Skip to content

horwitzlab/spect-sens-diffs

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Spectral sensitivity differences between rhesus monkeys and humans

This repository contains data used in the manuscript by Zachary Lindbloom-Brown, Leah J. Tait, Gregory D. Horwitz entitled "Spectral sensitivity differences between rhesus monkeys and humans: implications for neurophysiology", Journal of Neurophysiology, Dec 2014, 112 (12) 3164-3172; DOI: 10.1152/jn.00356.2014.

Here we provide three MATLAB .mat files that contain the data from three experiments presented in the manuscript.

Experiments 1 and 2 - detection of 15 Hz modulations of the green and blue phosphors or of all phosphors

  • monitor_spectra - Each column is the spectral power distribution of the red, green, and blue color channels (respectively, each 81 elements long) of the displays used in our experiments. Because we calibrated our displays periodically, each trial in trials.data (below) contains a column index into this matrix to indicate the calibration used at the time of data collection.
  • cone_sensitivities - The Stockman, MacLeod, and Johnson 10-degree human cone fundamentals that we use to compute cone contrasts.
  • background_cc - The background of our display in human cone contrast units. Just like monitor_spectra above, each column is indexed (in trials.data) to reflect the background color at the time of data collection.
  • wavelengths - The wavelength sampling lattice of monitor_spectra and cone_sensitivities.
  • trials - a struct array with two fields: 1) subject is the identity of the psychophysical observer, and 2) data houses the trial-by-trial data of each subject as a matrix. The columns of data are:
    1. the L-cone contrast;
    2. the M-cone contrast;
    3. the S-cone contrast;
    4. the retinal eccentricity of the stimulus in degrees (negative values indicate presentation to the left of the fixation spot);
    5. true if the observer detected the stimulus and false otherwise;
    6. a column index into monitor_spectra;
    7. a column index into background_cc.

Experiment 3 - low frequency scotopic detection

  • monitor_spectra - This has the same layout as above and was measured without the application of neutral density filters;
  • scotopic_lum_eff - The CIE scotopic luminous efficiency function;
  • nd_transmittance - The nominal transmittance of a 1.0-log unit Kodak Wratten 96 neutral density filter (digitized from Kodak's documentation);
  • wavelengths - The wavelength sampling lattice of monitor_spectra, scotopic_lum_eff, and nd_transmittance;
  • trials - Like the first two experiments, this struct array contains trial-by-trial data from each subject. The columns of data are:
    1. the scotopic luminance of the stimulus;
    2. the retinal eccentricity of the stimulus in degrees (negative values indicate presentation to the left of the fixation spot);
    3. true if the observer detected the stimulus and false otherwise;
    4. 1 for a green channel modulation or 2 for the blue channel;
    5. a column index into monitor_spectra.

About

Data from Lindbloom-Brown et al. (2014) "Spectral sensitivity differences between rhesus monkeys and humans". J. Neurophysiol. 112:3164–3172. DOI:10.1152/jn.00356.2014

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published