Script → Shot

Transform scripts into shot sequences through emotional rhythm

From dataset to model to narrative engine. Shot Rhythm Engine (PV Radar) with reproducible outputs, Azure MVP app, and research-grade emotion data.

Seed/Hash Reproducible Studio Presets Watermark & SHA — rolling out

Product

Script-to-shot tooling and emotion-driven narrative datasets for research and production.

Shot Rhythm Engine — API

Transform scripts into shot sequences with tags, transitions, and timing. Control via presets, reproduce via hash.

PV Radar — App (Azure MVP)

Narrative engine-based shot rhythm tool. Private access, session logging active.

Benchmark — One-Pager

Golden Set summary, coverage metrics, and scoring methodology.

Datasets

Scene-level emotion VAD curves with symbolic tags, manifests, and complete documentation.

SKU: EMOTION-V1

Emotion Dataset — Core100 + Episode Delta (v1.0)

Scene-level V/A/D curves with labeled centroids, distances, and margins. Includes 3-scene samples, data dictionary, and changelog.

  • • Files: emotion_curve.csv, scenes.jsonl, manifest.json
  • • Format: CSV + JSONL with complete manifest
  • • Language: EN (KR mirrored structure coming)

Sample Visualization

Emotion VAD sample chart — Valence, Arousal, Dominance across scenes
Preview only — from Emotion Dataset v1.0
Dataset Specification (manifest.json)
{
  "tables": {
    "emotion_curve": {
      "fields": [
        {"name": "scene_idx", "type": "integer"},
        {"name": "valence", "type": "number"},
        {"name": "arousal", "type": "number"},
        {"name": "dominance", "type": "number"},
        {"name": "label", "type": "string"},
        {"name": "label_distance", "type": "number"},
        {"name": "margin", "type": "number"}
      ]
    }
  },
  "version": "1.0",
  "updated": "2025-08-17"
}
Sample JSONL (2 lines)
{"scene_id":"S-0001","valence":0.12,"arousal":0.34,"dominance":0.56,"label":"neutral"}
{"scene_id":"S-0002","valence":0.75,"arousal":0.60,"dominance":0.90,"label":"pride"}

Stack

From foundational data to production tooling — every layer tracked with manifests and hash reproducibility.

Dataset Emotion VAD curves, symbolic tags, scene-level annotations
Model PV features, shot-tagging, transition rules
Narrative Engine Shot rhythm planner, presets, seeds
PV Radar (Tool) Shot sequence UI & API

Watermark & SHA manifest — planned/rolling out

App (Azure MVP)

Private access. Session logging active. Watermarking & SHA manifest — planned/rolling out. No redistribution permitted.

Benchmark — One-Pager Preview

Updated: 2025-10-05

Licensing

Tier Rights Model Training Seats Price
Evaluation Internal tests, private demos (30 days) 1 $149 / 30d
Research Non-commercial research △ (pre-approval) 3 $1,490 / year
Commercial In-product usage (limited) 10 $9,900 / year
Enterprise Expanded rights / sublicensing Custom Custom

* Model training on Research tier requires pre-approval. No redistribution of raw data or model weights.

Restrictions: no resale or open redistribution; maintain copyright notice; no re-identification attempts; no redistribution of model weights trained on the data.

Read full license preview
DATA LICENSE (Preview) — JsisOn OÜ

Scope: Scene-level emotion dataset (EMOTION-V1) and Shot Rhythm Engine API.
Grant of rights per purchased Tier.

You may:
• Download, process, and use internally within Tier scope
• Train models as specified in Tier permissions
• Access API endpoints with valid key

You may not:
• Re-sell or re-publish raw data
• Re-distribute model weights trained on data
• Attempt re-identification of source material
• Sublicense without Enterprise tier

Attribution: © JsisOn OÜ. Include LICENSE.txt with internal archives.
Jurisdiction: Estonia. Refunds per policy at /legal.
This preview is informative; executed license PDF governs.

API Key access is the primary licensing path. Direct dataset sales are currently on hold. Full results available under NDA upon request.

Contact

For licensing, collaboration, API access, or demo inquiries:
princess@batalstone.com

Please include your organization name, intended use case, and region.