Datasets / Asano_2015_Observer_CMFs

Asano (2015) Observer Function Database — 151 individual + 1000 Monte Carlo observer CMFs

observer color-matching functions (CMFs) and cone fundamentals — 151 color-normal individuals

observer-data

Description

Yuta Asano's (2015) database of color-matching functions and cone fundamentals from 151 individual color-normal observers measured at the RIT Munsell Color Science Laboratory, plus 10 categorical observer averages stratified by age/field-of-view, plus 1000 Monte-Carlo-simulated observers spanning the natural variation envelope. Each observer has full CMFs (x̄, ȳ, z̄) and LMS cone fundamentals at both 2° and 10° fields, sampled at 1 nm. Useful for two questions: (1) how much does inter-observer variation shift the position of any predicted ΔE? (2) what observer subgroups are over- or under-represented in CIE 1931 / CIE 1964 standard observers? Important for accessibility work — color-vision-deficient observers are not in this set, but the natural variation among normal observers is itself ±5–10 nm in cone peak λ, which affects ΔE predictions at low chroma.

asano_observers.zip
Contains canonical.csv + meta.yaml + raw/
6.8 MB Download

Notes

  • ★ This is OBSERVER CHARACTERIZATION data, not color-difference data. It tells you 'observer N matches color X with these primaries', not 'observer N rated this difference at level Y'.
  • Each Excel file has 5 sheets: xyz-02deg, xyz-10deg, lms-02deg, lms-10deg, parameters. Read directly from raw/ — `canonical.csv` is just a file index.
  • Use case for color-space work: compute an 'observer-uncertainty envelope' around any predicted ΔE by running it through all 1000 Monte Carlo observers and reporting the IQR. Essential for showing that a champion's STRESS advantage is not an artifact of CIE 1931 standard observer choice.
  • 151 individuals are a real demographic sample; 1000 Monte Carlo are model-generated to span the natural envelope; 10 categorical are population means.
  • Original distribution: rit-mcsl.org/StudentResearch/YutaAsano/ (mirrored to Zenodo DOI 10.5281/zenodo.3252742).

Sample rows

First 5 rows of canonical.csv:

filenamesize_bytescategorylines
Data_1000MonteCarloObs.xls4278272254 Monte Carlo observers (simulated; file named 1000 but BIFF8 256-column cap truncates to 254)
Data_10CatObs.xls28160010 categorical observers (age × field-size groupings)
Data_151Obs.xls2716160151 individual observers (CMFs + cone fundamentals, 2° + 10°)
urls.txt390source URLs + checksums

Raw files preserved

FileSize
raw/Data_1000MonteCarloObs.xls4.1 MB
raw/Data_10CatObs.xls275.0 KB
raw/Data_151Obs.xls2.6 MB
raw/urls.txt390 B

Source publications

Asano, Y. (2015). Individual Colorimetric Observers for Personalized Color Imaging. PhD dissertation, Rochester Institute of Technology.
Asano, Y., Fairchild, M. D. & Blondé, L. (2016). Individual Colorimetric Observer Model. PLoS ONE 11(2): e0145671.
Snapshot: Zenodo DOI 10.5281/zenodo.3252742 (mirrored from rit-mcsl.org)