STEPS to reduce ANU 2.3m Echelle data.

1. make lists to make before processing:
alllist - to be overscan corrected and trimmed by ccdproc
biaslist - biases to be combined
flatlist - flats (QI) to be combined
zerolist - images to be zerocorrected with ccdproc (all but biases)
unflatlist - images to be flattened (arcs and objs)
arclist
objlist

2. get rid of (delete) DATASEC and CCDSEC with hedit
   set gain - 2 e/ADU, readout noise = 4.6 electrons (only necessary if lacos_spec is used)

   in headers (wrong):
   BIASSEC = '[1:50,1:1074]'      / Bias section - user specified
   TRIMSEC = '[51:2148,1:1074]'   / Trim section
   set in ccdproc the measured values:
   BIASSEC = '[7:50,2:1024]'
   TRIMSEC = '[55:2100,2:1024]'

3. ccdproc: overscan correct and trim

4. zerocombine: combine biases

5. ccdproc: zero correct all images but the biases

6. flatcombine: combine flats (QI)

7. apflatten: normalize master flat
   before apflatten:
   epar apedit: set width = 20
   epar apfind: minsep = 1
   epar apresize: ylevel = 0.1
   epar apresize: bkg = no

   no to auto finding apertures

8. ccdproc: flatten arcs and objs

9. setjd: arcs and objs
   update (hedit) IMAGETYP to 'comparison' for arcs
   
10. move arcs and objs to a collective directory, 
    update master arclist and objlist in that directory with cat
   
11. doecslit: all the rest

set the aperture reference to the first star (obj) spectrum

order list:
the bottom order goes from 6520-6720

you need three UCLES ThAr pages to cover one order
e.g. orders 85-86-87 are needed for our bottom order
there are a couple of orders that are not possible to identify with the arc atlas
most of them are ok, but a few are simply too different
or, more precisely, that's with UCLES
there is a missing wavelength range that is not covered by the UCLES atlas

ORD  WAVEMIN WAVECENTR WAVEMAX   FSR
        (A)      (A)      (A)     (A)
  65  3440.42  3467.09  3493.76  53.34
  64  3493.84  3521.35  3548.86  55.02
  63  3548.94  3577.33  3605.72  56.78
  62  3605.78  3635.10  3664.41  58.63
  61  3664.47  3694.75  3725.04  60.57
  60  3725.08  3756.38  3787.68  62.61
  59  3787.72  3820.09  3852.46  64.75
  58  3852.48  3885.98  3919.48  67.00
  57  3919.50  3954.18  3988.87  69.37
  56  3988.87  4024.81  4060.74  71.87
  55  4060.74  4097.99  4135.25  74.51
  54  4135.23  4173.88  4212.53  77.29
  53  4212.50  4252.62  4292.74  80.24
  52  4292.71  4334.39  4376.06  83.35
  51  4376.02  4419.35  4462.68  86.65
  50  4462.63  4507.71  4552.78  90.15
  49  4552.73  4599.66  4646.60  93.87
  48  4646.53  4695.44  4744.35  97.82
  47  4744.28  4795.29  4846.31 102.03
  46  4846.22  4899.48  4952.73 106.51
  45  4952.64  5008.29  5063.94 111.30
  44  5063.84  5122.04  5180.25 116.41
  43  5180.14  5241.08  5302.02 121.89
  42  5301.90  5365.78  5429.66 127.76
  41  5429.53  5496.56  5563.59 134.06
  40  5563.46  5633.88  5704.30 140.85
  39  5704.15  5778.23  5852.31 148.16
  38  5852.15  5930.18  6008.21 156.06
  37  6008.04  6090.34  6172.64 164.60
  36  6172.45  6259.39  6346.33 173.87
  35  6346.12  6438.10  6530.07 183.95
  34  6529.85  6627.31  6724.77 194.92

Order 34 is our order 1 and then they go down in wavelength.

change xorder and yorder to get dispersion fit residual rms of < 0.02 A/pix