal quality.
  • Presence (1–4 kHz): Critical for clarity and intelligibility. A gentle boost near 2–3 kHz adds definition and helps the voice cut through music or background sounds. Be careful—over-boosting here creates an aggressive, piercing tone.
  • Brilliance (5–8 kHz): Adds “air” and openness. A small shelf boost around 8–10 kHz can make the voice sound more modern and crisp. Above 10 kHz may add sizzle but also brings up noise.
  • Sibilance (6–10 kHz): Harsh “s” and “sh” sounds live here. A narrow cut at the offending frequency can tame sibilance without dulling the entire track.

Types of EQ for Voice-Over

  • Parametric EQ: The most flexible type. You control frequency, gain, and Q (bandwidth). Use precise cuts and gentle boosts. Most DAW stock EQs (like FabFilter Pro-Q, Ableton EQ Eight, Logic’s Channel EQ) offer parametric control.
  • Shelving EQ: Boosts or cuts all frequencies above (high shelf) or below (low shelf) a set point. Useful for adding air (high shelf) or reducing excessive low-end (low shelf).
  • Graphic EQ: Fixed bands with sliders. Less precise for corrective work but can be useful for broad tonal shaping.

For voice-over, a parametric EQ with a high-pass filter is usually sufficient. Many engineers recommend starting with a high-pass filter at 80 Hz (adjust based on the speaker’s voice) to eliminate rumble, then listening for problematic frequencies (e.g., muddiness, boom, boxiness, harshness) and cutting them with narrow Q settings. Boost subtly—if you hear a need for more presence, try a wide Q boost of 1–2 dB at 2.5 kHz.

Common Voice-Over EQ Workflow

  1. High-pass filter: Set to 60–80 Hz. If the voice is very deep, start at 50 Hz. Listen to ensure you are not removing natural low-end body.
  2. Sweep for problems: Boost a narrow band (high Q) aggressively and sweep through the low mids (200–500 Hz) to find resonant frequencies that make the sound “honky” or “boomy.” Cut 2–3 dB at that frequency.
  3. Address sibilance: Sweep around 6–9 kHz. If you hear harsh “s” sounds, cut 2–4 dB with a narrow Q. Alternatively, use a dedicated de-esser.
  4. Add air: A high-shelf boost of 1–2 dB at 10–12 kHz can add openness without sounding harsh. For some voices, a gentle peak boost at 8 kHz works better.
  5. Check presence: Boost 2–3 kHz by 1–2 dB with a medium Q. A/B test to ensure clarity without harshness.
  6. Final listen: Play the track at low volume on small speakers. If it sounds muddy, cut more low mids. If too thin, reduce the high-pass filter frequency or add a small low shelf boost at 100 Hz.

Understanding Compression for Voice-Over

Compression reduces the dynamic range—the difference between the quietest and loudest parts of your recording. Voice-overs naturally have wide dynamics: a whisper can be very soft, while an emphatic word can peak. Compression evens out these levels, making the voice more consistent and easier to hear over background audio. It also adds perceived loudness and sustain.

Key Compressor Parameters

  • Threshold (dB): The level above which compression begins. For voice-over, start around -18 dB to -12 dB (depending on your recording level). Adjust so that the compressor engages on peaks but not on every quiet syllable.
  • Ratio (x:1): How much compression is applied above the threshold. A ratio of 3:1 or 4:1 is common for voice-over. Heavy compression (6:1 or higher) can sound unnatural but can be used for dramatic effect or for voice-over in loud environments.
  • Attack (ms): How quickly the compressor reacts after the signal exceeds the threshold. Fast attack (1–5 ms) grabs peaks and evens out sharp transients. Slower attack (10–30 ms) lets the natural attack of the voice through, preserving punchiness. For most voice-over, a medium attack (5–15 ms) works well.
  • Release (ms): How quickly the compressor stops applying gain reduction after the signal drops below the threshold. Faster release (20–50 ms) can cause pumping and breathing. Slower release (100–300 ms) yields smoother compression. For voice-over, aim for a release that resets naturally between phrases—usually 50–150 ms.
  • Knee: Determines how gradually compression begins. Soft knee creates a more musical, gradual compression; hard knee is more aggressive. Voice-over often benefits from a medium soft knee.
  • Make-up Gain (dB): After compression, the overall level drops. Use makeup gain to bring the track back to a desired output level. Compare with the original at the same perceived loudness.

