CD014 High-medium slit resolutions at 17.3um

Report on getting the high-res slit feature resolutions at the 580.5 cm-1 setting (17.3um).
The resolutions are surprisingly high, so I'm showing all my assumptions and methods here.

The data is from  /CD014/cd14_2013mar13/file17.fits-file22.fits

Instrument settings:
filter: 18a + KBr
V_decker=2.1614 V to maximize slit length
V_echelle=8.733 to center the 3 H2O lines in columns 200-300
pumping continuously on gas cell
33 ctdelay; 100 1 egos
Subtracted frame 3 from frame 0 for all data.

There are 3 strong H2O lines in this setting:
580.53650 cm-1
580.73199 cm-1
581.08783 cm-1

These lines appeared in 2 echelon orders, with the 580.7 cm-1 line having another order partially recorded at the bottom of an echelon order.

high-medium spectrum at 580.5 cm-1, with 3 H2O lines

I removed some odd/even by subtracting the median of each row from the image, end then fit vertical columns over the lines with 6-term Gaussians.
I corrected for the projection effect of the slant by multiplying by cos(angle-to-horizontal).
Angles for lines 580.54: -25.7deg   580.7:-24.8deg  581.1: -29.3deg

The reported FWHM are median(sigma*2.355*cos(ang))+/-stddev(the same)-- where sigma is the array of Gaussian widths across the lines.

filename  slit#    FWHM_580.54    FWHM_580.73    FWHM_581.09
file22    slit1    7.75+/-0.18    6.93+/-0.14    8.07+/-0.20
file21    slit2    7.92+/-0.20    7.21+/-0.18    8.33+/-0.17
file20    slit3    8.39+/-0.23    7.73+/-0.18    8.90+/-0.22
file19    slit4    9.38+/-0.24    8.98+/-0.24    10.21+/-0.22
file18    slit5    10.73+/-0.29   10.40+/-0.25   11.73+/-0.28
file17    slit6    12.73+/-0.31   12.71+/-1.65   14.11+/-0.33

To obtain the wavelength solutions, I fit lines vertical centers of the H2O lines (including the second 580.73cm-1), and I fit lines to the echelon order gaps. I took the intersection point of these lines and calculated the distance between them. For the 581.1-580.73 span, this was 507.4 pix, for 0.3558 cm-1
or 0.0007013 cm-1 per pixel @ 581.1cm-1 or 0.3619 km/s per pixel. For the echelon order with 580.54 and 580.73, 0.4036 km/s.

Illustration of how I found the wavelength scale(s)

Resolutions at each line
filename  slit#   R_580.54  R_580.73  R_581.09
file22    slit1    95800    119500    102700
file21    slit2    93800    114900    99400
file20    slit3    88500    107200    93100
file19    slit4    79200    92200     81200
file18    slit5    69200    79700     70600
file17    slit6    58400    65200     58700

The lower resolution of the 580.5 line is mostly due to the larger km/s per pixel value for that right echelon order. The difference in resolutions in 580.7 and 581.09 in the left echelon order is mostly due to the better focus position of the 580.7 cm-1.