Jp Mixing

Japanese-Style Mixing Techniques (jp-mixing)

Sources:

• Kosuke Nakamura, "DTMミックスのコツが一冊で分かる本" (mix recreations of landmark albums across 17 genres, Rittor Music)
• "クリエイター＆エンジニア直伝！プロの曲作りテクニック"
  • Hi-hat production techniques: SUI / Matt Cab / MK / mabanua
  • Rhythm-section saturation: Satoshi Fukuda
  • Low-end strategy (below 50Hz): Ayaka Toki
  • Vocal-triggered sidechain: Atsushi Asada (AmPm)
  • 7 vocal effects: MATZ / TeddyLoid / T-SK / JUVENILE / yasu / MK / murozo
• "プロの曲作りが分かる本" Chapter 7, engineer-style arrangement techniques (Nagie)

Core Principles

• Before mixing, form a clear mental image of the "ideal sonic picture"; there is no standard answer chart for EQ — no theory says "boost XXdB at XXHz on paper and it's good."
• Imagination comes from listening to lots of music and expanding your "drawers" (mental reference library) — with enough drawers, you can judge the best mixing approach the moment you hear the material.
• Workflow: repeatedly compare "full mix → Solo → full mix", and frequently A/B with plugin On/Off.
• Listening to a processed part in Solo for too long destroys your judgment — always return to the full mix to confirm.
• The mix "box" has a fixed size: if one thing comes forward, another must step back.
• Prioritize the low-frequency instruments first:
  • "The Kick is the star throughout the song" → the Bass yields;
  • "The Bass entrance is the hook" → set the Kick with that as the premise;
  • Piling both up to the max (morimori) never works — without a priority order there is no chance of winning.
• Every genre has a law of "make the highest-priority instrument bigger":
  • Reggae = bass (whole drum kit −5dB / bass +5dB)
  • Hip-hop = drums + vocals
  • Metal = guitar (mixing order: guitar → drums → bass)
  • Shoegaze = guitar wall (vocals half-buried in it)
  • New Soul = bass at the same volume as the whole drum kit
• All numbers are "starting points," not "correct answers" — adjust while the song plays and stop when it sounds cool.

1. Low-End Strategy (below 50Hz = the key to the modern sonic image)

1.1 Fundamentals

• Low-End ≈ below 50Hz; the lower it goes, the harder pitch is to perceive — less a "note" than an earth-rumbling "pressure."
• Turning point: around 2011, James Blake's breakthrough — extremely sparse note counts yet cool, with the Low-End playing a huge role.
• A track with a full low band makes even the highs feel extended — the entire vertical "sense of scale" is completely different.
• Why a Kick sounds fat: its energy is dense in the lows — even if the peak sits around 50Hz, you hear it together with the accompanying content up to about 100Hz.
• The 50Hz of a guitar and the 50Hz of a Kick, soloed at equal volume, actually differ little → the 50Hz content of the top-layer instruments (uwamono) should be cleaned out to make way.
• Music where the "Low-End works" fundamentally shapes its impression via the lows-to-Low-End of Kick and Bass.
• Monitoring checks:
  • Phone/laptop speakers cannot reproduce true lows, but the attack components and midrange of Kick and Bass must still convey the groove to pass — always verify balance via phone playback.
  • A small single-driver speaker (AURATONE 5C, the classic trusted by Dr. Dre) surviving a Mono check = solid low-end work; if it differs too much from your stereo monitors, find out why.
  • Sub-bass monitoring relies on headphones: FOCAL Clear Professional — 20-30Hz is not only "visible" but "trackable" (transients and release movement are rendered faithfully); judging whether gate/release processing is right depends entirely on this.

1.2 Core Low-End EQ Techniques

• Boost lows with a wide Q to expand "area":
  • Example: boost about 4dB with a wide Q around 46Hz, so the energy occupies a broad base rather than a narrow peak;
  • Toki's approach: use wide Q for both low-end boosts and cuts.
• Always check two side effects after boosting:
  1. Excess bloat appears above (around 120Hz) → cut it;
  2. A sub-low bump at the attack transient (around 17Hz) → use a dynamic EQ to suppress it only at the attack instant (FabFilter Pro-Q 3) — only in the offending moment, only in the specific band, without sacrificing the overall tone.
• For low-band attenuation, also avoid Low-Cut/Shelving blanket cuts:
  • Use small wide-Q Bell adjustments (example: cut about 1dB around 11.5Hz);
  • Reason: a Low-Cut cuts too much and destroys the balance with other low layers;
  • Do not kill everything below 20Hz with shelving — preserve natural sub-low extension.
• To make room for the bass's lows: precisely (pinpoint) attenuate the unneeded low content of Snare and similar.
• Directional summary: "Raise the band you 'want to hear a bit more of,' then cut and shape the excess that got raised along with it."

