@mthines/reaper-mcp 0.1.0

package/knowledge/reference/frequencies.md

@@ -0,0 +1,154 @@
+# EQ Frequency Reference
+
+A cheat sheet for the agent: what lives where, what to cut, what to boost, and why. All values are starting points — trust measurements over tables.
+
+## Frequency Band Definitions
+
+| Band | Range | Sensory Description | Common Issues |
+|------|-------|-------------------|---------------|
+| Ultra-sub | 20–40 Hz | Felt in chest; inaudible on most speakers | Rumble, mic handling noise — usually remove |
+| Sub | 40–60 Hz | Bass and kick fundamental; club system territory | Too much = boomy; too little = thin |
+| Bass | 60–120 Hz | Warmth, punch, weight | Mud if over-accumulated |
+| Upper bass | 120–250 Hz | Body of bass instruments; low-end of most other instruments | The "mud zone begins here" |
+| Low-mids | 250–500 Hz | MUD ZONE — where amateur mixes go wrong | Cut aggressively on non-bass instruments |
+| Mids | 500 Hz–1 kHz | Honk, nasality, the "telephone frequency" | Source of unpleasant boxiness |
+| Upper-mids | 1–2 kHz | Presence, forward bite | Nasal buildup on male vocals, guitars |
+| Presence | 2–5 kHz | The MOST SENSITIVE zone to human hearing | Harshness lives here — treat with care |
+| Upper presence | 5–8 kHz | Sibilance, consonant clarity, definition | Over-boosting causes fatigue |
+| Brilliance | 8–12 kHz | Crispness, snap, transient sparkle | Can sound harsh if overdone |
+| Air | 12–20 kHz | Shimmer, openness, studio sound | Subtle — ±1–2 dB is audible here |
+
+## High-Pass Filter Reference (Where to HPF each instrument)
+
+| Instrument | Recommended HPF | Slope | Notes |
+|------------|----------------|-------|-------|
+| Bass guitar | 40–60 Hz | 12–24 dB/oct | Preserve sub-60 Hz warmth |
+| Kick drum | 30–50 Hz | 12 dB/oct | Let sub punch through |
+| Snare drum | 100–150 Hz | 12–24 dB/oct | |
+| Hi-hat | 300–600 Hz | 24 dB/oct | No low-end needed |
+| Overhead mics | 200–400 Hz | 12–24 dB/oct | |
+| Room mics | 80–200 Hz | 12 dB/oct | Depends on room character |
+| Toms | 60–100 Hz | 12 dB/oct | Preserve tom weight |
+| Lead vocals | 80–100 Hz | 24 dB/oct | Below chest voice fundamental |
+| Backing vocals | 100–150 Hz | 24 dB/oct | Tighter than lead |
+| Electric guitar (rhythm) | 80–120 Hz | 24 dB/oct | Higher in metal (100–150 Hz) |
+| Electric guitar (lead) | 80–100 Hz | 18–24 dB/oct | |
+| Acoustic guitar | 80–120 Hz | 24 dB/oct | |
+| Piano | 40–60 Hz | 12 dB/oct | Preserve bass register |
+| Organ | 40–80 Hz | 12 dB/oct | |
+| Strings (full section) | 40–60 Hz | 12 dB/oct | |
+| Synth pads | 80–150 Hz | 18–24 dB/oct | Depends on register |
+| Synth lead | 100–200 Hz | 24 dB/oct | |
+| 808 bass | 25–40 Hz | 12 dB/oct | Ultra-sub only |
+| Reverb return | 100–200 Hz | 24 dB/oct | Prevent verb from muddying bass |
+| Delay return | 100–200 Hz | 24 dB/oct | |
+
+## Problem Frequencies by Symptom
+
+| Symptom | Frequency Zone | Action |
+|---------|---------------|--------|
+| Muddy, undefined | 250–500 Hz | Cut 2–4 dB on ALL competing instruments |
+| Boomy bass | 80–120 Hz | Cut on bass/kick; not everything else |
+| Boxy, phone-speaker sound | 400–800 Hz | Cut on the culprit instrument |
+| Nasal, honky | 800–1200 Hz | Cut on guitars, horns, vocals |
+| Harsh on extended listening | 2000–4000 Hz | Gentle cut on master or culprit |
+| Sibilant S sounds | 6000–10000 Hz | De-ess; dynamic EQ cut |
+| Lack of definition/clarity | 2000–5000 Hz | Gentle boost on lead element |
+| Lacks air, sounds closed | 10000–16000 Hz | Gentle high shelf boost |
+| Thin, lacks body | 100–250 Hz | Boost on instruments that should have weight |
+| Rumble, noise floor | Below 80 Hz | HPF everything except bass instruments |
+| No kick punch | 60–80 Hz | Boost on kick EQ |
+| Kick disappears in mix | 3000–5000 Hz | Boost beater click on kick |
+| Bass disappears on small speakers | 150–300 Hz | Boost harmonics on bass; add saturation |
+| Too bright, fatiguing | 8000–12000 Hz | Gentle cut or shelf |
+| Lacks excitement | 5000–8000 Hz | Slight boost (presence zone) |
+
+## Instrument Fundamental Frequencies
+
+Knowing the fundamental prevents accidentally HPF-ing the note itself.
