Suno Cohesion Conventions
1. Purpose
This document records established operational conventions that have proven effective in maintaining tonal, stylistic, and structural cohesion across multi-track EPs produced in Suno.
These conventions:
- are not canonical rules,
- do not define ritual meaning, belief, or structure,
- and do not impose hard governance constraints.
They exist to document execution context expectations that reduce stylistic drift and fragmentation during generation.
2. Scope and Applicability
These conventions apply when:
- Suno is used as the primary generation engine
- multiple tracks are intended to form a cohesive EP or ritual system
They do not apply to:
- single standalone tracks
- non-Suno generation environments
- canon definition or schema construction
3. Playlist Continuity
Convention
- EPs are expected to be generated within a single Suno playlist.
- Playlist continuity provides implicit contextual memory across generations.
- Earlier tracks in a playlist exert stronger stylistic and tonal influence on subsequent tracks.
Observed Effects
Using a shared playlist has been observed to:
- reduce stylistic variance between tracks
- stabilize orchestral density and pacing
- reinforce shared reverb, spatial, and temporal characteristics
- lower the need for increasingly restrictive style prompts
Deviation
If tracks are generated outside a shared playlist, this should be noted in lab documentation.
4. Persona Persistence
Convention
- A single persona is expected to be used per EP.
- Personas are treated as persistent contextual stabilizers, not creative agents.
- Persona changes mid-EP significantly increase drift risk.
Stabilized Aspects
Persona persistence has been observed to stabilize:
- vocal ontology (e.g. choir behavior, delivery style)
- tempo bias and pacing tendencies
- orchestral balance and density
- affective ceiling (reducing unwanted uplift or expressivity)
Deviation
If persona changes are introduced, they should be:
- explicit
- intentional
- documented as a contextual shift rather than an optimization
5. Relationship Between Playlist and Persona
- Playlists provide horizontal continuity (track-to-track memory).
- Personas provide vertical continuity (generation bias across sessions).
Used together, they form a contextual envelope that reduces reliance on:
- excessive negative prompt exclusions
- post-generation correction
- iterative prompt tightening
Neither replaces prompt discipline; they reinforce it.
6. Prompt Discipline (Contextual)
While not the focus of this document, effective use of playlists and personas has been observed to work best when combined with:
- long-form, descriptive style prompts
- explicit exclusion of undesired tropes
- avoidance of emotional or evaluative adjectives
- preference for first-pass acceptance over micro-iteration
These are supporting practices, not requirements.
7. Drift Considerations
Ignoring playlist or persona continuity in multi-track EPs has been observed to:
- increase stylistic divergence
- introduce unintended genre bleed
- destabilize pacing and tonal gravity
This does not invalidate results, but raises drift risk and documentation burden.
8. Relationship to Canon and Governance
- These conventions do not define system meaning or belief.
- They do not override frozen schemas or anchors.
- They do not authorize regeneration or modification.
They describe execution context only.
9. Use in Future EPs
For future Suno-based EPs:
- Playlist continuity and persona persistence should be assumed by default.
- Deviations should be explicitly noted, not justified.
No approval process is required.
10. Summary
These conventions document how cohesion was practically achieved, not how it must always be achieved.
They are intended to:
- increase reproducibility
- reduce accidental drift
- improve system legibility
They do not constrain creative intent or ritual ontology.