Argira: A Reproducible, Structure-Preserving Sonification Pipeline for Visual Art – Validation with r = 0.9289, Sobel Cross-validation, and a Documented Negative Result (2D FFT)

[EN] We present Argira, a deterministic sonification pipeline that translates the visual texture of paintings into sound while preserving measurable statistical properties. Using a dataset of 47 public-domain artworks (15 artists), we compute the Visual Texture Irregularity (VTI) the mean absolute difference between neighbouring pixels and map it linearly to a sonic frequency: F = 200 + VTI × 800 Hz. The resulting auditory representation is analysed via its spectral fractal dimension Ds. The correlation between VTI and Ds is r = 0.9289 (R² = 0.8629, p < 0.001, permutation test N = 1000, seed = 42). An independent edge-detection method (Sobel filter) confirms the same complexity ranking with r = 0.9570 and 0/49 divergences >2%, ruling out algorithmic artefacts. A 2D FFT experiment produced no differential signal; this negative result is documented for methodological transparency. The pipeline is openly available (CC BY-NC 4.0) and includes an interactive museum and a standalone web-based analyser. Argira demonstrates that structural properties of visual art can be preserved in the auditory domain without arbitrary cross-modal mappings, with direct applications in accessibility for blind and visually impaired audiences. RELATED INTERACTIVE TOOLS (supplementary to the paper): - Interactive sonified museum (listen to the complexity gradient from Malevich to Kandinsky): https://argira.eus/argira-sonification/ - Standalone scientific analyser (upload any image, compute 6 visual metrics and its acoustic signature): https://argira.eus/argira-analizador/ These tools are open access under CC BY-NC 4.0 and directly implement the validated pipeline described in this paper. [ES] Presentamos Argira, un pipeline de sonificación determinista que traduce la textura visual de pinturas a sonido preservando propiedades estadísticas medibles. Usando un dataset de 47 obras de dominio público (15 artistas), calculamos la Irregularidad de Textura Visual (VTI) la diferencia absoluta media entre píxeles vecinos y la mapeamos linealmente a una frecuencia sonora: F = 200 + VTI × 800 Hz. La representación auditiva resultante se analiza mediante su dimensión fractal espectral Ds. La correlación entre VTI y Ds es r = 0.9289 (R² = 0.8629, p < 0.001, test de permutaciones N = 1000, semilla = 42). Un método independiente de detección de bordes (filtro Sobel) confirma el mismo ranking de complejidad con r = 0.9570 y 0/49 divergencias >2%, descartando artefactos algorítmicos. Un experimento con FFT 2D no produjo señal diferencial; este resultado negativo se documenta por transparencia metodológica. El pipeline está disponible en acceso abierto (CC BY-NC 4.0) e incluye un museo interactivo y un analizador web autónomo. Argira demuestra que las propiedades estructurales del arte visual pueden preservarse en el dominio auditivo sin mapeos arbitrarios, con aplicaciones directas en accesibilidad para personas ciegas o con discapacidad visual. HERRAMIENTAS INTERACTIVAS RELACIONADAS (complementarias al artículo): - Museo sonoro interactivo (escucha el gradiente de complejidad desde Malevich hasta Kandinsky): https://argira.eus/argira-sonification/ - Analizador científico autónomo (sube cualquier imagen, obtén 6 métricas visuales y su firma acústica): https://argira.eus/argira-analizador/ Estas herramientas son de acceso abierto bajo licencia CC BY-NC 4.0 e implementan directamente el pipeline validado descrito en este artículo. IMPORTANT LICENSE NOTE: Due to a technical limitation of the Zenodo submission interface, the license shown here as "CC BY 4.0" is incorrect. The actual license of this work is CC BY-NC 4.0 (Creative Commons Attribution-NonCommercial 4.0 International), as stated in the PDF file and in all related documentation. The PDF, code, and dataset are released under CC BY-NC 4.0.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-05-11
DOI
https://doi.org/10.5281/zenodo.20123820
Primary Topic
Tactile and Sensory Interactions
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Argira: A Reproducible, Structure-Preserving Sonification Pipeline for Visual Art – Validation with r = 0.9289, Sobel Cross-validation, and a Documented Negative Result (2D FFT)

