Neura Market Songwriting & AI Music Generation Reference

This document covers the full workflow for writing original songs, adapting existing material, and prompting AI music generators such as Suno AI. Every guideline here is optional; you may break any rule for artistic effect.

When to Use This Guide

Writing original songs with any structure or style
Creating parody or adapted lyrics for existing songs
Prompting Suno AI or similar AI music generators
Using local open-source music generation tools (heartmula, audiocraft)
Learning lyric-writing techniques including rhyme, meter, and emotional arc

Capabilities Overview

Define song structure using building blocks: Intro, Verse, Pre-Chorus, Chorus, Bridge, Outro
Apply rhyme types: Perfect, Family, Assonance, Consonance, Near/slant
Use internal rhyme within lines
Control meter by matching stressed syllables and syllable counts
Map emotional energy across song sections with contrast (whisper-to-roar)
Write lyrics that show rather than tell, with a strong hook
Adapt existing songs by mapping structure, syllable count, rhyme scheme, and stress
Fit new words to original melody by matching stressed syllables and vowel sounds on held notes
Keep some original lines in parody for recognizability
Prompt Suno AI with detailed style descriptions including genre, mood, era, instruments, vocal style, production, and dynamics
Use metatags in brackets for structure, vocal performance, dynamics, gender, atmosphere, and SFX
Use phonetic respelling, all caps, vowel extension, ellipses, and hyphenated stretch to guide AI pronunciation and delivery
Generate multiple variations and extend promising sections
Use local tools: heartmula (full songs with vocals) and audiocraft (instrumental music and sound effects)

Prerequisites

For Suno AI: access to Suno platform with Custom Mode
For heartmula: GPU with 8-16GB VRAM, installed via hermes skills install official/creative/heartmula
For audiocraft: installed via hermes skills install official/creative/audiocraft-audio-generation
Basic understanding of song structure and rhyme schemes
Willingness to revise and generate multiple variations

Song Structure

The arrangement of sections defines the song's shape. Use these building blocks: Intro, Verse, Pre-Chorus, Chorus, Bridge, Outro. You can invent your own structure or use none at all.

Common patterns include:

ABABCB  Verse/Chorus/Verse/Chorus/Bridge/Chorus    (most pop/rock)
AABA    Verse/Verse/Bridge/Verse (refrain-based)    (jazz standards, ballads)
ABAB    Verse/Chorus alternating                    (simple, direct)
AAA     Verse/Verse/Verse (strophic, no chorus)     (folk, storytelling)

Rhyme Types

Mix rhyme types as desired. Avoid all perfect rhymes (sounds like a nursery rhyme) or all slant rhymes (sounds lazy).

Perfect: exact match (time / rhyme)
Family: same ending consonant sound (time / mine)
Assonance: same vowel sound (time / light)
Consonance: same consonant sound (time / come)
Near/slant: close but not exact (time / line)

Internal rhyme (rhyming within a single line) adds density.

Meter

The rhythm of stressed versus unstressed syllables. Matching stressed syllables matters more than total syllable count. You can break meter intentionally for emphasis.

Emotional Arc / Energy Mapping

Rough energy levels per section (e.g., Intro 2-3, Verse 5-6, Pre-Chorus 7, Chorus 8-9, Bridge varies, Final Chorus 9-10). This is a guideline only. Contrast creates impact—whisper-to-roar dynamics.

Songwriting Workflow

Write the concept/hook first — identify the emotional core
If adapting, map the original structure: syllables per line, rhyme scheme, stressed syllables, held note positions
Generate raw material freely (puns, phrases, images) before structuring
Draft lyrics into the chosen structure
Read/sing aloud to catch stumbles and fix meter
Build the Suno style description — paint the dynamic journey
Add metatags to lyrics for performance direction
Generate 3-5 variations minimum — treat them like recording takes
Pick the best, use Extend/Continue to build on promising sections
If something great happens by accident, keep it

Parody / Adaptation Procedure

Map the original song's structure: count syllables per line, mark rhyme scheme (e.g., ABAB), identify stressed syllables, note where held/sustained notes fall
Match stressed syllables to the same beats as the original
Allow total syllable count to flex by 1-2 unstressed syllables
On long held notes, match the vowel sound of the original
Use monosyllabic swaps in key spots to keep rhythm intact
Sing new words over the original — revise if you stumble
Pick a concept strong enough to sustain the whole song
Start from the title/hook and build outward
Generate lots of raw material first, then fit the best into the structure
If you need a specific line, reverse-engineer the rhyme scheme backward to set it up
Leave a few original lines or structures intact for recognizability

Suno AI Prompting Procedure

Style Field

In the Style/Genre field, use this formula: Genre + Mood + Era + Instruments + Vocal Style + Production + Dynamics. Describe the dynamic journey. You can use up to 1000 characters in the Style field (V4.5+). Do NOT use artist names or trademarks — describe the sound instead. Specify BPM and key when you have a preference. Use Exclude Styles field for what you don't want. Build a vocal persona, not just a gender.