1.3 Kick Processing Chain (House-style, goal: "lows → closer to the listener")

Step 1: Compression to unify the sense of speed (when the attack arrives before the lows and you want both to "hit at once")

• SOFTUBE Tube-Tech CL 1B: has a Dry/Wet knob so it can serve as parallel compression;
• Fully Wet sacrifices the lows → use Dry/Wet to find the point with "the ideal amount of body without burying the lows";
• Adjust Attack/Release Time to tune the compression's "speed feel" so the lows and the attack arrive nearly simultaneously;
• Ratio: basically a very low value is enough.

Step 2: Post-compression compensation EQ (Pro-Q 3)

• Boost 2.5dB around 30Hz — the part of a real Kick that "strikes the air" in a concert hall;
• After the boost, the area above sounds excessive → cut 0.7dB around 57Hz;
• Around 210Hz the pitch character is too strong → attenuate slightly (this band will be filled by chordal instruments; the Kick has no need to assert itself here);
• Slightly lift around 6kHz to restore the Kick's high-frequency information.

Step 3: Sub-low enhancement + Gate one-two punch

• BOZ DIGITAL LABS Bark of Dog (free version available): lifts 13Hz, enhancing centered on the sub-lows, giving the Kick a true "foundation";
• After boosting 13Hz the tail drags → cut the excess ring with Pro Tools Dynamics III Gate;
• Logic: "want the sub-low body, don't want the sluggishness it brings."

Step 4: Two-stage Release strategy to reinforce the release (in House, the Kick is the "axis" of the whole track)

• First plant a pivot with EQ: SONNOX Oxford EQ, about +1dB around 36Hz;
• iZOTOPE Ozone 8 Equalizer:
  • Boost around 2kHz for the "drumhead ring" to add dimensionality;
  • Afterwards it sounds a bit light → go back and boost 86Hz;
  • Boost about 2dB around 6kHz so lows and highs share the same sense of speed;
  • Cut the loose, dragging energy at the very bottom;
  • Set Q mostly at "1-strong" (slightly above 1).
• Ozone Dynamics single-track compression: Release about 37ms (tail stays tight; the goal is "bring the sound closer"), low-band output makeup about +1.7dB;
• Then insert a Maximizer: about 1dB of firm reduction;
• Drums & percussion Bus compressed again: Release about 467ms (on the low-to-sub-low band, letting the Kick's release blend snugly into the whole);
• Finally Maximize again: 3-4dB reduction.
• Summary: EQ boosts the lows → "compress → maximize" the same band; another round on the Bus — two stages with different Releases (37ms pulls closer / 467ms blends) reinforce the Kick's release.

Headroom management (pre-mastering)

• Peaks around −8dB;
• Levels of the main elements clustered around −16dB;
• Low-to-sub-low tracks within 2-3dB of each other (level gaps between low-frequency instruments should stay small);
• Overall around −18dB, leaving margin for mastering.

1.4 Synth Bass / 808 / Sub Layering

Three-layer structure (real example: Noritake Kinashi "OTONA feat. Toshinobu Kubota," mixed by Ayaka Toki)

• Layer (1): spans from "sharp" to "elastic and meaty," most audible, carries the pitch;
• Layer (2): adds flesh to (1) and pushes the center of gravity down;
• Layer (3): adds low-frequency push to (1), Low-End specialized;
• The bands covered by (1)→(3) get progressively lower.
• Processing order: set the rough balance against the drums → Solo and adjust (1) → bring in (2) and process → finally tune (3) together.
• Guiding principle: "raise the unity of the three layers" — "file down" whatever sticks out beyond the ideal tone and fill in what's lacking.

Always check phase when layering

• Waveform peaks opposed = phase offset; lows and highs get canceled and lost;
• Principle: bass wavelengths are long → heavily affected by phase offset (short high-frequency wavelengths always find partial alignment even when partly inverted);
• Tools:
  • Polarity flip: the Trim plugin bundled with Pro Tools;
  • Or nudge the waveform by a few to a few dozen samples (sometimes this actually adds dimensionality);
  • For MIDI instrument tracks, nudge slightly with Time Adjuster or sample delay.
• Alignment trick: rather than lining up each layer's "note onset," align the other layers' onsets to one layer's peak (the first crest after the note start) — sometimes it's "bouncier";
• With experience you can tell by ear alone "which band is being lost to phase misalignment."

Measured EQ parameters (Pro-Q 3, wide Q)

• Layer (1):
  • 260Hz +1dB (more elastic, meaty);
  • Correspondingly 3.5kHz +1dB or so (tonal balance follows along);
  • 30Hz about +0.7dB (the "glue with the other two layers" cohesion when all three play together).
• Layer (2) (after polarity flip):
  • 38Hz +1dB-plus ("want a bit more underneath" when playing with (1));
  • Cut about 1dB at 11.5Hz (wide-Q Bell, not a Low-Cut);
  • 120Hz +2.5dB (the band where you want the "core");
  • This then clashes with (1) → suppress about 1dB around 430Hz with dynamic EQ.