+
+| Instrument | Lowest Note | Frequency (Hz) |
+|------------|-------------|---------------|
+| Bass guitar (standard) | E1 | 41 Hz |
+| Bass guitar (drop D) | D1 | 37 Hz |
+| Electric guitar | E2 | 82 Hz |
+| Acoustic guitar | E2 | 82 Hz |
+| Piano | A0 | 27 Hz |
+| Male voice (bass) | ~80 Hz | — |
+| Male voice (baritone/tenor) | ~100–150 Hz | — |
+| Female voice (soprano) | ~200 Hz | — |
+| Kick drum (typical) | 60–100 Hz | varies by tuning |
+| Snare drum | 200–300 Hz | fundamental varies |
+| Bass clarinet | B1 | 58 Hz |
+| Cello | C2 | 65 Hz |
+| Double bass | E1 | 41 Hz |
+| Tuba | B1 | 58 Hz |
+
+## EQ Moves by Instrument (Quick Reference)
+
+### Kick drum
+
+| Move | Frequency | Amount | Why |
+|------|-----------|--------|-----|
+| HPF | 30–40 Hz | 24 dB/oct | Remove ultra-sub rumble |
+| Boost sub punch | 60–80 Hz | +2 to +4 dB Bell | The kick "thump" |
+| Cut boxiness | 300–500 Hz | -2 to -4 dB Bell | Reduce cardboard sound |
+| Boost beater click | 3000–5000 Hz | +2 to +4 dB Bell | Punch through mix |
+
+### Snare drum
+
+| Move | Frequency | Amount | Why |
+|------|-----------|--------|-----|
+| HPF | 100–150 Hz | 12–24 dB/oct | Remove unnecessary low-end |
+| Boost crack | 200–250 Hz | +2 to +3 dB Bell | The "crack" frequency |
+| Cut ring (if needed) | 400–600 Hz | -2 to -4 dB narrow Bell | Resonant ring |
+| Boost presence | 5000–8000 Hz | +1 to +3 dB Bell | Snap and presence |
+
+### Bass guitar
+
+| Move | Frequency | Amount | Why |
+|------|-----------|--------|-----|
+| HPF | 40–60 Hz | 12 dB/oct | Preserve fundamental |
+| Boost fundamental | 80–120 Hz | +1 to +3 dB Bell | Warmth and body |
+| Cut mud | 250–400 Hz | -2 to -3 dB Bell | Clarity |
+| Boost attack | 700–900 Hz | +1 to +2 dB Bell | Pick/pluck definition |
+| Boost small-speaker translation | 150–300 Hz | +1 to +2 dB Bell | Heard on earbuds |
+
+### Electric guitar (rhythm, high gain)
+
+| Move | Frequency | Amount | Why |
+|------|-----------|--------|-----|
+| HPF | 80–150 Hz | 24 dB/oct | Prevent bass-zone mud |
+| Cut mud | 250–500 Hz | -4 to -8 dB Bell | The metal scoop |
+| Boost presence | 2000–4000 Hz | +1 to +2 dB Bell | Cut through mix |
+| Cut harshness | 3000–5000 Hz | -1 to -2 dB Bell (dynamic) | Control if harsh |
+
+### Acoustic guitar
+
