ARGIRA Spatial Irregularity Gradient Experiment v1 — Controlled Spatial Perturbation Series

This dataset documents a controlled spatial perturbation experiment designed to isolate spatial irregularity (I) as an independent variable in the ARGIRA image sonification model A = f(H, S, I). The experiment was designed as a continuous extension of the discrete ARGIRA Synthetic Controlled Benchmark v1. 21 synthetic images were generated from a structured rainbow gradient base by applying progressively increasing Gaussian spatial displacement in coordinate space — approximately preserving the global chromatic histogram while increasing spatial disorder continuously across levels I00–I20. Three visual metrics were extracted per level: spatial irregularity, Shannon entropy (normalized /6), and hue_std (normalized ×1.4). Key findings: — Irregularity follows a non-linear growth curve with a saturation plateau at I16–I18, suggesting a bounded regime of spatial disorder. — Entropy decreases monotonically as irregularity increases — spatial chaos ≠ statistical chaos. — Mean saturation remains constant across all 21 levels (S = 1.0000), while hue_std remains approximately stable, validating controlled spatial perturbation independent from chromatic intensity. Acoustic metrics (Emergent Harmonic Count, Fractal Dimension) will be incorporated in a future extension following full audio-series rendering. Files: master CSV · generation script (Python) · figure PNG · README · 21 PNG images will be added in v2 following full local rendering. Related: DOI 10.5281/zenodo.20385068 (Synthetic Benchmark v1/v2) · DOI 10.5281/zenodo.20364540 (ICAD 2026 preprint)