Spread into the empty Side space (WAVES Center)

• Conventional wisdom says a Stereo Imager on Bass buries it, but look at the whole mix: the Sides often hold only the highs of the top-layer parts, with the low region nearly empty;
• Spreading the Bass moderately into the sides = "slotting into the gap" → presence actually increases, coexisting zone-by-zone with the centered Kick;
• Division of labor: (1) has edge and clear pitch → keep it Center-panned; hand everything below ~430Hz to (2) (the aforementioned 430Hz dynamic EQ serves exactly this);
• After widening, directly Cut the bands that overlap the Center and become "too much" in volume (left unattended, the overall level sounds doubled or tripled);
• Do not create a "sides-only" state (it feels unnatural and loses energy).

Low-end stereo widening / harmonic enhancement always via Send & Return, blended in sparingly (dry signal stays mono, centered)

• SONNOX VoxDoubler's Widen: pseudo-stereo kept "barely perceptible," send level deliberately low — the standard safe method when you want just a touch of width on the lows;
• WAVES Renaissance Bass: emphasizes around 40Hz (psychoacoustic harmonics make small speakers perceive the sub-lows), likewise layered thinly with Intensity kept in check.

Uneven levels between notes

• Tube-Tech CL 1B with slow attack/release tendencies to even them out (taking its "sticky thickness");
• Then Oxford EQ to shape the lows and the mid-high band where pitch identity lives, so the outline is discernible even at low volume.

1.5 Low-End Mixing Checklist

1. Before mixing, form a clear image of the ideal sonic picture;
2. Repeatedly verify in the order "full mix → Solo → full mix";
3. Watch the cohesion between low instruments like Kick/Bass;
4. Decide "what you want people to hear" and cut the unneeded bands accordingly;
5. Make the sense of speed of all low-frequency components consistent.

1.6 Trap Kick/808 Reference Tracks (listening-point cheat sheet)

2. Saturation and Harmonics (key to the rhythm section)

2.1 Principles

• Saturation = the faint distortion "just before obvious distortion": input exceeds the gear's limit point and waveform peaks get flattened (keep pushing and it eventually approaches a square wave).
• There is no clear boundary between it and Distortion; the call is made by sensibility and musicality.
• Distortion inevitably produces harmonics (overtones), and the harmonic type determines the character:
  • 2nd harmonic (2x the fundamental = one octave up): vacuum-tube circuits, tape → warmth;
  • 3rd harmonic (3x the fundamental = a perfect fifth): 70s transistor circuits → "bite," adds warmth to the low mids;
  • Non-integer harmonics: the source of the texture of percussion, husky voices, whisper vocals.
• Even adding only ~1% of harmonics meaningfully affects the tone; a spec sheet's THD can be read as the "harmonic addition rate."
• More harmonics = more information = a more three-dimensional tonal character.
• Tape compression (Tape Comp): peaks flattened → tails relatively lifted → parts sound more continuous → loudness feel, grain, and groove all rise.
• A compressor's character (kyarakutā) is likewise built on saturation.
• Quick character notes on classic units:
  • NEVE 33609: "Glue"-style "quiet" compression — plants the torso firmly, effective without sounding compressed; the fixed Attack of about 3ms is superb on drums; harmonics spread upward with a "matte" texture;
  • UREI 1176LN: Attack 20μs-0.8ms, extremely fast; harmonics "stretch upward," suited to Snare/Bass that need speed; on Kicks in rock and Black music it's actually used less.
• Gain staging matters: passing signal at the proper level yields "subtle" saturation — electrically it's distortion/noise, musically it's a "hidden flavor" that brings persuasiveness and tonal sheen.
• Gear with tubes + discrete circuits + input/output transformers (common in vintage units) easily produces good saturation; modern IC designs chasing clean sound have a hard time producing good saturation.
• Over-EQing is a cause of "thin sound":
  • When a band recording loses its energy, it's usually excessive EQ;
  • Endlessly EQing because "I want a bit more highs" never converges on the right answer;
  • When the musical appeal of the ensemble is shrinking, adjusting the center of gravity with saturation gets you to the finish line faster.