+| Move | Frequency | Amount | Why |
+|------|-----------|--------|-----|
+| HPF | 80–120 Hz | 24 dB/oct | Rumble and body resonance |
+| Cut mud | 200–400 Hz | -2 to -3 dB Bell | Body mud |
+| Boost body | 120–200 Hz | +1 to +2 dB Bell | Warmth if too thin |
+| Boost string detail | 5000–8000 Hz | +1 to +2 dB Bell | Pick attack and string noise |
+
+### Lead vocals (female)
+
+| Move | Frequency | Amount | Why |
+|------|-----------|--------|-----|
+| HPF | 80–100 Hz | 24 dB/oct | Mic rumble |
+| Cut chest resonance | 200–400 Hz | -1 to -3 dB Bell | Proximity effect |
+| Boost presence | 2000–4000 Hz | +1 to +2 dB Bell | Vocal clarity |
+| Dynamic cut sibilance | 7000–9000 Hz | -3 to -5 dB (dynamic) | Harsh S sounds |
+| Boost air | 10000–15000 Hz | +1 to +3 dB Shelf | Breathy, airy quality |
+
+### Lead vocals (male)
+
+| Move | Frequency | Amount | Why |
+|------|-----------|--------|-----|
+| HPF | 80–100 Hz | 24 dB/oct | Rumble |
+| Cut mud | 250–400 Hz | -2 to -4 dB Bell | Boxy proximity |
+| Cut honk | 500–800 Hz | -1 to -2 dB Bell | Nasal character |
+| Boost presence | 2000–4000 Hz | +1 to +2 dB Bell | Clarity and cut |
+| Dynamic cut sibilance | 6000–8000 Hz | -3 to -5 dB (dynamic) | Sibilance control |
+| Boost air | 12000–16000 Hz | +1 to +2 dB Shelf | |

package/knowledge/reference/metering.md

@@ -0,0 +1,166 @@
+# Metering Reference
+
+Standards, targets, and interpretations for all metering types used in audio production. These are the numbers the agent compares against when analyzing a mix.
+
+## LUFS (Loudness Units relative to Full Scale)
+
+LUFS is the standard loudness measurement for broadcast, streaming, and delivery. Unlike peak meters, LUFS measures perceived loudness weighted by human hearing sensitivity.
+
+### Types of LUFS Measurements
+
+| Measurement | Window | Use Case |
+|-------------|--------|----------|
+| Momentary LUFS | 400 ms | Instantaneous loudness — how loud right now |
+| Short-term LUFS | 3 seconds | Section loudness — how loud this passage |
+| Integrated LUFS | Full duration | Overall loudness of the entire program — delivery standard |
+| LRA (Loudness Range) | Full duration | Dynamic range expressed in LU — difference between quiet and loud |
+
+The **Integrated LUFS** is the number that streaming platforms normalize to.
