Program call parameters:

  ccolor -cp 30 -im 1 -hs 3 -ap 1 -cc 64 -g 1 -cl 100 -sm 01 
         -ih $bhp/ill_ap01_im01_sp01_cp26_sm01.hist 
         -sh $bhp/surfb_ap01_im01_sp01_cp26_sm01_sm01.hist 
         -ti $1_illum.ppm $1_obj.ppm $outfile

Explanation:

  -cp 30
        = illuminant estimation algorithm
          30 = bayesian em, same as 29, but uses independent bootstrap 
               samples (probably an invalid assumption)
          29 = bayesian em, must specify histogram files
  -im 1 
        = mode for selecting initial variances for EM 
  -hs 3 
        = m-estimate for histogram, m = 0.001
  -ap 1 
        = better approximation mode for cached statistics based EM 
  -cc 64 
        = use 64 components in the mixture model
  -g 1 
        = image data has a gamma of 1.0
  -cl 100 
        = clip the image data at a brightness of 100
  -sm 1 
        = EM approximation mode, 1d projection mode summary, 1024 bins
  -ih $bhp/ill_ap01_im01_sp01_cp26_sm01.hist 
        = the illumination color prior (histogram) can be found in the file
  -sh $bhp/surfb_ap01_im01_sp01_cp26_sm01_sm01.hist 
        = the surface color prior (histogram) can be found in the file
  -ti $1_illum.ppm 
        = the true target illumination can be found in this file

These settings are equivalent to my script results "64a".