Algorithmic Point · Line · Plane

Composition MMXXVI

Music-to-Graphic Generator — Inspired by Wassily Kandinsky
v5.23.2 · Composition MMXXVI
01
BUILD YOUR PLAYLIST
Search, add tracks, choose WHEN, preview the algorithmic impression, then generate.
API / no track selected
backend not checked
Search for a song to begin.
TRACKSAdd songs. No per-track rating is required.
ORDERED TRACK IDENTITY0 TRACKS
Audio features are resolved first. Meaning, metadata, relationship, color, variance, and sequence remain bounded structural evidence.
Search above and add at least two tracks.
READ ONLY · Analysis details
AUDIO EVIDENCE COVERAGE0 / 0
WAITING FOR TRACKSAudio-first analysis starts when tracks are added.
OPTIONAL · 聴く時間・シチュエーション
聴く場面を添える(任意)未選択でも生成できます。時間とシチュエーションの影響は合計20%以内。色相は曲から決まります。
WHEN · 任意
SITUATION · 任意
ALGORITHMIC PLAYLIST IMPRESSION PREVIEWREAD ONLY · Updates automatically as your playlist changes.
Add at least two tracks to build the profile.
OPTIONAL · Add words or refine the interpretation
LYRICS / SCENE TEXT
Optional. Paste lyrics, image words, or scene fragments. The parser looks for colors, weather, time-of-day, landscapes, motion, and material cues, then pushes background color and composition bias.
No semantic scene detected yet.
Palette cue: —
OPTIONAL · Interpretation settingsInner · audio / semantic mix
INNER
The shared 12-axis interpretation layer connecting song analysis to the visual composition engine.
interpretation layer
Search and select a track. The resulting Audio Inner and Title/Text Inner will be combined here.
Interpretation mix
Audio 75%Semantic 25%
Audionone
Tempo
Dynamics
Spectrum
OPTIONAL · Composition settingsSeed · density · direction · core / extension
Seed
Used as the local-axis reference in No Dominant Axis mode.
Controls local directional differences and spread.
02
ARTWORK
The preview stays visible while you search songs and refine Inner. Selecting a track auto-generates the first result, while title / lyrics cues push color and scene interpretation.
Seed —
GENERATION STATUSSelect a song to auto-generate artwork.No track selected.
OVERLAYS
SELECT A SONGArtwork is generated only after the selected track has been analyzed.
No neutral placeholder composition is used in API mode.
READ ONLY · Why this image?Song imprint · scene · binding · affect

Session Status

Song selection auto-generates the first image.

Song Imprint

How this song is currently steering the image

Semantic Scene

title / lyrics → color, environment, and atmosphere cues

Song Binding

Descriptors → Intermediate Inner → Image Directives

Visual Program

Musical Gestalt → Affective Archetype → Palette / Form / Rhythm Grammar

Visual Affect

Affective Archetype → Envelope → Palette Constraints → V Score
RESEARCH / AUDITSong → Musical Gestalt → Visual Program → Image → K / X / M / B / V / P

Reference Seed Analysis

Seed 1809818100 / current composition profile

Algorithm Extension — Research

Computational Extension

Mode Details

Saved References

0 references

Profile Notes

Reference seeds store compositional metrics, not image files.

Current vs References

Compare compositional conditions without copying one layout.

Preference Profile

Aggregate saved references to estimate a preferred composition region.

Snapshot

Same seed / current mode
current mode snapshot