Computational analysis of the photograph of the artwork F# minor - Research on Harmony (AQC0602) [1] by Arnaud Quercy [2], according to IDS-CMP-2025 [3], on an image 1000 pixels on its longer side, on 2025-10-03. Each colour is located in the Munsell system and named according to the ISCC-NBS system (267 names, National Bureau of Standards, 1976). The image divides into 12 flat areas. The largest are light bluish green (53%), moderate bluish green (25%) and very dark green (10%). Among the others, those that stand out most sharply from their surroundings are moderate yellow in the centre (0.5%), deep yellow at the bottom left (2.5%) and moderate yellow in the centre (5.3%). By colour name family: green 90% and yellow 6.3%.
Context
Analysis performed according to IDS-CMP-2025 [3]: (a) colours named by Munsell and ISCC-NBS, areas and flat areas, (b) texture, (c) brightness and contrast, (d) spatial distribution. Source image: 2081x2775 pixels. Analysis date: 2025-10-03.
A) Color Analysis
Flat areas
Ranked by contrast with their surroundings (CIEDE2000). Strokes thinner than 10 pixels are excluded (3).
| Rank |
Colour |
ISCC-NBS name |
Munsell |
Area |
Contrast |
Position |
| 1 |
D0AD53 |
moderate yellow |
2.3Y 7.1/7.4 |
0.5% |
49.6 |
in the centre |
| 2 |
0E2A24 |
very dark green |
8.4G 1.5/2.9 |
10% |
37.3 |
on the left |
| 3 |
B1801D |
deep yellow |
1.3Y 5.7/8.7 |
2.5% |
30.1 |
at the bottom left |
| 4 |
DBC675 |
moderate yellow |
4.4Y 7.9/6.0 |
5.3% |
27.0 |
in the centre |
| 5 |
C2C6B5 |
yellowish gray |
0.9GY 7.8/1.0 |
0.6% |
26.3 |
at the bottom |
| 6 |
E5E7D2 |
yellowish white |
1.1GY 9.0/0.9 |
0.6% |
24.1 |
at the top |
| 7 |
59BBA6 |
light bluish green |
0.8BG 6.9/6.4 |
53% |
15.6 |
on the right |
| 8 |
1F817E |
moderate bluish green |
4.4BG 4.9/6.0 |
25% |
12.2 |
at the bottom left |
| 9 |
398F7E |
moderate bluish green |
1.6BG 5.2/5.4 |
2.4% |
4.3 |
at the bottom right |
| 10 |
359482 |
moderate bluish green |
2.5BG 5.5/6.0 |
0.1% |
3.2 |
on the right |
| 11 |
4E9584 |
light bluish green |
1.5BG 5.5/4.9 |
0.1% |
3.0 |
at the top left |
| 12 |
3D937F |
moderate bluish green |
9.1G 5.5/6.0 |
0.1% |
3.0 |
at the top |
Colour name families
| Family |
Area |
| green |
90% |
| yellow |
6.3% |
| orange yellow |
0.8% |
| brown |
0.7% |
| yellow green |
0.6% |
| gray |
0.5% |
B) Texture Analysis
| Metric |
Value |
| Global Roughness |
0.181 |
| Mean Local Roughness |
0.026 |
| Roughness Uniformity |
0.021 |
| Edge Density |
0.141 |
| Mean Gradient Magnitude |
0.208 |
| Gradient Variance |
0.055 |
| Gradient Smoothness |
0.0 |
| Directional Coherence |
0.016 |
| Pattern Complexity |
0.142 |
| Pattern Repetition |
1.0 |
| Detail Frequency Ratio |
0.651 |
| Spatial Variation |
0.124 |
| Texture Consistency |
0.436 |
C) Brightness & Contrast Analysis
| Metric |
Value |
| Mean Brightness |
0.511 |
| Brightness Variance |
0.181 |
| Brightness Uniformity |
0.646 |
| Brightness Skewness |
-0.606 |
| Brightness Entropy |
7.317 |
| Rms Contrast |
0.181 |
| Michelson Contrast |
1.0 |
| Weber Contrast |
0.713 |
| Mean Local Contrast |
0.029 |
| Contrast Uniformity |
0.245 |
| Dynamic Range |
1.0 |
| Effective Dynamic Range |
0.639 |
| Shadow Percentage |
12.136 |
| Midtone Percentage |
69.871 |
| Highlight Percentage |
17.994 |
| Shadow Clipping |
0.013 |
| Highlight Clipping |
0.0 |
| Tonal Balance |
0.048 |
| Fine Contrast |
0.014 |
| Medium Contrast |
0.036 |
| Coarse Contrast |
0.048 |
| Multiscale Contrast Ratio |
0.286 |
| Edge Contrast |
0.208 |
| Contrast Clustering |
0.564 |
D) Spatial Distribution Analysis
| Metric |
Value |
| Spatial Coherence |
0.77 |
| Color Clustering |
0.496 |
| Color Transition Smoothness |
0.49 |
| Transition Uniformity |
0.656 |
| Sharp Transition Ratio |
0.1 |
| Transition Directionality |
0.019 |
| Mean Saturation |
0.561 |
| Saturation Variance |
0.027 |
| Low Saturation Ratio |
0.043 |
| Medium Saturation Ratio |
0.75 |
| High Saturation Ratio |
0.207 |
| Saturation Clustering |
0.999 |
| Hue Concentration |
0.863 |
| Complementary Balance |
0.003 |
| Analogous Dominance |
0.913 |
| Temperature Bias |
-0.83 |
Methodology
Computational analysis of a photograph of the artwork, reduced to 1,000 pixels on its longer side and filtered by a 3×3 median to remove grain. Colours: each pixel (sRGB rounded to 5 bits per channel) receives its Munsell notation and ISCC-NBS name from a table computed once from published data (Munsell renotation; ISCC-NBS blocks, NBS Special Publication 440, 1976), never from the artworks. Flat areas: connected regions of one name, merged with their neighbours while the CIEDE2000 difference between their mean colours stays below 10; an area is kept from 0.05% of the image; contrast with the surroundings measured 4 to 12 pixels from the edge. Texture: intensity standard deviation, local standard deviation, edge density (Canny), local binary patterns. Brightness: mean, variance, RMS, Michelson and local contrast, shadow / midtone / highlight distribution. Distribution: spatial coherence, transitions, saturation. Methods are deterministic and reproducible. Limits: the measurement concerns the photograph, including its lighting and white balance; two neighbouring areas less than 10 apart are merged.