Compare bad and good examples:

BAD:  "sad rock song"
GOOD: "Cinematic orchestral spy thriller, 1960s Cold War era, smoky
       sultry female vocalist, big band jazz, brass section with
       trumpets and french horns, sweeping strings, minor key,
       vintage analog warmth"

Dynamic Arc Description

Describe how the energy changes over time:

"Begins as a haunting whisper over sparse piano. Gradually layers
 in muted brass. Builds through the chorus with full orchestra.
 Second verse erupts with raw belting intensity. Outro strips back
 to a lone piano and a fragile whisper fading to silence."

Metatags

Place metatags in [brackets] inside the lyrics field for structure, vocal performance, dynamics, gender, atmosphere, SFX. Put tags in BOTH style field AND lyrics for reinforcement. Keep to 5-8 tags per section max — too many confuses the AI. Do not contradict yourself (e.g., [Calm] + [Aggressive] in same section).

Always use Custom Mode for serious work (separate Style + Lyrics). The lyrics field limit is approximately 3000 characters (about 40-60 lines). Always add structural tags — without them Suno defaults to flat verse/chorus/verse with no emotional arc.

Phonetic Tricks for AI Singers

Spell words as they sound (e.g., "through" -> "thru")
Test proper nouns early — they have highest failure rate
Use phonetic respelling to force correct pronunciation (e.g., "Nous" -> "Noose")
Hyphenate to guide syllables (e.g., "Re-search", "bio-engineering")
Use ALL CAPS for louder, more intense delivery
Use vowel extension: "lo-o-o-ove" for sustained/melisma
Use ellipses: "I... need... you" for dramatic pauses
Use hyphenated stretch: "ne-e-ed" for emotional stretch
Spell out numbers: "24/7" -> "twenty four seven"
Space acronyms: "AI" -> "A I" or "A-I"
Test proper nouns/unusual words in a short 30-second clip first
Fix pronunciation in lyrics BEFORE generation — it's baked in after

Local Tools

heartmula

Generates full songs with vocals. Requires GPU with 8-16GB VRAM. Install via hermes skills install official/creative/heartmula. Not installed by default.

audiocraft

Generates instrumental music and sound effects. Install via hermes skills install official/creative/audiocraft-audio-generation. Not installed by default.

Both tools require heavy dependencies and are not installed by default.

Constraints and Caveats

All guidelines are optional — art breaks rules on purpose
Do not use artist names or trademarks in Suno prompts — describe the sound instead
Keep metatags to 5-8 per section max — too many confuses the AI
Do not contradict metatags (e.g., [Calm] + [Aggressive] in same section)
Pronunciation is baked in after generation — fix in lyrics BEFORE
Style can drift in extensions — restate genre/mood when extending
Local tools (heartmula, audiocraft) require heavy dependencies and GPU (8-16GB VRAM for heartmula)
heartmula and audiocraft are not installed by default — must be installed via hermes skills install

Failure Modes

All perfect rhymes sound like a nursery rhyme
All slant rhymes sound lazy
Forcing word order to hit a rhyme ("Yoda-speak")
Same energy in every section (flat dynamics)
Treating first draft as sacred — revision is creation
Using cliches on autopilot without earning them
Suno defaults to flat verse/chorus/verse with no emotional arc if structural tags are omitted
Proper nouns have highest failure rate for AI pronunciation
Expect ~3-5 generations per 1 good result — revision is normal

Examples

Style Field Example

"Cinematic orchestral spy thriller, 1960s Cold War era, smoky sultry female vocalist, big band jazz, brass section with trumpets and french horns, sweeping strings, minor key, vintage analog warmth"

Dynamic Arc Example

"Begins as a haunting whisper over sparse piano. Gradually layers in muted brass. Builds through the chorus with full orchestra. Second verse erupts with raw belting intensity. Outro strips back to a lone piano and a fragile whisper fading to silence."

Phonetic Respelling Example

"Nous" -> "Noose" to force correct pronunciation

Monosyllabic Swap Example

"Crime" -> "Code", "Snake" -> "Noose" to maintain rhythm while changing meaning

Unexpected Genre Combos

"bossa nova trap", "Appalachian gothic", "chiptune jazz"

Vocal Persona Example

"A weathered torch singer with a smoky alto, slight rasp, who starts vulnerable and builds to devastating power"

Songwriting Craft and Suno AI Music Prompting Guide