Compressor Types and Voice-Over

Different compressor topologies impart different colors:

  • VCA (Voltage Controlled Amplifier): Transparent and clean, great for general voice-over compression. Example: dbx 160, SSL bus comp.
  • FET (Field Effect Transistor): Fast and punchy. Adds a slightly aggressive character. Good for voice-overs that need energy. Example: Universal Audio 1176.
  • Optical (Opto): Smooth and gentle, with slower attack and release. Ideal for smooth leveling. Example: LA-2A. Many voice-over engineers use opto compressors for a natural, warm sound.
  • Vari-Mu (Variable-Mu): Tube-based, warm and musical. Can add pleasant coloration. Example: Manley Variable Mu.

For beginners, start with an optical compressor emulation (many DAWs include one). Set threshold so you see 2–6 dB of gain reduction on peaks, ratio around 3:1, attack medium (10 ms), release around 100 ms. Listen for any unnatural pumping or breathing and adjust release accordingly.

Common Compression Workflow for Voice-Over

  1. Set ratio to 3:1.
  2. Lower threshold until the compressor shows 2–4 dB of gain reduction on the loudest phrases.
  3. Set attack to 10 ms (medium). If the voice has very sharp peaks (like a plosive), faster attack 3–5 ms may help.
  4. Set release to 100 ms. Listen to a section with varying dynamics. If the compressor “pumps” (level jumps up and down audibly), lengthen the release to 200 ms or more. If the voice feels too squashed, shorten release.
  5. Apply makeup gain so the output level matches the input level (or slightly louder). A/B test with and without compression at the same perceived volume.
  6. Check for distortion or unnatural artifacts. If you hear the compressor working too obviously, back off the threshold or lower the ratio.

A classic voice-over chain often uses two compressors in series: a fast FET compressor (2:1 ratio, fast attack, catching peaks) followed by an optical compressor (4:1 ratio, slow attack, gentle leveling). This yields controlled dynamics without sounding overly processed.

Combining EQ and Compression: Order Matters

The traditional approach is to EQ before compression. The reasoning: cleaning up the audio first ensures the compressor isn’t reacting to unwanted frequencies (like rumble or sibilance). For example, a compressor that triggers on low-frequency rumble will pump unnecessarily, creating an uneven sound. EQ first removes that rumble, so the compressor only processes the desired voice signal.

However, some engineers advocate compressing before EQ. The theory is that compression can bring out subtle resonant frequencies or sibilance, which you then tame with EQ. In practice, both approaches work, and the best order depends on the material and your goal. A safe workflow:

  1. First: High-pass filter (part of EQ) to remove subsonic content.
  2. Second: Compress lightly to even out dynamics.
  3. Third: Additional EQ to shape tone, add presence, and reduce harshness.

This sequence avoids triggering compression on rumble yet allows compression to “smooth” the voice before you apply more transparent EQ shaping. Many modern plugins allow this within a single channel strip (e.g., FabFilter Pro-Q and Pro-C in series).

Advanced Techniques for Professional Results

Multiband Compression

Multiband compression splits the audio into frequency bands (e.g., low, mid, high) and applies independent compression to each band. This is useful for controlling specific issues without affecting the whole signal. For voice-over, you might apply heavier compression to the low mids to control boominess, while leaving the high presence band more dynamic. De-essers are essentially multiband compressors that only compress the sibilance region. Use multiband compression sparingly—it can quickly make a voice sound artificial.

Dynamic EQ

Dynamic EQ combines EQ and compression: it only boosts or cuts a frequency band when the level in that band exceeds a threshold. This is perfect for taming resonances that only occur on certain words or phrases. For example, a dynamic EQ can cut 200 Hz only when the speaker hits a low note that causes muddiness, leaving the rest of the track untouched. Dynamic EQ is more transparent than multiband compression for these spot corrections.

Serial Compression (Two-Stage)

Using two compressors in series, as mentioned earlier, can give you the benefits of both a fast, peak-catching compressor and a slower, leveling compressor. The first compressor (fast attack, low ratio) catches transients; the second (slower, higher ratio) smooths the overall level. This approach is common in professional voice-over chains. For example: first stage – 1176 style (4:1 ratio, fast attack 4 ms, fast release 50 ms) catching 2–3 dB; second stage – LA-2A style (optical, ~4:1 ratio, medium attack 10 ms, auto release) adding another 2–3 dB of gain reduction.

