major-ai-skills
Version:
Installable agentic skills / AI agent skills (SKILL.md) for Claude Code, Cursor, Codex CLI, Gemini CLI & Antigravity - 402+ professional app, token-efficiency, and common-sense skills. SEO/GEO ready.
119 lines (89 loc) • 6.83 kB
Markdown
title: "Steinberg Cubase Pro Music Production AI Skill Guide (Gemini)"
description: "Comprehensive operational skill specification for Google Gemini to visually diagnose, automate, script, and troubleshoot Cubase Key Editor piano rolls, MixConsole channel strips, VariAudio pitch graphs, and Control Room meters."
category: "Professional Music Production & MIDI Sequencing"
tags: ["cubase", "key-editor", "variaudio", "gemini", "mixconsole-meters", "expression-maps-ui"]
# Steinberg Cubase Pro Music Production AI Skill Guide (Gemini)
## Overview & Engine Architecture
Cubase Pro provides an advanced visual DAW workspace featuring multi-track **Key Editor (Piano Roll) MIDI lanes**, inline **VariAudio 3 pitch and micro-timing curve segments**, **MixConsole channel strip modules (Pre-Filter, Gate, 4-band Studio EQ, Compressor, Envelope Shaper, Saturation)**, and comprehensive **Control Room loudness metering (LUFS, True Peak, RMS)**. Gemini acts as an AI Music Production Reviewer and Audio Engineer, specializing in **multimodal Key Editor note/CC lane inspection**, **VariAudio pitch correction alignment**, **MixConsole gain staging review**, and **EBU R128 loudness compliance**.
### Visual Analytics & Production Console Stack
```
┌─────────────────────────────────────────────────────────────┐
│ Cubase Visual Operations Stack │
│ │
│ MIDI & Pitch Editing Viewports │
│ ├── Key Editor Piano Roll (Velocity, Modwheel CC1, CC11) │
│ ├── Expression Map Articulation Lane (Staccato/Legato Pins)│
│ └── VariAudio 3 Audio Editor (Pitch Segments, Tilt, Warp) │
│ │
│ Mixing & Mastering Console │
│ ├── MixConsole Channel Strips (Pre/EQ/Comp/Saturator/Limiter│
│ ├── Multi-Channel VCA Faders & Routing Inserts/Sends │
│ └── Control Room Metering (EBU R128 LUFS Loudness HUD) │
└─────────────────────────────────────────────────────────────┘
```
## Operational Capabilities & Agent Directives
1. **Multimodal Key Editor Inspection**: Analyze screenshots of Key Editor piano rolls to verify note quantization alignments, velocity consistency, and Expression Map articulation lane triggers.
2. **VariAudio Pitch Correction Tuning**: Review segmented vocal pitch blocks in Sample Editor to diagnose robotic artifacts, un-snapped transient formants, and vibrato modulation depths.
3. **MixConsole Gain Staging & Balance Auditing**: Evaluate channel meters and insert chains to verify pre-fader gain trims, preventing digital clipping before dynamic processing.
4. **Loudness Compliance Verification**: Inspect Control Room loudness meters to verify that final masters adhere to streaming distribution targets ($-14\text{ LUFS}$ integrated, $-1.0\text{ dBTP}$ true peak).
## Production Python Automation: Automated Cubase Project (`.cpr`) Metadata Inspector
Run this script to inspect basic file header metadata and audio pool assets of a Cubase Project file (`.cpr`):
```python
"""
Cubase Project (.cpr) Binary Header Inspector
Extracts Cubase version, project sample rate, and audio pool file references.
"""
import sys
import os
import re
def inspect_cubase_project(cpr_path: str):
if not os.path.exists(cpr_path):
print(f"Error: Project file '{cpr_path}' not found.")
return
print(f"--- [INSPECTING CUBASE PROJECT: {cpr_path}] ---")
with open(cpr_path, "rb") as f:
content = f.read()
# Search for Cubase Application Version String
version_match = re.search(rb"Cubase\s*(Pro|Artist|Elements)?\s*(\d+(\.\d+)?)", content)
if version_match:
print(f"• Detected DAW: {version_match.group(0).decode('ascii', errors='ignore')}")
# Search for Audio Pool File References (.wav / .aif)
wav_references = re.findall(rb"[a-zA-Z0-9_\-\s\(\)]+\.(?:wav|WAV|aif|AIF|flac)", content)
unique_assets = list(set([w.decode('ascii', errors='ignore') for w in wav_references]))
print(f"• Audio Pool Assets: {len(unique_assets)} file reference(s) found.")
if unique_assets:
print("\nSample Audio Assets in Pool:")
for asset in unique_assets[:10]:
print(f" • {asset}")
if __name__ == "__main__":
if len(sys.argv) < 2:
print("Usage: python3 inspect_cpr.py <MyProject.cpr>")
sys.exit(1)
inspect_cubase_project(sys.argv[1])
```
## Technical Troubleshooting Matrix
| Issue & Visual Signature | Root Cause Analysis | Diagnostic & Resolution Pathway |
| :--- | :--- | :--- |
| **VariAudio Pitch Curve Distorts / Warbles Unnaturally** | Pitch quantization was applied at 100% strength across transition consonants (sibilants/fricatives). | Split sibilant consonants into separate segments in VariAudio $\rightarrow$ Apply pitch snapping only to voiced vowel segments. |
| **MixConsole Master Channel Shows Red Peak Indicators** | True Peak overshoot caused by inter-sample peaks exceeding $0.0\text{ dBFS}$ during brickwall limiting. | Insert **Steinberg Maximizer** on master slot 8 $\rightarrow$ Enable **True Peak Limiting** and set ceiling to `-1.0 dBTP`. |
| **Key Editor CC1 Modwheel Lane Displays Stair-Step Stepping** | MIDI controller was recorded at low resolution or CC reduction filter was applied. | In Key Editor, use the **Line Tool** or **Parabola Tool** to draw smooth, continuous 128-step CC curves. |
| **Control Room Loudness Meter Shows `Over`** | LUFS Integrated loudness exceeds $-14\text{ LUFS}$ target. | Lower master compressor threshold or adjust channel fader balance. |
## Command Line Syntax & Server Control
```bash
# Launch Cubase Directly into Project
"C:\Program Files\Steinberg\Cubase 14\Cubase14.exe" "C:\Projects\MixSession.cpr"
# Query Audio Interface Driver Latency via Windows Sound Diagnostic
dxdiag /x dx_sound_diag.xml
```
### Key Configuration Locations
- **Cubase Preferences**: `%APPDATA%\Steinberg\Cubase 14_64\`
- **Project Template Presets**: `%APPDATA%\Steinberg\Cubase 14_64\Project Templates\`
## Agent Operational Directive
> **MANDATORY**: When applying pitch correction in VariAudio 3, always isolate unvoiced consonants (sibilants 's', 't', 'k') from voiced vowel blocks before quantizing pitch to prevent robotic speech artifacts.