+
+### Streaming Platform Targets
+
+| Platform | Integrated LUFS Target | True Peak Ceiling | Notes |
+|----------|----------------------|------------------|-------|
+| Spotify | -14 LUFS | -1 dBTP | Normalizes louder content to -14; quieter passes through |
+| YouTube | -14 LUFS | -1 dBTP | Same normalization as Spotify |
+| Apple Music | -16 LUFS | -1 dBTP | More conservative target |
+| Tidal | -14 LUFS | -1 dBTP | |
+| Amazon Music | -14 LUFS | -2 dBTP | |
+| SoundCloud | -14 LUFS (approx) | -1 dBTP | Less strict normalization |
+| Deezer | -15 LUFS | -1 dBTP | |
+| Pandora | -14 LUFS | -1 dBTP | |
+| CD (Red Book) | No standard | 0 dBFS | Pre-streaming; aim for -9 to -6 LUFS for competitive CDs |
+| Vinyl | No standard | N/A | Requires specific mastering considerations |
+| EBU R128 (broadcast) | -23 LUFS | -1 dBTP | European broadcast standard; very conservative |
+| ATSC A/85 (US broadcast) | -24 LUFS | -2 dBTP | US broadcast standard |
+
+### Genre LUFS Targets (beyond platform standards)
+
+Even for streaming platforms, different genres have conventional loudness levels within their norms:
+
+| Genre | Integrated LUFS Range | Notes |
+|-------|----------------------|-------|
+| Classical/orchestral | -23 to -16 LUFS | Maximum dynamic range; don't over-limit |
+| Jazz/acoustic | -18 to -14 LUFS | Preserve natural dynamics |
+| Singer-songwriter | -16 to -13 LUFS | Intimate feel |
+| Rock | -11 to -9 LUFS | Competitive and energetic |
+| Pop | -14 to -10 LUFS | Platform-normalized target |
+| Hip-hop | -10 to -7 LUFS | Bass-heavy loudness; check with streaming normalization |
+| EDM/electronic | -10 to -6 LUFS | Club context is loud; streaming normalization applies |
+| Heavy metal | -11 to -8 LUFS | Very dense mixes |
+
+Note: For streaming, anything louder than -14 LUFS will be turned DOWN by the platform. Delivering at -10 LUFS for Spotify does NOT make your song louder — it gets normalized to -14 anyway. You just lose dynamic range unnecessarily.
+
+### What Streaming Normalization Means for Mixing
+
+- If your integrated LUFS is louder than the platform target: platform turns your song down
+- If your integrated LUFS is quieter than the platform target: your song plays at its natural level (not boosted on most platforms)
+- **Conclusion**: Hit the target, not louder. Exceeding target costs dynamic range with no benefit.
+
+## True Peak (dBTP)
+
+True peak is not the same as sample peak. When audio is decoded and played back (especially in MP3/AAC encoding), intersample peaks can exceed 0 dBFS even if the original samples never did. True peak meters catch these intersample excursions.
+
+| Target | When to Use |
+|--------|-------------|
+| -1.0 dBTP | Standard streaming delivery (Spotify, YouTube, Apple Music) |
+| -0.3 dBTP | Club/EDM delivery; more headroom used for louder levels |
+| -2.0 dBTP | Broadcast (extra margin for encoding artifacts) |
+| 0 dBFS (sample peak) | Not sufficient — always use true peak measurement |
+
+Use Pro-L 2's ISP mode or any true peak limiter to ensure compliance. A sample peak of -0.1 dBFS can produce intersample peaks of +0.5 to +1.0 dBTP in some content.
+
+## Peak Level Standards
+
+| Measurement Point | Standard Peak Level | Notes |
+|------------------|---------------------|-------|
+| Individual track pre-processing | -12 to -6 dBFS peak | Gain staging target for headroom |
+| Individual track average (RMS) | -18 dBFS (±3 dB) | Gain staging target |
+| Bus (drum, vocal, guitar) output | -10 to -8 dBFS peak | Before mix bus |
+| Mix bus pre-limiting | -6 to -3 dBFS peak | Headroom before limiter |
+| Mix bus pre-limiting (average) | -18 to -14 dBFS RMS | |
+| Final export (peak) | -0.3 to -1.0 dBTP | True peak after limiter |
+
+## Crest Factor (Dynamic Range Health)
+
+Crest factor = peak level - RMS level, measured in dB. A higher crest factor means more dynamic range — the difference between the loudest peaks and the average level.
+
+| Crest Factor | Interpretation | Likely Cause |
+|-------------|---------------|-------------|
+| 20+ dB | Excellent dynamics | Minimal compression, natural performance |
+| 14–20 dB | Good dynamics | Appropriate compression for most genres |
+| 10–14 dB | Moderate compression | Rock/pop normal range |
+| 8–10 dB | Heavy compression | EDM, hip-hop; approaching limiting territory |
+| 6–8 dB | Over-compressed | Mix will sound fatiguing; losing dynamics |
+| Below 6 dB | Heavily over-limited | "Loudness war" territory; sounds flat and lifeless |
+
+**How to measure**: Take peak reading minus RMS reading over the same time window.