Sidechain EQ (De-essing)

A de-esser is a specialized compressor triggered by a sidechain that is EQ-d to focus on sibilant frequencies (typically 6–9 kHz). When the level in that band exceeds the threshold, the compressor attenuates the whole signal (or just that band, in multiband de-essers). Most DAW compressor plugins have a sidechain input; you can create a de-esser by duplicating the track, adding an EQ that boosts sibilance heavily on the duplicate, and routing that to the compressor’s sidechain. However, dedicated de-esser plugins (like FabFilter Pro-DS or Waves DeEsser) are simpler and often more effective.

Practical Tips for Best Results

Monitoring and Room Acoustics

Equalization and compression decisions are only as good as your monitoring environment. Use closed-back headphones (e.g., Beyerdynamic DT 770 Pro or Sony MDR-7506) for critical listening, especially if your room is not acoustically treated. Open-back headphones can provide a more natural sound but may leak audio into the mic. Always check your mix on multiple systems: earbuds, laptop speakers, car audio, and a subwoofer if possible.

Room acoustics affect how you perceive low frequencies. Boomy rooms can trick you into cutting too much low-end; dead rooms can lead to excessive low-end boost. If you are recording in an untreated space, treat it with bass traps and absorbers. Even a portable vocal booth can help.

Gain Staging

Before applying EQ or compression, ensure your recording levels are healthy but not clipping. Aim for peaks around -12 dBFS to -6 dBFS in your DAW. Leave headroom for processing. Each plugin can add gain, so watch the overall output level to avoid digital clipping. Use a true peak limiter as a safety measure at the end of your chain.

Use Your Ears, Not Your Eyes

It is tempting to look at frequency analyzers and gain reduction meters, but your ears should be the final judge. A/B test your processing often. Solo the track and listen for changes in tone, clarity, and naturalness. Compare with reference tracks from professional voice-over work. Over time, you will develop a mental map of what “good” sounds like.

Avoid Over-Processing

Less is often more. Many professional voice-over engineers use very gentle EQ and compression—1–2 dB of boost or cut, 2–4 dB of gain reduction. Aggressive processing can make the voice sound thin, lifeless, or “squashed.” If you need heavy processing, consider whether the source recording quality is adequate. A good recording requires minimal processing.

Use Reference Tracks

Import a professionally recorded voice-over (e.g., from a commercial or audiobook) into your session. A/B between your processed track and the reference at the same volume. Listen for differences in tonal balance, presence, dynamics, and overall clarity. This practice trains your ears and sets a target to aim for.

Consider the End Delivery Format

Voice-over for a podcast may require less compression than for a TV commercial, where every word must cut through music and sound effects. Audiobooks need a consistent level over long durations. Film narration often uses a wider dynamic range. Tailor your processing to the medium. For example, iZotope’s guide to EQ provides a solid foundation for understanding how different genres apply EQ.

Common Mistakes and How to Fix Them

  • Boosting too much low-end: Makes the voice sound muddy or boomy. Instead, cut the low mids. Listen to Sound on Sound’s voice-over recording tips for corrective EQ strategies.
  • Over-compression: Results in a lifeless, unnatural voice that lacks dynamics. Reduce the ratio or threshold. Use a compressor with auto-makeup gain or manual careful makeup to avoid constant high level.
  • Wrong attack/release times: Too fast attack can squish transients and rob the voice of punch. Too slow release can cause pumping. Experiment with timing to match the speaker’s natural pacing.
  • Not addressing sibilance: Harsh “s” and “sh” become worse after compression. Use a de-esser before or after compression. WikiHow’s de-esser tutorial explains the steps clearly.
  • Ignoring the room sound: EQ cannot fix a boxy, reverberant room. Treat the recording space first. Post-processing can only do so much.

Final Thoughts

Mastering EQ and compression is a journey of listening, experimentation, and refinement. Start with simple high-pass filtering and gentle compression, then slowly add more advanced techniques as your ear develops. The key is to preserve the natural quality of the voice while enhancing its clarity and consistency. Every voice is different—what works for one speaker may not work for another. Trust your ears, use appropriate tools, and always check your work on multiple playback systems.

For further reading, consult Production Expert’s voice-over processing tutorial and the free resources from BBC Academy’s voice-over guide. Practice consistently, and soon you will be able to dial in a polished voice-over sound in minutes.