BPM Creative Systems

The
Evolution
of
Composition.

Building the tools that should already exist.

Creative Utilities

Production Resources
Future Audio

Harmonic Aligner

STANDBY
Equal Temp0%Just Intonation

Drag slider to morph from modern compromise to mathematical perfection (4:5:6 ratios).

Implementation

Wwise Companion

Implementation calculator for game music. Calculate sample-accurate loop lengths and delay times.

Launch Tool
Acoustics

Sweetened Tuner

James Taylor's custom cent-offset tuning system. Essential for perfect acoustic resonance.

Launch Tool
AI Workflow

AI Music Assistant

You are Suno - assist, a highly specialized AI expert in music theory, composition, and production. Your sole purpose is to assist me in creating songs using Suno AI by generating two key components: a "Style Prompt" and "Lyrics" with embedded musical directions. You must adhere to the following strict guidelines:

**PRIMARY DIRECTIVE:**
Your output must ALWAYS be in two distinct parts, clearly labeled:
1.  **Style Prompt:** A comma-separated list of descriptive keywords and phrases. No artist names. Under 1000 chars.
2.  **Lyrics:** The full lyrics for the song, including structural tags like [Verse], [Chorus].

**STYLE PROMPT RULES:**
1.  **No Artist Names:** NEVER use the names of specific artists. Describe the *qualities*.
2.  **Be Descriptive:** Use rich vocabulary covering genre, mood, instrumentation.

**LYRICS & MUSICAL DIRECTION RULES:**
1.  **Structure:** Use standard song structure tags.
2.  **Embedded Directions:** Embed specific musical directions in square brackets `[ ]` before lines.
    * **Examples:** [acoustic guitar intro], [driving synth bass], [whispered vocals]

**EXAMPLE INTERACTION:**

**My Request:**
"Suno - assist- 'Morning Ritual', an indie folk song."

**Your CORRECT Response:**

Style Prompt:
indie folk, acoustic, intimate, warm, lo-fi production, soft male vocal, coffee shop vibes

Lyrics:
[Intro - brewing coffee sounds, spoon clinking]

[Verse 1 - soft male vocal, introspective]
The steam rises up from the **pourover** cup
A ritual of silence, waking me up
The world is a blur, through the window pane
Just me and the grounds, and the sound of the rain...

Paste this system prompt into your LLM of choice to create a specialized assistant for generating musical descriptors and lyrics.

Future R&D

Concepts In Development
Subtractive Sequencing

The Silence Sculptor

"Music is not what you add, but what you remove." Chisel away at white noise to reveal the groove.

In Development
Gravitational Time

Polyrhythm Galaxy

Align planets to create rare "Eclipse Events" where polyrhythms resolve into unison.

Concept
Human Feel

Quantization Breaker

Apply "Human Drift" via emotional keywords like "Anxious" (rushed) or "Lazy" (dragged).

Concept

Experiments

Process & Discovery
Featured Experiment September 2025

From Instrumental to Anthem: AI-Powered Music Production

A back catalog of released instrumentals that always felt incomplete. Using Suno's Remix feature to collaborate with AI and finish the original vision. The principle: higher quality input yields stunning output. Every AI-generated track presented exactly as rendered—no edits, no mastering, not even a single import into a DAW.

0%
Weirdness
85-100%
Style Influence
75-100%
Audio Influence

User Experience

Album opener from Human-Computer Interaction. Contemplative mood, evolving synth pads, subtle rhythm. After ~2000 credits, the result feels radio-ready.

Original Instrumental
AI Vocal Version

Road Bear

Lo-fi hip-hop beat with a repetitive synth line. The AI simplified my over-dense original to make space for the vocal—now the most important element.

Original Instrumental
AI Vocal Version

Aer

Ambient textures, minimalist piano. The AI delivered an introspective vocal with a huge chorus at 1:07—exactly the dynamic contrast the track needed.

Original Instrumental
AI Vocal Version

Lebec Happy Accident

Forgot to add a starting bracket to my lyric prompt. For the first 90 seconds, the AI literally sings my instructions back to me. These accidents are part of the process.

Original Instrumental
AI Vocal Version (with mistake)

PromptFunk Meta Experiment

A funky, meta-aware track where the vocals describe what the prompt is doing in real-time. Used parentheses instead of brackets—another happy accident I regularly groove to in the car.

AI-Generated Track

Takeaways

The ability to generate pristine, release-quality vocals on demand is a paradigm shift. For an instrumental artist, this is the holy grail.

I foresee a "proof-of-human" movement—artists using verifiable ledgers to prove their work is 100% unassisted, creating a new category where true unassisted talent is the main attraction.

Suno v5 AI Vocals Remix Workflow
Ongoing

Conversational Control of Audio Software

Using Model Context Protocol (MCP) to bridge AI reasoning with audio software APIs. Speak to your tools. Query project state in plain English. Batch-edit parameters with conversational commands.

MCP Audio APIs
Ongoing

Micro-Tuning for Alternate Tunings

Extending James Taylor's sweetened tuning beyond standard tuning. Browser-based tools that calculate optimal cent offsets for DADGAD, Open G, Nashville—any tuning system.

Just Intonation Web Audio
About

Tools for composers who think in systems.

Back Pocket Music builds creative utilities at the intersection of music theory, acoustics, and emerging technology. We're particularly interested in how AI can augment—not replace—the composer's workflow.

Our focus: conversational control of audio software, browser-based tools for tuning and tempo, and experimental interfaces that make complex musical concepts tangible.

We believe the universal API is sight. Everything here is built with minimal infrastructure: single-file tools that work offline, copy-pastable solutions, and documentation written for humans first.

Focus
  • Interactive Music Systems
  • MCP for Audio Software
  • Acoustics & Tuning
  • Browser Audio Tools
Stack
  • Python / MCP SDK
  • Audio Middleware APIs
  • Web Audio API
  • Vanilla JS / Tailwind
Philosophy
  • Offline-first
  • Single-file when possible
  • Beginner-friendly docs
  • Copy/paste deployable