ARGIRA Station — Emergent Corpus v1.0: Spectral Geometry from Minimal Hue→Frequency Image Sonification

ARGIRA — Emergent Corpus v1.0 is a frozen research corpus and analysis pipeline investigating whether stable spectral organizations can emerge from a minimal image sonification mapping without explicitly programmed harmonic rules. This release should be interpreted as evidence for emergent organization within a global chromatic spectral regime, not as proof of universal physical laws independent of algorithmic implementation. The repository contains three frozen scripts: sonify_emergent_v1.py — emergent sonification pipeline mapping image hue distributions to frequency distributions without explicit harmonic assignment spectral_geometry.py — geometric analysis of the resulting spectral space validation_experiments.py — statistical validation via permutation/shuffle tests and dimensional ablation The system intentionally removes earlier explicit controls (direct harmonic assignment, conditional granular synthesis thresholds, rhythm/BPM assignment, rule-based harmonic biasing). The retained core consists only of hue→frequency mapping, spectral mixture construction, and FFT-based spectral analysis. The current implementation operates on global chromatic distributions (I(x,y) → P(h)) rather than spatial image structure (I(x,y) → S(f,t)). As a consequence, the pipeline detects global spectral states rather than local compositional structure. The corpus contains 14 public-domain paintings (Wikimedia Commons) spanning low-variance and high-variance chromatic regimes, including works by Malevich, Rembrandt, Van Gogh, Velázquez, Monet, and Kandinsky. Main findings: Separable spectral regimes emerge without explicit harmonic programming Statistically significant geometric clustering (92.9% agreement with chromatic regimes, shuffle p < 0.005) Spectral organization is multidimensional: spectral_spread and temporal_stab are independent axes, both necessary Strong correlation between chromatic variance and emergent spectral spread (hue_std ↔ spectral_spread, Spearman r = +0.86, p < 0.001) Ds (fractal dimension) acts as structural invariant, not discriminative variable Validation: independent-column permutation tests (Shuffle B1), label permutation tests (Shuffle B2), and leave-one-out dimensional ablation. Optimal reduced vector identified (6D, silhouette 0.427 → 0.520). Frozen under tag v1.0-corpus-frozen. Intended as reproducible baseline for future work including larger corpora, spatially-aware sonification models, and attractor analysis.