+
+### Crest Factor Targets by Genre
+
+| Genre | Healthy Crest Factor |
+|-------|---------------------|
+| Classical/orchestral | 20–30 dB |
+| Jazz/acoustic | 14–22 dB |
+| Singer-songwriter | 12–18 dB |
+| Country | 10–14 dB |
+| Rock | 10–14 dB |
+| Pop (modern) | 8–12 dB |
+| Heavy metal | 8–10 dB |
+| Hip-hop | 8–12 dB |
+| EDM/club | 6–10 dB |
+
+## Stereo Correlation
+
+Stereo correlation measures the relationship between the left and right channels.
+
+| Correlation Value | Interpretation | Action |
+|------------------|---------------|--------|
+| +1.0 | Perfectly mono (L and R identical) | No stereo information |
+| +0.8 to +1.0 | Mono-compatible stereo | Excellent — sounds good mono and stereo |
+| +0.5 to +0.8 | Wide stereo | May have some center cancellation in mono |
+| 0 to +0.5 | Very wide | Check mono compatibility — some content may disappear |
+| -0.5 to 0 | Out-of-phase stereo | Problem — content cancels in mono sum |
+| -1.0 | Fully out of phase | Complete mono cancellation — this is broken |
+
+**Target for music delivery**: Correlation above +0.5, ideally +0.7 or higher. The bass register (below 100 Hz) should always be at +0.99 or higher (near perfect mono).
+
+## RMS Levels (Approximate Mix References)
+
+These are typical readings when the master bus is playing back a well-balanced mix in the chorus:
+
+| Genre | Short-term LUFS (chorus) | Mix Bus RMS |
+|-------|-------------------------|------------|
+| Classical (tutti) | -18 to -12 LUFS | -20 to -14 dBFS |
+| Acoustic ballad | -18 to -14 LUFS | -20 to -16 dBFS |
+| Pop (chorus) | -12 to -8 LUFS | -14 to -10 dBFS |
+| Rock (chorus) | -10 to -8 LUFS | -12 to -10 dBFS |
+| Hip-hop (drop) | -10 to -6 LUFS | -12 to -8 dBFS |
+| EDM (drop) | -8 to -4 LUFS | -10 to -6 dBFS |
+
+## Headroom at Each Processing Stage
+
+For a healthy mix signal chain:
+
+| Stage | Headroom Available | Peak Level |
+|-------|------------------|-----------|
+| Microphone preamp output | 6–12 dB before clip | -12 to -6 dBFS peaks |
+| A/D converter input | 6 dB | -6 dBFS peaks maximum |
+| Individual track (post-gain stage) | 6–12 dB | -18 dBFS average |
+| Compressor input | 6 dB above threshold | Threshold at -18 to -24 dBFS typically |
+| Bus input | 6 dB | -12 to -6 dBFS peaks |
+| Mix bus pre-limiter | 6–10 dB | -10 to -6 dBFS peaks |
+| Limiter output | 1 dBTP headroom | -1.0 dBTP |
+| Final export | Compliant with platform | Per platform spec |
+
+## Common Metering Mistakes
+
+| Mistake | Correct Practice |
+|---------|-----------------|
+| Using sample peak as delivery spec | Use true peak (dBTP) measurement |
+| Measuring LUFS on 30 seconds for integrated | Integrated LUFS requires full song duration |
+| Ignoring correlation meter | Check before delivery; correlations below +0.5 indicate phase issues |
+| Setting gain for momentary LUFS | Set gain for integrated LUFS |
+| Not accounting for intersample peaks | Always use a true peak limiter (ISP mode) |
+| Comparing short-term LUFS between songs | Compare integrated LUFS for an apples-to-apples comparison |
+| Ignoring crest factor | A mix with good integrated LUFS but low crest factor is over-compressed |

package/knowledge/workflows/drum-bus.md

@@ -0,0 +1,211 @@
+---
+name: Drum Bus
+id: drum-bus
+description: Build a drum bus processing chain for glue and impact
+---
+
+# Drum Bus
+
+## When to Use
+
+When a session has multiple drum tracks (kick, snare, toms, overheads, rooms) that need to be unified into a coherent, punchy drum sound. The drum bus is the glue between individual drum elements.
