danielritchie/cinematic-mood-palette

Cinematic Mood Palette

Curated mappings between affective states and cinematic visual expression. The goal is to describe how filmmakers translate psychological affect into color and perceptual parameters.

~80 mappings, including emotional states, cinematic aesthetics, and spatial calibration points.

---

What This Is

A collection of anchor points in a 5-dimensional emotional space, each paired with corresponding cinematic color and perceptual parameters.

It functions as a reference map showing how affective states can be expressed through visual design choices used in film and photography. Nearby points in this space can be meaningfully interpolated to derive intermediate visual treatments. Input coordinates are deliberately amplified to emphasize emotional extremes, creating visually distinctive reference points rather than modeling naturalistic affect distributions.

The emotional space extends the classic Valence–Arousal–Dominance (VAD) model with:

• Complexity (visual activity/richness)
• Coherence (organizational harmony)

These additions help describe how visual systems express mood through light, color, and composition.

---

Why This Might Be Useful

This dataset documents a systematic relationship between affect theory and cinematic color language.

• Each mapping represents a documented pattern in how emotion is stylized visually
• The values encode relationships between emotional dimensions and visual parameters
• It provides a structured vocabulary for translating mood into color/light choices

---

Potential Uses

This manifold could support:

• Translating emotional data into expressive color palettes
• Providing emotional constraints for generative visual systems
• Studying relationships between affect dimensions and design choices
• Prototyping mood-driven visual interfaces

Note: With ~80 samples, this works best as a reference structure or semantic anchor rather than bulk training data.

---

Structure

{
  "name": "pure_joy",
  "source": "Wes Anderson pastels",
  "input": [0.95, 0.85, 0.9, 0.65, 0.8],
  "output": [1, 0.85, 0.7, 0.8, 0.7]
}

All values are normalized to [0, 1].

Input Dimensions (Emotional Space)

Dimension	Meaning
Valence	Positive ↔ Negative emotional tone
Arousal	Calm ↔ Energized intensity
Dominance	Passive ↔ Powerful presence
Complexity	Minimal ↔ Rich visual activity
Coherence	Chaotic ↔ Harmonious organization

Output Dimensions (Cinematic Color Parameters)

Dimension	Meaning
R	Red channel
G	Green channel
B	Blue channel
Energy	Visual activity/liveliness (calm ↔ dynamic)
Intensity	Effect prominence (subtle ↔ pronounced)

Notes:

• RGB values create the base color palette
• Energy represents how 'alive' or 'active' the visual should feel - independent of the colors themselves
• Intensity controls how strongly the treatment is applied - high energy can be displayed subtly, or low energy can be pronounced

---

Contents

~80 curated mappings spanning:

• Emotional states (joy, rage, meditation, anxiety, awe, grief, etc.)
• Aesthetic qualities (sunset warmth, forest calm, storm energy, clinical detachment)
• Cinematic references (film color grading, lighting moods, production design)
• Geometric anchors that define boundaries of the space

The source field documents the visual or cultural inspiration behind each mapping.

---

Limitations

• Small scale (~80 mappings): useful as anchors/references, not comprehensive coverage
• Culturally specific: primarily draws from Western cinematic tradition
• Interpretive: mappings reflect observed patterns in film/photography, not objective measurements
• Output parameters are descriptive rather than rigidly standardized across tools
• Designed as a reference structure; practical utility will vary by application

---

Files

• train.json — the manifold mappings

---

Usage

from datasets import load_dataset

dataset = load_dataset("danielritchie/cinematic-mood-palette")

sample = dataset['train'][0]
print(sample['name'], sample['input'], sample['output'])

---

In One Line

A reference map from psychological affect space to cinematic color language.

---

Citation

@dataset{cinematic_mood_palette,
  title={Cinematic Mood Palette},
  author={[Daniel Ritchie]},
  year={2026},
  publisher={Hugging Face}
}