Mixing Vocals Cheat Sheet

Most vocal advice online repeats round numbers for high-pass filters, mud cuts and de-essers without saying where they came from. This mixing vocals cheat sheet puts the EQ frequencies, compression ratios, gain-reduction targets, de-esser ranges, reverb decay times and delay times on one page. Every quoted figure links to the guide or spec sheet it comes from (iZotope, Waves, Universal Audio, Valhalla DSP and Sound On Sound), and the one table that’s calculated says so. Where those sources disagree, you’ll see both figures, and where none of them commits to a number, this page doesn’t invent one.

Written by YECK, founder of MixingGPT. MixingGPT isn’t one of the sources or tools in this cheat sheet.

Mixing Vocals Cheat Sheet: Every Starting Point on One Page

These are starting points, not presets. Each one assumes a reasonably recorded lead vocal; the sections below explain when to move away from it. The source column links to the page each number comes from.

Vocal EQ Frequency Chart for Mixing Vocals

The ranges below come from iZotope’s vocal EQ guide, its mixing glossary and its pop vocal EQ article, with the fundamental ranges from iZotope’s article on creating space for vocals. The low-mid rows overlap: “mud,” “body” and “thickness” describe the same region from different angles: too much of it, the right amount and too little.

The sources don’t draw the bands identically. iZotope’s guide puts presence at 3–5 kHz, its glossary at about 2–6 kHz, and Waves’ vocal production guide calls 1–5 kHz the upper mids for clarity and presence. Waves also splits the low end differently: 100–250 Hz for warmth (too much sounds boxy) and 250 Hz–1 kHz for body (too much sounds honky or nasal). Treat the chart as a map of where to listen first, not a list of fixed cuts.

For a longer walk through lead, backing and ad-lib EQ, see our vocal EQ frequency guide.

High-Pass Filter Settings for Vocals

iZotope’s vocal EQ guide says a high-pass filter is commonly used to cut rumble below 80–100 Hz, and in its worked example the filter sits at 90 Hz with a 48 dB per octave slope. Its pop vocal article gives a starting guideline of around 100 Hz for a male vocal and 150 Hz for a female vocal.

Why not just filter higher to be safe? Because the voice itself starts low. iZotope puts the fundamental of a male vocal at 100–300 Hz and a female vocal at 200–400 Hz. A filter at 150 Hz is already inside the bottom of a low male voice, so the number depends on the singer, not the gender label. iZotope’s high-pass filter article adds three cautions: steep slopes can ring or distort around the cutoff, stacking high-passes across a mix makes it feel anemic, and the same setting shouldn’t go on every track.

Finding Boxy, Nasal and Ringing Frequencies

The chart tells you where to look; boost-and-sweep tells you exactly where the problem is on this voice. iZotope publishes two versions of it, differing mainly in how wide and loud the search band starts:

• Narrow and loud: its resonant frequencies article suggests a narrow bell boosted by about 12 dB, swept across the range while the problem section plays, then turned into a cut once the resonance jumps out.
• Wide, then narrowed: the pop vocal article starts with a wide bell with a Q of around 1, boosted by about 5 dB and swept; once a resonance shows up, it narrows the Q and cuts at that frequency.

One useful detail from iZotope’s EQ guide: in its example, a nasal resonance at 931 Hz is described as fairly constant for a given singer, because it depends on the singer’s nasal cavity more than on the song’s key or melody. In the 1.5–5 kHz range the same guide keeps moves to no more than 1 or 2 dB, because that’s where intelligibility lives. If the muddiness is the main problem, our guide to fixing muddy vocals in the 200–500 Hz zone goes deeper, and fixing vocal harshness covers the 2–5 kHz harshness region with dynamic EQ and resonance suppressors.

Presence and Air: How Much to Boost

iZotope’s and Universal Audio’s air figures sit between about 10 and 16 kHz (Waves draws its “air” band lower, from 5 kHz up). iZotope’s glossary defines air as content above 12 kHz and “lift” as an airy boost often at 10–16 kHz. In its EQ guide example, iZotope uses a high shelf of about 3 dB from 12 kHz on a dark recording and says it would start the shelf at 16 kHz on a bright one. Universal Audio’s vocal guide suggests a gentle boost around 16 kHz on its Pultec-style EQ.

Watch the sibilance when you boost. The same iZotope example adds about 1.5 dB at 9,357 Hz for sparkle and notes that it also raises sibilance, and the pop vocal article warns to be careful boosting around 4–9 kHz for the same reason. That’s the practical argument for putting a de-esser after your brightening EQ, or for brightening less.

