@mthines/reaper-mcp 0.1.0

package/knowledge/workflows/vocal-chain.md

@@ -0,0 +1,246 @@
+---
+name: Vocal Chain
+id: vocal-chain
+description: Build a complete processing chain for lead vocals
+---
+
+# Vocal Chain
+
+## When to Use
+
+When a lead vocal track needs a complete processing chain built from scratch or audited for problems. This workflow builds a professional vocal chain in the correct plugin order with genre-appropriate settings.
+
+Use this workflow when:
+- A new vocal track has been added with no processing
+- The vocal is present but not cutting through the mix
+- The vocal sounds harsh, muddy, or uncontrolled
+- The user says "fix the vocals" or "make the vocals sound professional"
+
+## Prerequisites
+
+- Lead vocal track identified (track name should contain "voc", "lead", or similar)
+- Gain staging already complete (-18 dBFS average on vocal track)
+- Genre is known (settings vary significantly by genre)
+- Track has audio playing — meters must be readable
+
+## Step-by-Step
+
+### Step 1: Save a snapshot
+
+```
+tool: snapshot_save
+params:
+  name: "pre-vocal-chain"
+  description: "Before building vocal processing chain"
+```
+
+### Step 2: Identify the vocal track
+
+```
+tool: list_tracks
+```
+
+Find the lead vocal track by name. Get its `trackIndex`.
+
+### Step 3: Read the current vocal spectrum
+
+```
+tool: read_track_spectrum
+params:
+  trackIndex: [vocal track index]
+```
+
+Look for:
+- Low-frequency buildup below 100 Hz (rumble, handling noise)
+- Presence of 200–400 Hz (boxiness / proximity effect)
+- Peak in 2–5 kHz (harshness zone — check if it's causing problems)
+- Sibilance energy above 6 kHz (will need de-essing)
+- Overall spectral tilt (too bright, too dark?)
+
+### Step 4: Add EQ (first in chain)
+
+Select EQ based on plugin resolver: prefer Pro-Q 3, fall back to ReaEQ.
+
+```
+tool: add_fx
+params:
+  trackIndex: [vocal track index]
+  fxName: "Pro-Q3" # or "ReaEQ" if not available
+```
+
+Set EQ parameters:
+
+```
+tool: set_fx_parameter
+params:
+  trackIndex: [vocal track index]
+  fxIndex: 0  # EQ is first FX
+  paramName: "Band 1 Shape"
+  value: [High Cut = 4]  # HPF
+
+tool: set_fx_parameter
+params:
+  trackIndex: [vocal track index]
+  fxIndex: 0
+  paramName: "Band 1 Frequency"
+  value: [80 to 100 Hz — normalized to 0.0–1.0 range]
+```
+
+Standard vocal EQ moves:
+- Band 1: HPF at 80–100 Hz (slope: 24 dB/oct)
+- Band 2: Bell cut at 300–400 Hz, -2 to -3 dB, Q=1.0 (proximity/boxiness)
+- Band 3: Dynamic EQ or Bell cut at 7–9 kHz for sibilance (set dynamically if Pro-Q 3)
+- Band 4: High shelf boost at 10–12 kHz, +1 to +2 dB (air)
+
+### Step 5: Add first compressor (FET-style, for control and character)
+
+Select compressor: prefer Pro-C 2 in FET mode, fall back to JS 1175, fall back to ReaComp.
+
+```
+tool: add_fx
+params:
+  trackIndex: [vocal track index]
+  fxName: "Pro-C2"
+```
+
+Set parameters for FET style (generic genre settings; adjust for genre from genre file):
+- Style: FET
+- Ratio: 4:1
+- Attack: 10–15 ms
+- Release: 50–80 ms
+- Threshold: Set for 4–6 dB GR while playing chorus
+- Knee: 2–4 dB
+
+```
+tool: read_track_meters
+params:
+  trackIndex: [vocal track index]
+```
+
+Play chorus and adjust Threshold until GR meter shows 4–6 dB.
+
+### Step 6: Add second compressor (Opto-style, for smoothing)
+
+```
+tool: add_fx
+params:
+  trackIndex: [vocal track index]
+  fxName: "Pro-C2"  # second instance
+```
+
+Set parameters for Opto style:
+- Style: Opto
+- Ratio: 3:1–4:1
+- Attack: 30–50 ms
+- Release: Auto
+- Threshold: Set for 2–3 dB GR
+- Knee: 8–10 dB (very soft)
+
+### Step 7: Add de-esser (if sibilance is a problem)
+
+If spectrum analysis in Step 3 showed prominent 6–9 kHz energy, or if sibilance is audible:
+
+Option A: Use Pro-Q 3 dynamic EQ band (set up in Step 4 above, Band 3 dynamic cut at 7–9 kHz).