2.2 Five Practical Techniques (Satoshi Fukuda)

Technique 1: Ambience-mic saturation = depth and dimensionality

• Place one mono ambience mic in front of the kit (where Kick/Snare/Hi-Hat balance well) (NEUMANN M49 tube condenser);
• Into a TELETRONIX LA-2A, pushing the GAIN knob up to create saturation;
• Judging the amount: the critical line is that a hard Crash hit must not distort to the point of transient collapse — take loud sections to "just about to overflow";
• Principle: attacks get flattened → tails and reverb components relatively rise → the "contour" of the recording space (wall-reflection information, etc.) is highlighted → the close-mic sound gains depth and dimensionality;
• Difference from Reverb: Reverb "pastes on a separate layer of spatial decay," whereas this "highlights the real room reverb of the recording" — tightness is not lost, so it still suits drums;
• Saturation's characteristic "shattering, scattering" quality is the point (it cannot be produced merely by raising Release with compression);
• Plugins: SOUNDTOYS Decapitator or Devil-Loc (the UAD LA-2A plugin's distortion grain is too smooth and pretty to yield dimensionality; Devil-Loc Deluxe = a model of the SHURE Level-Loc, self-described "Audio Level Destroyer").

Technique 2: Saturated ambience changes the Groove = laid-back feel

• Once the distorted ambience is blended in: notes ring longer and feel gently pulled backward → the perceived groove shifts "behind the beat," groovier;
• Non-integer harmonics increase → hit outlines blur slightly → timing errors stop being conspicuous — even the rhythm's wobble becomes pleasant;
• Practical advice: don't rush to fix timing before mixing — apply saturation first; very often things simply "sit right," then decide whether fixing is still needed;
• The same goes for vocal timing/pitch; Black music especially calls for loose handling — use saturation to preserve the "human feel" of instruments and vocals.

Technique 3: Parallel saturation for punch

• Snare with 2 top + 1 bottom mics (SM57×3): one top mic through an ALTEC 1567A (60s tube 6-in/1-out mixer amplifier, also used by early Motown), slammed hard into the head amp;
• "Unrecognizably mangled" is the requirement — if it still sounds "decent, ringing normally" in solo, the saturation isn't deep enough;
• Blend in parallel with the other, cleanly recorded mic: the sound's "front face (outline)" stays clean while only the "back side" texture is rewritten — grit and punch rise;
• Essentially the parallel-compression concept (compressed signal blended with the original);
• Same method at mix stage: duplicate the Snare track → saturate → blend with the original;
• Plugins: Decapitator, Radiator (= 1567A model; BASS/TREBLE/INPUT/OUTPUT/MIX);
• Additional "grit just by passing through" plugins: UAD Harrison 32C, Neve 1073 Legacy — with the EQ section engaged the mids stand up with a "zaku" bite and turn punchy; better suited to Black-music snares than the newer UAD Neve 1073.

Technique 4: Controlling the center of gravity

• ALTEC 436C as a pure saturator:
  • Threshold at max, Release at fastest (= almost no compression), input level pushed to about 2 o'clock;
  • Adds "zara-zara to bari-bari" (gritty to crackling) distortion; the sound's outline stands up;
  • On low instruments it means "staying low, yet slightly pulled upward" — the center of gravity doesn't sag but gains outline.
• GATES Sta-Level (Vari-Mu tube limiter):
  • The compressed signal stacked back onto the original → the sound "stretches downward," saturation forms in the lows → the center of gravity drops;
  • Merely passing through it lowers the center of gravity;
  • Real example: the Bass on ENDRECHERI "HYBRID FUNK" — Line + Mic summed and fed into the Sta-Level; a "firmly planted" heavy feel, with no EQ inserted at all.
• Clap penetration comes not from boosting high-band EQ but from saturation:
  • Decapitator on an AUX track for Send & Return saturation;
  • Additive thinking — the Send amount decides "how much flies over"; aggressive, like "hurling the sound at the ceiling."
• Kick example (Shunské G & The Peas "Groove Me"):
  • Beater (impact point) / shell resonance / Sub low recordings summed to an AUX track → insert Decapitator;
  • DRIVE 2-3 (about 9-10 o'clock), emphasizing the shell's "boko" component;
  • Use the TONE knob as an EQ = skip difficult EQing and instantly get the desired tone.
• Decapitator = the author's "My Best": universal for Clap/Snare/Hi-Hat/Kick/Tom/Overhead — all rhythm material;
• FABFILTER Saturn: finer, more delicate distortion grain → suited to EDM and other electronic styles and acoustic+electronic hybrid dance music; rich presets.

Technique 5: Turning a "vertical sound" "horizontal" (widening the image)

• Applies when the Kick feels "tall and narrow" and the sides feel lonely (COMA-CHI "water" example):
  1. Send the Kick via Bus to an ALTEC 1567A and GATES Sta-Level, blend the processed signal in parallel with the original — the ALTEC makes the sound "slowly bleed outward"; the vertically extended sound broadens horizontally and thickens;
  2. Send & Return the original Kick again through a UREI 1178 (stereo version of the 1176) — running mono material through stereo outboard can sometimes yield a bit of stereo feel = width;
  3. Also return an extremely short Room Reverb at only −35.2dB — it bleeds, widens horizontally, and gains fullness/presence within the arrangement.
• Fast, dense Bass lines: UAD UA 1176LN Rev E to emphasize the outline (note its extremely fast attack compresses "fast"; when you want the part laid-back, switch to another compressor);
• Material that's "a bit thin": run it through AVID Fairchild 660 (bundled standard with Pro Tools; three knobs INPUT GAIN/THRESHOLD/TIME CONSTANT) — the moment it passes through it "gently" thickens.
• Judgment when mixing for clients: sounds that deserve saturation = those that "want one more push of heaviness" or are "too polished," lacking harmonics and grit — add sparingly, step by step.