+
+Use this workflow when:
+- Drums feel loose and un-glued despite individual track processing
+- The user says "the drums don't punch" or "drums sound separate"
+- Starting a new session with a live drum kit recording
+- Building a parallel compression setup for maximum impact
+
+## Prerequisites
+
+- Individual drum tracks exist (kick, snare, at minimum)
+- A drum bus (folder track / submix) exists or can be identified
+- Gain staging complete on individual drum tracks
+- Play a section with full drum performance (not just a single hit)
+
+## Step-by-Step
+
+### Step 1: Save snapshot
+
+```
+tool: snapshot_save
+params:
+  name: "pre-drum-bus"
+  description: "Before drum bus processing"
+```
+
+### Step 2: Identify drum tracks
+
+```
+tool: list_tracks
+```
+
+Identify:
+- Individual drum tracks: kick, snare, hi-hat, toms, overheads, room mics
+- Drum bus/folder track (receives signal from all drum tracks)
+- Parallel drum bus (if it exists — if not, create one in Step 4)
+
+Note the `trackIndex` for the drum bus.
+
+### Step 3: Verify drum bus routing
+
+```
+tool: get_track_routing
+params:
+  trackIndex: [drum bus index]
+```
+
+Confirm that kick, snare, and other drum tracks are sending to the drum bus. If routing is incorrect, flag to the user — cannot fix routing via the current MCP toolset.
+
+### Step 4: Add compressor to drum bus (VCA glue compressor)
+
+Select: Pro-C 2 Bus mode preferred, ReaComp as fallback.
+
+```
+tool: add_fx
+params:
+  trackIndex: [drum bus index]
+  fxName: "Pro-C2"
+```
+
+Settings (VCA glue — tighten, don't squash):
+
+| Parameter | Value | Why |
+|-----------|-------|-----|
+| Style | Bus | VCA bus compression character |
+| Ratio | 4:1 | Moderate |
+| Attack | 10–20 ms | Preserve kick and snare transient attack |
+| Release | Auto or 80–150 ms | Musical, follows drum rhythm |
+| Threshold | Set for 3–5 dB GR | Light glue |
+| Knee | 4–6 dB | Soft knee for transparent onset |
+| Sidechain HPF | 80–100 Hz | Prevent kick fundamental from triggering comp too hard |
+
+```
+tool: set_fx_parameter
+# Set Sidechain HPF to 80 Hz (prevents low-frequency triggered pumping)
+```
+
+Play the chorus/main groove section while adjusting threshold until GR meter shows 3–5 dB GR. The drums should feel "tighter" without losing punch.
+
+### Step 5: Add EQ after compressor on drum bus
+
+Subtle EQ after the compressor can shape the drum bus tone.
+
+```
+tool: add_fx
+params:
+  trackIndex: [drum bus index]
+  fxName: "Pro-Q3"  # or ReaEQ
+```
+
+Standard drum bus EQ (gentle):
+- No HPF needed (bus inherits from individual tracks)
+- Optional: Bell boost at 60–80 Hz (+1 to +2 dB) if kick punch needs reinforcement
+- Optional: Bell cut at 200–350 Hz (-1 to -2 dB) if drum bus sounds boxy
+- Optional: Bell boost at 8–10 kHz (+1 to +1.5 dB) for cymbal air and openness
+
+Keep EQ moves subtle — ±2 dB maximum on a bus.