+
+Option B: Dedicated de-esser plugin if available:
+```
+tool: add_fx
+params:
+  trackIndex: [vocal track index]
+  fxName: "ReaXcomp"  # or any de-esser plugin
+```
+
+Target: 6500–9000 Hz, threshold set so it only triggers on harsh S sounds.
+
+### Step 8: Set up reverb send/return
+
+Reverb should be on a dedicated return bus (send/return architecture), not as an insert.
+
+If no reverb bus exists:
+- Ask user to create one or use `add_track` if available
+- Alternatively, add reverb as an insert with Dry signal at -inf dB (pure send simulation)
+
+Reverb settings (genre-specific; these are neutral defaults):
+- Pre-delay: 20–30 ms (separates vocal from reverb onset)
+- RT60 / Size: 1.5–2.0s (medium hall or chamber)
+- High-frequency damping: Moderate (reverb should be darker than the dry vocal)
+- Mix: 100% wet on the return bus
+
+Adjust send level from vocal to reverb bus:
+- Dry/intimate genres: -20 to -24 dB send (reverb barely audible)
+- Wet/ambient genres: -12 to -15 dB send (reverb is part of the sound)
+
+### Step 9: Set up delay send (optional but recommended)
+
+```
+tool: add_fx
+params:
+  trackIndex: [reverb return / delay return track]
+  fxName: "ReaDelay"
+```
+
+Settings:
+- Tempo-sync: 1/4 note (classic) or 1/8 note (modern)
+- Feedback: 20–40%
+- LPF: 6000 Hz (dark repeats, don't compete with dry)
+- HPF: 200 Hz (keep delay from muddying bass)
+- Send level from vocal: -12 to -20 dB
+
+### Step 10: Final level check
+
+Play the mix with the vocal chain active.
+
+```
+tool: read_track_meters
+params:
+  trackIndex: [vocal track index]
+```
+
+After the compressors, the vocal output should:
+- Average: -18 to -15 dBFS (slightly hotter than raw input to account for GR recovery)
+- Peak: Not exceeding -6 dBFS
+
+Adjust output gain on the second compressor if needed.
+
+### Step 11: Save post-vocal snapshot
+
+```
+tool: snapshot_save
+params:
+  name: "post-vocal-chain"
+  description: "Vocal chain complete: EQ + FET comp + Opto comp + reverb/delay sends"
+```
+
+## FX Chain Order
+
+The correct insert order for vocal processing:
+
+1. **Pitch correction** (if needed — not covered by this workflow; manual setup)
+2. **EQ** — Corrective EQ first (HPF, problem frequencies)
+3. **FET compressor** — First compression stage; character and control
+4. **Opto compressor** — Second stage; smoothing and transparency
+5. **De-esser** — After compression (compression can make sibilance worse)
+6. **Saturation** (optional — adds warmth, presence)
+7. **EQ (optional second EQ)** — Additive EQ after dynamics if needed
+8. **Output trim** — Level matching after chain
+
+Sends go to: Reverb bus, Delay bus (post-fader sends by default in REAPER).
+
+## Verification
+
+After building the chain, verify:
+
+1. Vocal sits above the mix in the 1–4 kHz presence range without harshness
+2. No audible compression pumping or breathing on sustained notes
+3. Sibilant S sounds are controlled but not lispy
+4. Reverb adds space without washing the vocal into the mix
+5. Vocal cuts through when listening at low monitoring levels
+
+## Common Pitfalls
+
+- **Reversing compressor order**: FET first (character), then Opto (smoothing). Not the other way.
+- **De-esser before compression**: Compression amplifies sibilance; de-ess after the compressors.
+- **Reverb as insert**: Always use send/return for reverb. Insert reverb doubles the dry signal internally and causes comb filtering.
+- **Too much attack time on FET comp**: If the attack is slow (>30 ms), the compressor catches the middle of the word, not the consonant. Vocals need 10–15 ms attack.
+- **Not checking sibilance after compression**: A vocal that sounds fine raw may have harsh S sounds after 6+ dB of compression. Always check sibilance with the chain active.
+- **Long reverb pre-delay causing rhythmic issues**: If the reverb onset is delayed 50+ ms, it can sound like an echo rather than space. 20–30 ms is the sweet spot.