Iron rule for the upper limit: saturation isn't just "roughing up the sound" — it serves space, groove, punch, center of gravity, and width; never overdo it — the ceiling is "the attack must remain clearly perceptible."

2.3 Other Favorite Saturation Tools

• UAD Studer A800: the most-used tape sim; switching tape type and IPS (tape speed) gives clear-cut changes; excellent harmonic feel, also used "just to pass signal through."
• U-HE Satin: continuously variable tape speed is the appealing point; just-right grit.
• EMPIRICAL LABS Distressor EL-8: the name itself = Distortion + Compressor; Dist2/Dist3 modes add distortion content; frequently used on Bass and Kick.
• SHADOW HILLS The Equinox: summing amp and monitor controller; the sound "bleeds with quality richness," close to the feeling of working on an SSL console.

3. Sidechain Applications

3.1 Vocal-Triggered M/S Sidechain (core recipe, Atsushi Asada)

• Goal: the backing automatically steps aside the instant the vocal appears — the vocal is clear without pushing its level up.
• Added value: the linked "when A appears, B yields" motion makes the two perceptually bond into one unit → vocal and backing gain cohesion, keeping loudness while securing intelligibility.
• Compress only the Mid content of the top-layer instruments; leave the Side untouched:
  • The vocal sits in the center of the field (Mid), and what clashes with it is mainly the backing's Mid content;
  • Preserving the Side → stereo width is unharmed and the processing is better concealed.
• Plugin: BRAINWORX BX_DynEQ V2 (a dynamic EQ supporting M/S mode + external sidechain — better than an ordinary compressor at limiting "which band, which channel component" yields).
• Parameter recipe:
  • Processing target: the Mid of the top-layer parts (M/S mode, Mid only);
  • Attenuated band: around 350Hz (the low mids that clash with vocal body; the main battlefield of vocal thickness/presence is roughly 150-500Hz);
  • Key signal: SC band set to ext (external input = the vocal track);
  • EQ the key signal itself to 681Hz — triggering off the vocal's energy around 681Hz tracks the moments of "actually singing" more accurately; confirm the key content with the SC solo button;
  • Attack 0.06ms (extremely fast), Release 12ms (very fast), Threshold −47.7dB;
  • Attenuation about 2.2dB — the empirical sweet spot is 1.5~3dB for the most natural result: "you can't hear the processing, yet the vocal is somehow clear"; too much gives it away.
• Operating order: SC solo to confirm the key → lower the threshold until you see the desired triggering → set max gain around 2dB → play the whole song and confirm the groove isn't ruined by a "sucking" sensation.
• Comparison-experiment conclusions:
  • Compressing the whole top layer (Mid+Side): the backing "loses its axis" and the vocal paradoxically floats without blending (worse as attenuation approaches the 5dB range) → preserving the Side and moving only the Mid is clearly superior;
  • Vocal-triggered compression of Bass (Mid): the bands barely overlap so the effect is weak, but "the lows dipping slightly the instant the vocal appears" can create a breathing groove.
• Synth-Lead-triggered version (EDM Drop, "make the main melody definitely land"):
  • Faster Attack, slightly fast Release, attenuation by ear around 4dB;
  • M/S compressing only the Mid: Side intact, the field doesn't collapse (the author's preference);
  • Ordinary stereo compression: the melody juts out more, more momentum — both are valid; choose by the track's character.
• Recommended sidechain plugins:
  • FABFILTER Pro-C2: first in usability — Lookahead, 8 compression styles, M/S support, SIDE CHAIN LEVEL plus sidechain EQ/Audition;
  • DMG AUDIO Essence: an M/S-capable De-esser (mastering grade) = essentially "a compressor whose sidechain EQ is pre-configured," working band freely selectable — for the detail-oriented.

3.2 Classic Kick-Triggered Pumping (Nagie)

• Feed a four-on-the-floor Kick into a compressor's sidechain and compress "everything except the Kick":
  • At the eighth-note point after the Kick's decay ends, the bass surges out with a "whoosh" = the Pump;
  • Kick decay and bass surge in inverse proportion, with zero gaps in the sound — ultimate loudness music (Daft Punk / Justice-style Electro).
• Not just bass: piano, synths, vocals can all be hooked up — sounds vanish and surge back in unexpected places, generating interesting phrases.
• The sidechain input needn't be the Kick: feed a pattern different from the Kick → more complex pumping effects.
• Routing: bass group / synth group / vocal group each get an AUX with a compressor; the Kick is sent to every compressor's sidechain input → all groups sum into the master fader with a bus compressor pushing loudness to the max.
• The decisive parameters = the compressor's Attack Time and Release Time (getting the imagined compression curve rarely happens on the first try); with multi-stage compression watch out for DAW latency.
• At this level, balance is no longer controlled by mix faders but by "sound combinations + how the sidechain acts" — that determines the music's entire balance.
• You can stack a Transformer-style effect on top (AVID Transfuser etc., tempo-synced "chopping" of the audio) for a mix of pumping and stuttering.