Jose Ranero García
Zenodo (CERN European Organization for Nuclear Research)
Tactile and Sensory Interactions
article

Argira: A Reproducible, Structure-Preserving Sonification Pipeline for Visual Art – Validation with r = 0.9289, Sobel Cross-validation, and a Documented Negative Result (2D FFT)

Jose Ranero García
article en

Abstract

[EN] We present Argira, a deterministic sonification pipeline that translates the visual texture of paintings into sound while preserving measurable statistical properties. Using a dataset of 47 public-domain artworks (15 artists), we compute the Visual Texture Irregularity (VTI) the mean absolute difference between neighbouring pixels and map it linearly to a sonic frequency: F = 200 + VTI × 800 Hz. The resulting auditory representation is analysed via its spectral fractal dimension Ds. The correlation between VTI and Ds is r = 0.9289 (R² = 0.8629, p < 0.001, permutation test N = 1000, seed = 42). An independent edge-detection method (Sobel filter) confirms the same complexity ranking with r = 0.9570 and 0/49 divergences >2%, ruling out algorithmic artefacts. A 2D FFT experiment produced no differential signal; this negative result is documented for methodological transparency. The pipeline is openly available (CC BY-NC 4.0) and includes an interactive museum and a standalone web-based analyser. Argira demonstrates that structural properties of visual art can be preserved in the auditory domain without arbitrary cross-modal mappings, with direct applications in accessibility for blind and visually impaired audiences. RELATED INTERACTIVE TOOLS (supplementary to the paper): - Interactive sonified museum (listen to the complexity gradient from Malevich to Kandinsky): https://argira.eus/argira-sonification/ - Standalone scientific analyser (upload any image, compute 6 visual metrics and its acoustic signature): https://argira.eus/argira-analizador/ These tools are open access under CC BY-NC 4.0 and directly implement the validated pipeline described in this paper. [ES] Presentamos Argira, un pipeline de sonificación determinista que traduce la textura visual de pinturas a sonido preservando propiedades estadísticas medibles. Usando un dataset de 47 obras de dominio público (15 artistas), calculamos la Irregularidad de Textura Visual (VTI) la diferencia absoluta media entre píxeles vecinos y la mapeamos linealmente a una frecuencia sonora: F = 200 + VTI × 800 Hz. La representación auditiva resultante se analiza mediante su dimensión fractal espectral Ds. La correlación entre VTI y Ds es r = 0.9289 (R² = 0.8629, p < 0.001, test de permutaciones N = 1000, semilla = 42). Un método independiente de detección de bordes (filtro Sobel) confirma el mismo ranking de complejidad con r = 0.9570 y 0/49 divergencias >2%, descartando artefactos algorítmicos. Un experimento con FFT 2D no produjo señal diferencial; este resultado negativo se documenta por transparencia metodológica. El pipeline está disponible en acceso abierto (CC BY-NC 4.0) e incluye un museo interactivo y un analizador web autónomo. Argira demuestra que las propiedades estructurales del arte visual pueden preservarse en el dominio auditivo sin mapeos arbitrarios, con aplicaciones directas en accesibilidad para personas ciegas o con discapacidad visual. HERRAMIENTAS INTERACTIVAS RELACIONADAS (complementarias al artículo): - Museo sonoro interactivo (escucha el gradiente de complejidad desde Malevich hasta Kandinsky): https://argira.eus/argira-sonification/ - Analizador científico autónomo (sube cualquier imagen, obtén 6 métricas visuales y su firma acústica): https://argira.eus/argira-analizador/ Estas herramientas son de acceso abierto bajo licencia CC BY-NC 4.0 e implementan directamente el pipeline validado descrito en este artículo. IMPORTANT LICENSE NOTE: Due to a technical limitation of the Zenodo submission interface, the license shown here as "CC BY 4.0" is incorrect. The actual license of this work is CC BY-NC 4.0 (Creative Commons Attribution-NonCommercial 4.0 International), as stated in the PDF file and in all related documentation. The PDF, code, and dataset are released under CC BY-NC 4.0.

Zenodo (CERN European Organization for Nuclear Research)
Openalex Percentile: Top 6%
Tactile and Sensory Interactions
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.