ARGIRA Synthetic Controlled Benchmark v2 — Emergent Harmonic Count, Fractal Dimension and Spatial Irregularity in Image Sonification

This dataset was designed to isolate independent visual variables under controlled synthetic conditions. ARGIRA Synthetic Controlled Benchmark v2 contains 10 artificially generated images, their corresponding audio outputs (Pipeline 09, sine additive), extracted metrics, and a formal hypothesis document. The benchmark provides reproducible evidence that structured chromatic distributions preserve harmonic richness better than stochastic distributions despite equivalent hue variance. The central finding is the controlled contrast between images 04 (random_noise, stochastic_texture) and 05 (rainbow_gradient, structured_gradient): identical hue_std = 0.5000, opposite saturation and spatial structure, completely different acoustic profiles (6 vs 13 harmonics). This demonstrates that hue_std alone does not predict acoustic complexity saturation and spatial irregularity are independent dimensions. Formal model: A = f(H, S, I) — acoustic complexity as a function of chromatic dispersion, mean saturation, and spatial irregularity.Files: 10 PNG images · 10 WAV audio outputs · CSV master file (14 columns) · sonification pipeline · hypothesis PDF · README · 2 publication-ready figures (v2) v2 update: two publication-ready figures added. (1) Main benchmark figure four panels: Emergent Harmonic Count, hue_std vs Saturation, Spatial Irregularity, and Scatter Irregularity × Emergent Harmonic Count with annotated contrast between images 04 and 05. (2) Acoustic Complexity Metrics figure Emergent Harmonic Count and Fractal Dimension per image, with baseline reference at 1.0. Related: DOI 10.5281/zenodo.20364540 (ICAD 2026 preprint) · DOI 10.5281/zenodo.20366726 (Pipelines 10–13)