3.3 Lightweight / Derivative Sidechain Uses

• Hi-Hat sidechained with the Kick as trigger (mabanua): the Hi-Hat only "peeks out" in the Kick's gaps, adding groove to just the beat; try different depths.
• Easy tool: NICKY ROMERO Kickstart (Mix 100%);
  • Trap layering trick: time-stretch a Cymbal about 4x longer (without changing pitch) → apply Kickstart → the long tone "breathes" on the downbeats, padding behind the Hi-Hat for dimensionality.
• Sidechain origins: broadcast ducking (announcer's voice automatically lowers the BGM) and the De-esser (sidechain compression with an EQ'd key signal) — understand this and you can freely mutate the technique.
• The Key-In triggering of Gated Reverb is also a sidechain application (see 5.1).

4. Vocal Mixing Chain and 7 Vocal Effects

4.1 Mixing Order That Makes Vocals Stand Out (Nagie)

1. Drums and bass first:
   • Once the rhythm section is set, bring the vocal in to test;
   • When the bass collides with the vocal: EQ-cut the bass around 1kHz (the zone that easily obstructs vocals) while keeping the bass tone natural;
   • If bass volume swings are big enough to hit the vocal: flatten the level with a fast-Attack compressor.
2. Yield the center to the vocal:
   • Guitar panned left, piano panned right;
   • Harmonies (Cho) widened left/right with chorus or short delay (creating an out-of-phase "valley" in the center);
   • Bass sits at the bottom;
   • EQ the instruments whose bands clash with the vocal — carve out a "seat" for the Vo in the 1kHz vocal zone;
   • Goal at this stage: a backing track that "sounds cool even without the vocal."
3. After placing the vocal back, set the basic position first (heavy-rhythm tracks slightly bury the vocal; a flagship single opens it fully), then two cleanup methods:
   • Method 1 (rework the offender's sound): EQ away from the vocal band / tame runaway peaks with compression / change panning away from the song / deepen its reverb to push it behind the vocal;
   • Method 2 (volume control via automation): listen carefully for the moments something blocks the song and duck that instrument momentarily;
   • Never do the reverse — push the vocal up: raising the vocal → other instruments then feel lacking → everything gets raised → the master level instantly clips; a death spiral;
   • Until the backing is sorted, resist touching the vocal fader.
4. Finishing:
   • Automate expressive vocal-level detail (fine dynamics);
   • Multiband compressor on the master fader to shape loudness, final fine-tuning;
   • Advanced: group the whole backing to an AUX with a compressor, rebalance "backing group vs vocal," then a compressor on the master fader → stronger vocal presence;
   • But beware: cohesion between song and backing can thin out; with many compression stages, automation may stop behaving — weigh and experiment.

• Respond to each singer's voice character (which band its features live in, whether the voice cuts through) — preset-style moves like "cut 3dB at 4kHz" don't work.
• 60s retro vocal compression: Attack 4ms / Release 2s, about 5dB GR — the short gap before the Attack means initial breaths and sibilants escape compression, which is exactly the "flavor."

4.2 Seven Vocal Effect Recipes

(1) Vocoder / robot voice (MATZ; reference: Daft Punk "Harder, Better, Faster, Stronger")

• Principle: voice = modulator, synth = carrier; pitch comes not from the voice but is "played" via MIDI on the carrier side.
• ABLETON Live's built-in Vocoder on the vocal track; carrier = a fat LENNARDIGITAL Sylenth1 sawtooth;
• Program the original melody one octave down as MIDI for the carrier;
• Vocoder settings: Carrier set to External, Audio From set to the Synth track, taking the Post FX signal point;
• Larger Bands (band-split count) = higher intelligibility; smaller = rougher and more retro (example: Bands 8, Range 80Hz-12.0kHz); the band bars can be hand-drawn up/down like an EQ;
• Daft Punk-style "human touch" trick: switch Carrier to Noise + Unvoiced Sens. about 55% — unpitched content like consonants/breath gets filled with noise for a "whispery feel"; the robot voice ends up sounding more human;
• Write pitch-bend automation into the carrier MIDI: slide the pitch at phrase starts or ends, tuning "toward sounding more human";
• Redux for old-sampler texture: Bits=8, Rate=40.0kHz (Jitter/Shape 0%);
• Alternatives: SOFTUBE OTO Biscuit, D16 Decimort 2, WAVES OVox (+Morphoder), TAL-Vocoder (hardware-like texture).

