Datasets / MacAdam_1942_Ellipses
MacAdam (1942) discrimination ellipses (Observer PGN)
JND threshold ellipses
Description
The classic 25 just-noticeable-difference (JND) chromaticity ellipses measured by David MacAdam in 1942 under daylight illumination, using a single highly trained observer (Perley G. Nutting Jr., 'PGN'). Each ellipse maps the locus of color matches around a fixed center — the inside of the ellipse is the threshold of perceptual indistinguishability. Foundational data for color discrimination thresholds, the inspiration for the entire 'perceptually uniform color space' research line, and the reason CIELAB and CIELUV were developed.
Notes
- ★ This is the REAL JND dataset — not to be confused with MacAdam (1974) OSA UCS which is suprathreshold. Many papers cite '1942' but use '1974' data, or vice versa. Always check.
- Single observer (PGN). MacAdam later replicated with 2 more observers in 1949 but the 1942 PGN data is the canonical set used in every textbook.
- Ellipse semi-axes (a, b) are stored in CIE chromaticity units, NOT the source paper's ×10³ convention — divide-by-1000 already applied in canonical.csv.
- Illuminant C (Munsell-era daylight standard); apply chromatic adaptation if using under D65.
Sample rows
First 5 rows of canonical.csv:
| center_id | x_c | y_c | a | b | theta_deg | illuminant |
|---|---|---|---|---|---|---|
| macadam1942_01 | 0.16 | 0.057 | 0.00085 | 0.00035 | 62.5 | C |
| macadam1942_02 | 0.187 | 0.118 | 0.0022 | 0.00055 | 77.0 | C |
| macadam1942_03 | 0.253 | 0.125 | 0.0025 | 0.0005 | 55.5 | C |
| macadam1942_04 | 0.15 | 0.68 | 0.0096 | 0.0023 | 105.0 | C |
| macadam1942_05 | 0.131 | 0.521 | 0.0047 | 0.002 | 112.5 | C |
Source publications
MacAdam, D. L. (1942). Visual Sensitivities to Color Differences in Daylight. Journal of the Optical Society of America 32(5), 247–274.
Wyszecki, G. & Stiles, W. S. (2000). Color Science: Concepts and Methods, Quantitative Data and Formulae, Table 2(5.4.1). Wiley.
Snapshot: colour.DATA_MACADAM_1942_ELLIPSES (BSD-3 licensed source data)