Offered by Dead Brands
Universal, tool-agnostic audio engineering: gain staging, EQ, compression, reverb and delay, the stereo field, reference-track workflow, and loudness targets.
Use it over MCP: connect to https://mcp.deadbrands.co/hub/mcp and call skills_get with slug mixing-and-production-basics.
# Mixing and Production Basics Credit: SteTy — published free in the Skills Hub Universal, tool-agnostic audio engineering: gain staging, EQ, compression, reverb and delay, the stereo field, reference-track workflow, and loudness targets. Every number here is a professional starting point that works in any DAW, on any console, for any genre — the goal is mixes that translate from studio monitors to earbuds to a car. ## Overview Mixing is balancing: setting levels, carving frequency space so sources stop masking each other, controlling dynamics so performances feel steady, and placing elements in a stereo image with a sense of depth. The order of operations matters more than any plugin: a well-gain-staged, well-balanced rough mix with faders alone will beat a plugin pile on a chaotic session. Mastering — final loudness and tonal polish of the stereo file — is a separate, later step; this skill covers mixing thoroughly and the loudness targets your finished master must hit. ## Core Principles - **Gain staging first.** Analog-modeled processors are voiced for signals around **−18 dBFS RMS (≈ 0 VU)**. Stage individual tracks to peak roughly −18 to −12 dBFS, and keep the finished mix peaking around **−6 to −3 dBFS** with no master-bus processing — headroom is what lets every later stage sound clean. Too quiet is also a fault: you raise the noise floor when you push it back up. - **Cut before you boost.** Most EQ problems are excess, not absence. Muddy mix? Cut 250–400 Hz rather than boosting highs. Harsh vocal? Find the spike between 2–6 kHz and cut it. Cuts remove the cause; boosts mostly add loudness and fatigue. - **Louder always "sounds better" — lie detector required.** Human hearing (the Fletcher–Munson / equal-loudness contours) perceives louder as fuller and brighter. Every A/B comparison — plugins, references, your own master — must be level-matched, or you are comparing volume, not quality. - **Compression is a budget.** Gain reduction spent on the track, again on its bus, again on the mix bus is cumulative even though each meter looks innocent. Typical budgets: track 3–6 dB, subgroup 2–4 dB, mix bus 1–2 dB. - **Mono is the truth serum.** If elements vanish or the low end gets hollow in mono, you have phase/width problems no stereo trick will fix. Check mono at every stage. ## Techniques **EQ — the frequency map.** Human hearing runs 20 Hz–20 kHz: - **20–60 Hz sub-bass:** felt, not heard; 808 and kick fundamentals. Inaudible rumble here eats headroom — high-pass everything that isn't kick or bass. - **60–250 Hz bass:** kick punch (fundamental ≈ 60–100 Hz), bass body. Decide which one owns the very bottom and cut the other there. - **250–500 Hz low-mids:** warmth — and the **mud zone** (200–400 Hz). Nearly every instrument has energy here; small wide cuts (2–4 dB) on non-bass sources are the fastest professional-sounding move in mixing. - **500 Hz–2 kHz mids:** body and vocal intelligibility; excess sounds honky/nasal. - **2–6 kHz presence:** attack, consonants, beater click (kick ≈ 2–5 kHz), snare crack. Also the harshness/fatigue zone — cut surgically. - **6–20 kHz air/brilliance:** cymbal shimmer, vocal "air" (≈ 10–16 kHz gentle shelf); excess = hiss and sibilance (vocals' sibilance lives ≈ 5–8 kHz — treat with a de-esser, not a static cut). Working method: the **sweep-and-cut** — narrow band, +10 dB boost, sweep slowly; the ugliest resonance is the problem; flip to a cut and widen. Per-source landmarks: kick boxiness ≈ 300–500 Hz; snare body ≈ 150–250 Hz, snap ≈ 5–10 kHz; bass growl ≈ 700 Hz–1.5 kHz; vocal clarity ≈ 2–5 kHz, warmth ≈ 150–300 Hz; guitars bite ≈ 2–4 kHz, mud ≈ 300–500 Hz. High-pass starting points: vocals 80–100 Hz, guitars 80–120 Hz, overheads 100+ Hz. **Compression — the five controls.** Threshold (where it starts working), ratio (how hard), attack (how fast it clamps — slow attack lets transients through = punch; fast attack = control but flattens), release (how fast it lets go — watch the gain-reduction meter return toward zero between hits; too fast = pumping, too slow = choking), and makeup gain (restore level, then A/B at matched loudness). Starting points: - **Lead vocal:** 3–4:1, attack 10–25 ms, release 60–150 ms, 3–6 dB GR on peaks. Two gentle compressors in series beat one slamming. - **Kick:** 4:1, attack 5–10 ms, release 50–100 ms, 3–6 dB. **Snare:** 4:1, attack 15–30 ms for punch, release 80–150 ms. - **Bass:** 4:1, attack 10–30 ms, release 100–250 ms (keep release above ~100 ms — faster lets the compressor track individual bass wave cycles and distort), 3–7 dB. - **Drum bus:** 2–4:1, attack 10–30 ms, 2–4 dB. **Mix bus:** 1.5–2:1, attack ~30 ms, release auto or 100–200 ms, **1–2 dB maximum** — glue, not demolition. - **Parallel compression:** blend a heavily compressed copy under the uncompressed signal — density with transients intact; the classic drum and vocal trick. **Reverb and delay — manufactured space.