(2) Auto-Tune "kero-kero" voice (ケロケロ・ボイス, hard-tuned croaky vocal) (TeddyLoid; reference: Perfume "ポリリズム")

• ANTARES Auto-Tune Pro switched to CLASSIC mode — "the Auto-Tune 5-era sound"; this is what sparked the original boom, and pros prefer this mode for kero-kero;
• Auto Mode: set KEY/SCALE to the song's key, choose INPUT TYPE (Soprano/Alto/Tenor etc.) per the singer's range;
• RETUNE SPEED turned fully left = 0 (fastest) — pitch snaps instantly and rigidly to the nearest scale degree; the source of the "frog" effect;
• Simultaneously HUMANIZE=0, NATURAL VIBRATO=0, completely removing humanization for maximum mechanical feel;
• FORMANT on; adjust timbre with THROAT (throat-length modeling, example value 93): formants shifted up = cute and slight, down = thick and full — changes the voice color without changing pitch.

(3) Digital Choir (T-SK; reference: Kanye West "Lost in the World feat. Bon Iver")

• Use: chorus pads or Drops — vocals stacked into a "digital choir," sacred and coldly gleaming;
• iZOTOPE VocalSynth 2 in MIDI Mode (not Auto Mode); create a MIDI track and route its MIDI output to the plugin;
• Enter chords on the MIDI track (write whatever harmony you want the choir to sing);
• Note: no sound where there are no MIDI notes — "from where to where it sounds" is precisely controlled by note lengths;
• Start from a thick-sounding MIDI Mode Preset (categories like Classic/Clear/Deep/Edgy/Lush), then fine-tune as needed;
• Finish with Pro-Q 3: low-cut to clean the bottom + attenuate around 774Hz where it clashes with the lead vocal to make way; optionally add a stereo imager (widening) and a De-esser (de-harshing).

(4) Talkbox (JUVENILE; reference: Bruno Mars "24K Magic")

• First treat the dry vocal (a talkbox must "speak clearly" as a prerequisite): low-cut below 200Hz, control the consonant/sibilance zone above 3kHz to keep consonants clear, compress to gather dynamics;
• VocalSynth 2 with the Talkbox module enabled + MIDI Mode;
• MIDI = unison with the main melody + one layer an octave up — the upper octave layer adds cut-through and thickness;
• Drive (distortion amount → funkiness) and Speaker (speaker resonance character) are the two most important knobs; always adjust them; the module also has OSC waveform/Sub/LFO;
• Afterwards add EQ, reverb, chorus; consonants get buried easily — keep them clear.

(5) Megaphone voice (yasu; reference: Ado "阿修羅ちゃん")

• MCDSP FutzBox (lo-fi multi-effect), starting from the built-in preset "Red Megaphone" but not taking it as-is:
  • EQ: low-cut below about 600Hz, high-cut above about 8kHz, boost about 6dB around 2-3kHz (the megaphone's narrow-band midrange honk);
  • FILTERS section: high-pass about 400Hz, low-pass 14kHz lightly (don't over-cut the highs);
  • DISTORTION section: gradually raise AMOUNT to "just about to break" — past that it's only noise;
  • In the right-hand EQ, boost 2.8kHz another 3-4dB and the megaphone flavor snaps into focus.
• Finish: iZOTOPE Ozone's Exciter, mode "Tape", Amount about 8.4 — too much becomes harsh.

(6) Formant Changer (MK; reference: Porter Robinson & Madeon "Shelter")

• Principle: formants = spectral peaks determined by the vocal-tract shape; shift only the formants, not the pitch → male/female/child/elderly voices — the source of an "uncanny hook";
• Method: create a formant shift amount automation lane under the vocal track and draw it up/down at the right moments — sustained offsets or intermittent "twitches," creating a dislocated feel of male/female or near/far flickering;
• Plugins: WAVES Vocal Bender, ZYNAPTIQ Morph 2, SOUNDTOYS Little AlterBoy, FL Studio's built-in feature — any plugin with a formant parameter works;
• Advanced: layer a formant-processed duplicate behind the lead, or stack chorus/effects on top of it.

(7) Telephone / radio voice (murozo; reference: YOASOBI "たぶん")

• Positioning: a great etude for EQ fundamentals;
• Pro-Q 3 band-pass: low-cut about 300Hz + high-cut about 4kHz as the basic shape;
• The steeper the filter slope (6/12/…/72dB/oct up to Brickwall), the stronger the lo-fi feel;
• Don't memorize frequencies: move the cutoffs while the song plays and find the spot where "the vocal still opens up and stays intelligible" (the optimum differs per singer and between male/female voices);
• Bonus: light saturation/distortion makes it more convincing; layering noise pushes it from "telephone" toward "radio";
• The key to usage = contrast with the dry lead vocal: use the narrow-band voice at scene changes like intro/interlude/bridge, then return to the full-range vocal at the verse — the contrast makes the lead more dimensional.