Vocal Compression Settings: Ratio, Gain Reduction, Attack and Release

Ratio

iZotope’s vocal compression article (Chrissy Tignor) gives three tiers: 5:1 as a good place to start with pop vocals, 7:1 or 8:1 for heavier compression, and 2:1 or 3:1 for a more natural sound. It sets the threshold relative to the quietest word in a phrase rather than a fixed dB value, and says to add back as much makeup gain as the meter shows in reduction (5 dB of reduction, 5 dB of makeup).

Gain reduction

iZotope’s eight-step vocal mixing article and its basic vocal chain article both land on about 3–7 dB of gain reduction on a main vocal. Universal Audio’s vocal guide aims for 2–5 dB on an LA-2A and says to A/B the plugin on and off. The ranges overlap at 3–5 dB.

Attack and release

None of the sources here commits to a single attack or release time for vocals, so the useful numbers are the ranges of the two compressor designs that both the Waves and Universal Audio chains on this page use:

Even the 1176’s slowest attack, 800 microseconds, is under a millisecond, which fits the job both Waves and Universal Audio give it in a vocal chain: catching consonants and peaks. The LA-2A’s release helps explain the “smooth compression” role Universal Audio gives it: half the gain reduction recovers in 0.06 seconds and the rest over 0.5 to 5 seconds. Universal Audio’s guide suggests starting an 1176 at medium-to-fast attack, fast release and moderate gain reduction, and reminds you that its attack knob works backwards (clockwise is faster).

Compression can’t do all the leveling. Mike Senior makes the point in Sound On Sound’s Compression Made Easy: a compressor isn’t intelligent enough to keep a lead vocal in place on its own, so fader automation finishes the job. Our vocal compression techniques guide covers serial setups, and the vocal automation guide covers the rides.

De-Esser Settings: Where Vocal Sibilance Actually Sits

iZotope’s de-essing guide and Waves’ de-essing article both put sibilance between 4 and 10 kHz. Waves’ vocal production guide narrows it to 5–10 kHz, and iZotope’s example de-esser band starts at 4.4 kHz. The exceptions are the useful part of the cheat sheet:

• Lower than you’d expect: iZotope says sibilance can drop to about 1.5 kHz with certain singer and microphone pairings, and its author has heard it below 2 kHz on home recordings.
• Belting: iZotope’s de-essing guide says singers hit esses harder around 3 kHz when belting. Separately, its basic vocal chain article sets the de-esser to tame sibilance above 3 kHz.
• Range before threshold: Waves suggests starting with threshold and range at 0 dB, then setting the range (the maximum gain reduction) high and lowering it until the result sounds natural.

How to tell you’ve gone too far: iZotope’s EQ guide warns that over-de-essing gives the singer a lisp and hurts intelligibility. Our de-esser plugin comparison compares the dedicated tools if your DAW’s stock de-esser isn’t enough.

Vocal Reverb Settings: Decay, Pre-Delay and Reverb EQ

Use a send. Universal Audio’s guide says not to insert reverb directly on the vocal track, iZotope’s vocal reverb tips recommend a 100% wet reverb on an aux channel, and Valhalla DSP’s best practices note that sends are more CPU-efficient and let you shape the effect’s tone separately.

• Decay: Waves’ vocal production guide suggests about 0.5–1.5 seconds for an intimate sound and 2 seconds or more for a spacious one; its plate example uses a decay of about 1 second.
• Reverb EQ: Valhalla DSP’s general reverb tips suggest a low cut at 200–400 Hz, or higher, for a clearer reverb, and a high cut set fairly low, such as 2–4 kHz. That advice is for reverb in general, not vocals specifically.
• Pre-delay: iZotope’s algorithmic reverb article says pre-delay separates the dry vocal from the start of the reverb and helps keep a lead vocal intelligible, and Waves’ reverb explainer says more pre-delay makes the singer sound closer. Neither gives a millisecond figure, so set it by ear: raise it until the start of each word is clear of the reverb.
• Diffusion: Valhalla’s Sean Costello explains that a reverb with high diffusion can sound metallic on vocals, and that lower diffusion usually sits better in a mix.
• Level: Universal Audio’s test is to raise the send until you clearly hear the reverb, then pull back until you’d miss it if it were gone. iZotope’s rule of thumb: when in doubt, back it off by 10–25%.

Reverb choices by genre are in our vocal reverb plugins and settings guide.

Vocal Delay Settings: Slapback, Synced Echoes and Doubling

• Slapback: Waves’ vocal production guide puts slapback at around 80–150 ms with minimal feedback, and its famous vocal tricks article uses about 100 ms for a Phil Collins-style slap.
• Rhythmic delay: quarter- or eighth-note timing synced to tempo, with feedback around 10–20% for subtle echoes, per the same Waves guide. The Bowie “Let’s Dance” example in the tricks article uses a synced dotted 1/8 note.
• Doubling: Waves’ double-tracking article suggests duplicating the vocal twice, delaying one copy by 10 ms and the other by 20 ms, and panning them hard left and right. Different times on each side keep it from collapsing into a mono slapback.