** Reverb types have jobs: **plate** = bright, dense, the pop vocal standard; **hall** = large and lush for ballads and orchestral sources; **room/chamber** = short, believable realism; **spring** = vintage splash (guitar). Set **pre-delay** (20–100 ms) so the dry attack stays intelligible before the wash arrives; set decay so the tail ends before the next phrase — tempo-synced starting points: at 120 BPM a quarter note = 500 ms, eighth = 250 ms. Delay: **slapback** 75–150 ms (rockabilly, vocal thickening), tempo-synced dotted-eighth (the U2/The Edge signature) for rhythmic space, ping-pong for width. Send effects from a shared bus rather than inserting per track — one coherent space glues a mix; a different reverb on every track builds a swamp. Keep low end out of reverbs (high-pass the return ≈ 150–250 Hz). **The stereo field.** Pan by function: kick, snare, bass, lead vocal stay center (the mono-compatible core); supporting elements spread. Use **mid/side** thinking: keep subs mono, let width live in the mids and highs. The Haas effect (≈ 5–35 ms delay to one side) widens but risks mono phase cancellation — verify in mono. Check the mix on one small speaker and earbuds: most listeners hear your mix there first. **Reference tracks.** Pick 2–3 professionally mixed songs in your genre. Level-match them to your mix (they're mastered louder — turn them *down* to match). A/B in short loops comparing: low-end weight, vocal level relative to the track, presence-region brightness, and how loud the chorus feels versus the verse. Mix toward the reference's *balances*, not its loudness. **Loudness targets (finished masters).** Streaming normalization has ended the loudness war — louder masters simply get turned down. Spotify, YouTube, Tidal, Amazon: **−14 LUFS integrated**; Apple Music: **−16 LUFS**; true-peak ceiling **−1 dBTP** (leave −1 to −2 dBTP for lossy encoding). A dynamic master at −14 LUFS with intact transients will sound punchier on-platform than a crushed −8 LUFS file turned down by the same service. ## Worked Examples **The muddy mix rescue.** A dense rock mix sounds "behind glass." Diagnosis path: solo-safe nothing; pull up the spectrum — low-mid buildup. Fix: high-pass guitars at 100 Hz, cut 300 Hz by 3 dB (wide bell) on guitars, piano, and backing vocals; leave bass and kick untouched; re-check. Clarity returns without any boosts, and the master bus peak drops 1.5 dB — free headroom. **Setting a vocal compressor in 60 seconds.** Ratio 3:1, attack 15 ms, release 100 ms. Lower threshold until peaks show 4 dB GR. Makeup gain to match bypass level. A/B: if consonants vanished, slow the attack; if the level "breathes" audibly, slow the release. Then add a second compressor at 2:1 catching only the loudest 2 dB — broadcast-stable vocal. **Tempo-synced space.** Song at 100 BPM: quarter = 600 ms, dotted eighth = 450 ms. Vocal delay: dotted-eighth ping-pong, feedback ~25%, mixed low. Reverb: plate, pre-delay 40 ms, decay ~1.8 s, return high-passed at 200 Hz. The vocal now sits in a space that pulses with the song instead of washing over it. ## Common Pitfalls - Mixing in solo: EQ decisions made on isolated tracks routinely destroy the blend. Solo to diagnose, decide in context. - Master-bus compression (or a limiter) left on while mixing — you're mixing into a moving target. - Boosting instead of cutting; double-digit dB moves as a first resort. Professional EQ is mostly 1–4 dB. - Reverb/delay inserted per-track with no shared space, and unfiltered reverb returns muddying the low end. - Comparing your unmastered mix to a loud commercial master without level-matching, then "fixing" the difference with a limiter. - Ignoring mono and small-speaker checks until release day. ## Practice Drills 1. **Faders-only mix:** balance a multitrack with volume and pan only, 30 minutes. Live with it a day before allowing yourself one EQ. 2. **Sweep hunt:** on five different sources, find the ugliest resonance with a boosted narrow sweep from memory of the map, then cut it. 3. **Attack/release lab:** one drum loop, one compressor at 10 dB GR; sweep attack slowest→fastest, then release; describe each change in one sentence. Now you can hear the controls forever. 4. **Reference ladder:** level-match one reference and A/B your mix in 10-second loops, writing down exactly three balance differences — fix only those. 5. **One-space mix:** mix a song using a single reverb and a single delay on sends, no inserts. Learn what coherence sounds like. ## When to Use and Limits - **Use it when:** a task calls for mixing and production basics, as laid out in the Techniques and Worked Examples above. - **Watch first for:** (see Common Pitfalls). - **Limits:** this is a general reference drawn from public sources; it is a starting point for practice, not a full course, and attributions and dates should be checked against the cited sources before quoting them. ## Quick Reference - Gain staging: tracks ≈ −18 dBFS RMS / peaks −18…−12 · mix peaks −6…−3 dBFS, bus clean - Mud zone 200–400 Hz · presence 2–6 kHz · sibilance 5–8 kHz · air 10–16 kHz · high-pass non-bass sources - Compression: vocal 3–4:1 / 10–25 ms / 60–150 ms · bass release > 100 ms · mix bus 1.5–2:1, ≤ 2 dB GR - Space: pre-delay 20–100 ms · slapback 75–150 ms · dotted-eighth = ¾ of a beat · filter reverb returns - Loudness: −14 LUFS (Spotify/YouTube/Tidal) · −16 LUFS (Apple) · true peak ≤ −1 dBTP - Golden rules: cut before boost · level-match every A/B · check mono · compression is a budget --- *Trademarks and copyrights mentioned belong to their respective owners. Educational reference only; not affiliated with or endorsed by them.*