Bonus: a 60s trick — MOOG-filtered vocal (The Doors "Strange Days")

• Peaking EQ, narrow Q, boost 30-40dB at 2.5kHz (if one EQ can't boost that much, chain 2; real example: 2538Hz, Gain +18, Q 6.50);
• → low-pass filter cutting everything above 2.5kHz;
• → Delay 167ms (imitating the tape-machine delay time at 15ips tape speed).

5. Classic Techniques Toolbox

5.1 Gated Reverb (the signature 80s sound)

• Origin anecdote: during the recording of "Peter Gabriel III (Melt)" (1980), the drum room's talkback mic (with compression + Gate on it) accidentally picked up the drums — the over-compressed/gated room sound was unexpectedly cool (Steve Lillywhite × Hugh Padgham).
• Basic flow: close-miked snare → tightened with a Gate → the gaps between hits filled with digital reverb → that reverb tail cut off again with a Gate.
• Method (1): Key-In sidechain triggering
  • Send the snare track to a reverb track (UAD Lexicon 224 or another digital reverb; many recent plugins recreate vintage digital reverbs);
  • Insert a Gate after the reverb — but a Gate alone won't cut the tail crisply;
  • Key point: feed the dry snare into the Gate's Key-In (sidechain input) to trigger it open and shut;
  • WAVES C1 example parameters: sidechain source Bus 10, Gate Open=−25.8, Gate Close=−31.7, Attack=0.15ms, Release=63ms, Hold=257.63ms;
  • The reverb length is set by Hold — the right length varies with tempo; turn Hold while the song plays and stop where it sounds cool.
• Method (2): compressed ambience mic as the reverb
  • Put a fast-Attack / slow-Release compressor on the ambience mic to lift the air and use it as reverb (a major element of the original Gated Reverb);
  • WAVES CLA-76 (1176-style): Attack fastest (fully right) / Release slowest (fully left) — note that both its ATTACK/RELEASE knobs get faster the further right they're turned.
• Vintage-reverb-flavor tricks:
  • Lower the Density (reverb density) a little = slightly grainy sound, more 80s atmosphere;
  • For "reverb applied yet the sound stays up front": turn off the Reverb Tail and add only the Early Reflections.
• Recording note: close-mic the cymbals and handle them separately from the snare — otherwise gating the close snare track can't remove the tails bleeding into the Overheads.

5.2 Parallel Compression

• Principle: compressor on a Bus, Send & Return blending only the needed amount of the squashed signal back with the original — deep compression without losing the sound's "core."
• Single-unit option: a compressor with a Mix/Blend knob, e.g. EMPIRICAL LABS Arousor (example: IN=6.25dB, Att=31.3ms, Rel=625ms, OUT=11.75dB, BLEND 3.13; Soft Clipping steps: Pinned 14%THD / Toasty 8% / Hot 3% / Warm 0.5%).
• Application (1), compression as reverb (Tchad Blake style): deeply compress off-mic material → attacks flattened → room tails (reverberant content) float up = an exaggerated space with a flavor different from digital reverb.
• Application (2), parallel saturation: original + "unrecognizably mangled" saturated signal blended = rewriting only the back-side texture of the sound (see 2.2 Technique 3).

5.3 Advanced Compression

• Two in series (the Roger Nichols principle):
  • For compression to sound natural, GR on a single source should not exceed 3dB;
  • Need more = 2 compressors doing 3dB each — far more natural than 6dB from one;
  • Division of labor: unit 1 with slow Attack for a natural compressed foundation; unit 2 dedicated to catching the attack transients the first one missed;
  • More "musical" peak control than reaching straight for a Maximizer;
  • Example: WAVES CLA-2A (optical) + CLA-76 (FET) in series.
• "Foul-play" usage (Tchad Blake style):
  • At 15-20dB GR the tone is no longer flat: mids bulge, highs roll off, so the lows sound lifted;
  • Extremely short Release produces distortion;
  • Deliberately breaking the rules draws out a compressor's "other face" (Tchad used the SPECTRA SONICS 610, SHURE Level-Loc);
  • Choose plugins that "change the sound the moment they're inserted": SOUNDTOYS Devil-Loc, WAVES Abbey Road RS124 (most plugins compress with a flat tone and don't suit this).
• Taming slap bass: Ratio 8:1 or higher, shortest Attack (0.01ms), Release about 50ms (WAVES C1 example: 7.95:1, Threshold −15.0, Makeup 5.0); keep the settings and deepen GR until the phrase is clear — the Reduction meter swings 5-10dB on the "smack" moments.
• Vintage slow-At

…(truncated)