If your delay plugin doesn’t sync to tempo, the arithmetic is simple: one beat (a quarter note) lasts 60,000 ÷ BPM milliseconds. Halve it for an eighth note, take 75% of it for a dotted eighth, and quarter it for a sixteenth.

Calculated from 60,000 ÷ BPM and rounded to one decimal place where needed.

At 120 and 140 BPM, a synced sixteenth (125 ms and 107.1 ms) falls inside Waves’ 80–150 ms slapback range, so a sixteenth-note delay can double as a slapback that sits in time. For tape, digital and character delays, see our delay plugin comparison.

Saturation Settings for Vocals

Saturation has fewer published numbers than EQ or compression. iZotope’s tape emulation guide describes the speed switch on tape plugins this way: 30 IPS is more transparent with better high-frequency response, 15 IPS is the standard balance of warmth and fidelity, and 7.5 IPS is lo-fi and bass-forward. It also suggests tape on a bright or harsh vocal to soften the top end gently, which it says can sound more natural than an EQ cut or de-esser. In its eight-step vocal example, iZotope runs Nectar’s Tape saturation at about 50%. Our saturation plugin comparison covers the options.

Vocal Chain Order for the Settings on This Cheat Sheet

The order changes what each setting does: a compressor after a bright EQ reacts to the boost, and a de-esser before that EQ can’t catch the sibilance the EQ adds. iZotope’s basic vocal chain runs:

1. Pitch correction, first, because it works best on a dry vocal
2. EQ
3. Compression
4. De-essing, to tame the sibilance EQ boosts can bring up
5. Delay and reverb, in whichever order suits the sound

Waves’ plugin order article agrees on tuning before EQ, saturation, reverb and delay, and puts noise reduction (and de-reverb, if the recording needs it) before the tuner, because noise and room tails confuse pitch detection.

Where the sources split: two compressors. Waves’ CLA-2A and CLA-76 article puts the 1176-style CLA-76 first for consonant control and the LA-2A-style CLA-2A second for smooth leveling across phrases. Universal Audio’s LUNA chain does the reverse: LA-2A for smooth compression, then a Pultec EQ, then an 1176 at the end to catch the transients the LA-2A lets through. Both are documented approaches; the first leans on fast peak control early, the second on smooth leveling early. For a full chain built step by step, see how to mix vocals step by step.

Common Mistakes When Using a Mixing Vocals Cheat Sheet

• Copying a number without listening. Every frequency above is a range because voices, microphones and rooms differ. iZotope’s de-essing guide alone places sibilance anywhere from about 1.5 kHz to 10 kHz.
• High-passing into the voice. A filter above the singer’s fundamental thins the vocal. Check against the 100–300 Hz and 200–400 Hz ranges.
• Big moves in 1.5–5 kHz. iZotope keeps them to 1–2 dB, and its glossary puts harshness at 2.5–5 kHz, inside that same range.
• Brightening after the de-esser. An air boost added later can bring back the sibilance the de-esser removed.
• Reverb on an insert at full mix. Use a send at 100% wet so the dry vocal stays clean and the reverb can be EQ’d on its own.
• Expecting compression to replace automation. Sound On Sound’s point stands: ride the fader for the words that matter.

If the vocal still floats on top of the track or disappears into it after all of this, our guide to getting vocals to sit in the mix lists the usual causes, and the AI mixing plugins guide compares the AI assistant plugins that work alongside a manual chain.

How This Cheat Sheet Was Put Together

Every quoted frequency, ratio, time and percentage on this page comes from a published source, linked where it’s used. In summary: iZotope’s blog articles and glossary on vocal EQ, high-pass filtering, resonances, compression, de-essing, reverb and tape; Waves’ articles on vocal production, de-essing, double-tracking, plugin order, reverb, compressor stacking and famous vocal effects; Universal Audio’s vocal guide and its 1176LN and LA-2A spec pages; Valhalla DSP’s reverb articles; and one Sound On Sound technique article. The delay-time table is calculated, not quoted. Nothing here comes from hands-on testing or listening tests. Where no source gave a number (reverb pre-delay, vocal attack and release times), this page says so instead of filling the gap.

Mixing Vocals Cheat Sheet: Frequently Asked Questions

A note on freshness: every figure on this page was checked against its linked source on October 13, 2026. Vendors revise their guides and occasionally move articles, so if a link has changed, check the vendor’s current version of that page.