Looks like stepsize plays no role in err=2 sigmas
Following is an old one with original sigmas, much too small - set to sigI
Clean this up if decided to go with kludged sigF input.
  For err=2 mode=1
  	  0.104       2.036       2.179      0.1429         0.1       2.043
      0.207       1.198       1.354      0.1565      0.0193       65.77
      0.311      0.9409      0.9673     0.02643      0.0127        4.33
      0.415       1.245       1.379      0.1339      0.0148       81.91
      0.519      0.4568      0.4559  -0.0009956       0.005     0.03965
      0.622      0.1763      0.1603    -0.01595       0.005       10.18
      0.726      0.4306      0.3831    -0.04751       0.005        90.3
       0.83           0     0.03927     0.03927       0.005       61.69
      0.934      0.5835      0.5609    -0.02259       0.005       20.41
      1.037      0.1288     0.09778    -0.03104       0.005       38.54
      1.141      0.1371     0.09589    -0.04121       0.005       67.94
      1.245      0.3157      0.3485     0.03282       0.005       43.08
 chi2red = 48.62   sigma^2							input values