+
+### Step 6: Set up parallel compression bus
+
+The signature "rock/metal drum impact" sound requires parallel compression:
+
+If a parallel drum bus doesn't exist, alert the user that one should be created (routing setup required). If it does exist:
+
+```
+tool: add_fx
+params:
+  trackIndex: [parallel drum bus index]
+  fxName: "Pro-C2"  # or ReaComp
+```
+
+Settings (heavy crush, then blend back):
+
+| Parameter | Value | Why |
+|-----------|-------|-----|
+| Style | FET | Aggressive character |
+| Ratio | 20:1 | Near-limiting |
+| Attack | 0.1–2 ms | Kill all transients |
+| Release | 50–150 ms | Some pump for energy |
+| Threshold | Set for 15–20 dB GR | Crush the drums |
+| Mix (if available) | 100% | Wet only — blend with dry on fader |
+
+```
+tool: set_fx_parameter
+# Set ratio to 20:1 via normalized value
+```
+
+Adjust parallel bus fader to blend: start at -12 dB and increase until drums have more punch and weight without sounding over-compressed. Typical final blend: -6 to -9 dB fader on parallel bus relative to main drum bus.
+
+### Step 7: Individual track cleanup (if not already done)
+
+Check these essential individual tracks:
+
+**Kick drum:**
+```
+tool: read_track_spectrum
+params:
+  trackIndex: [kick track index]
+```
+Look for: punch at 60–80 Hz, remove boxiness at 300–400 Hz, ensure click/attack at 3–5 kHz.
+
+**Snare:**
+```
+tool: read_track_spectrum
+params:
+  trackIndex: [snare track index]
+```
+Look for: crack at 200 Hz, remove excessive ring (narrow cut where ring resonates), presence at 5–8 kHz.
+
+### Step 8: Check mono compatibility of drum bus
+
+Overhead mics and room mics create stereo information. Ensure the drum bus is mono-compatible:
+
+Play the mix with overheads — then listen to how it sounds in mono. If cymbals disappear, the overhead panning or phase relationship is causing comb filtering. This is an advanced issue to flag to the user rather than fix automatically.
+
+### Step 9: Set drum bus level relative to mix
+
+```
+tool: read_track_meters
+params:
+  trackIndex: [drum bus index]
+```
+
+Drum bus output target: -10 to -8 dBFS peak before going to mix bus. This leaves room for the rest of the mix. If the drum bus is louder than this, reduce the drum bus fader.
+
+### Step 10: Save snapshot
+
+```
+tool: snapshot_save
+params:
+  name: "post-drum-bus"
+  description: "Drum bus: VCA glue comp + EQ + parallel compression blend"
+```
+
+## FX Chain Order (on the drum bus)
+
+1. **Bus compressor** (VCA/Bus style) — glue and control
+2. **EQ** — subtle tonal shaping after compression
+3. **Saturation** (optional) — subtle harmonic content, especially for sample-based drums
+4. **Limiter** (optional safety) — prevent drum bus from clipping; set at -0.5 dBFS ceiling
+
+Parallel compression is on a separate bus, not an insert.
+
+## Verification
+
+After completing the drum bus workflow:
+
+1. Play the full mix — drums should feel tight, cohesive, and punchy
+2. Check that the compressor's GR is working (3–5 dB reduction, not 0 or 15+)
+3. Kick transient should still click through clearly despite compression
+4. Snare crack should be present
+5. No "pumping" that sounds unintentional (test by adjusting sidechain HPF if pumping is audible)
+6. Parallel compression blend: remove parallel bus and confirm the difference — it should add weight without changing the character
+
+## Common Pitfalls
+
+- **Too-fast attack on drum bus**: Attack below 5 ms kills the kick and snare transients — the most common drum bus mistake. Start at 10–20 ms and adjust.
+- **No sidechain HPF**: Without filtering the sidechain, the kick's sub frequencies trigger the compressor too hard, causing audible pumping. Always set HPF to 80–100 Hz on drum bus sidechain.
+- **Parallel compression too loud**: The parallel bus should add weight, not presence. If the parallel bus is louder than the main bus, the uncompressed transients are lost. Blend conservatively.
+- **Bus compression on already over-compressed individual tracks**: If kick and snare are already clamped hard at the track level, bus compression has nothing musical to do. Check individual track compression first.
+- **Not using the drum bus gain structure**: The drum bus fader should control how loud the drums sit in the mix. Don't mix drums by adjusting individual track faders after the bus is set up.