Makee: HIRES Reduction Cookbook (with examples) (as of December 2000)

Reducing a single observation with makee.

Reducing an entire observing run (multiple reductions) using makeepipe.

Reducing a single observation with makee.

Before running makee you must select your four input images: (1) object , (2) trace star , (3) flat field , and (4) arc lamp . The arclamp is optional (you just won't get a wavelength scale.) It is also recommended that you specify a short dark image . See the makee syntax for more details. Also see ideal input files for MAKEE.

Select your input files by referring to your logsheets and/or a listing of your HIRES exposures Below is an example listing using rhh :

  78  hires0078   0.2600 -0.6370  RED  quart  ug5     12s  quartz
  90  hires0090   0.2600 -0.6375  RED  none   clear  300s  BD+284211
  91  hires0091   0.2600 -0.6375  RED  none   clear 3000s  Q2206-199N
  92  hires0092   0.2599 -0.6375  RED  ThAr1  clear    1s  thar

This listing shows the observation number, filename (.fits is assumed), the ECHANGL value, XDANGL, cross disperser type (RED or UV), lamp name, lamp filter, exposure, and frame name.

In this example, we want to reduce observation #91, a 3000 second exposure of the quasar 2206-199. We will use the UV bright standard star BD+28 4211 (observation #90) to define the echelle trace curve.

The MAKEE command for this would be:

makee hires0091 hires0090 hires0078 hires0092 log=hires0091.log

Here we used a single quartz flat, although a sum of flats is usually recommended. We also used one arclamp, although only one was required. We have included the option "log=" to send the output to a log file, alternatively we could have had the output sent to screen (default).

First, look at the log file for this reduction, hires0091.log . Look for error and warning messages or excessive cosmic ray rejections (more than 300 in the object, or 2000 in the sky background).

Next, look at the trace PostScript file, . This shows the trace curve subtracted from a straight line through the end points of the curve. The deviations away from the fit should be less than about 0.2 pixels with larger residuals perhaps at the edges of some orders.

Then look at the profiles PostScript file, . The dotted lines show which regions were extracted as object and background. If the automatically selected regions are inadequate, you can manually select the object and background regions (see Controlling the Object Extraction Boundaries ).

Then look at the wavelength calibration PostScript file, . Each plot shows the residuals in the wavelength calibration polynomial fits to line identifications in each echelle order. The vertical range of the plot is fixed at about +/-0.4 pixels. The residuals should fall within about 0.2 pixels in most cases. The dotted curve shows points derived from an iterpolation (or extrapolation) of the polynomial fits to adjacent orders. These points allow for reliable wavelength calibration in regions of an echelle order with little or no real line identifications.

The second-to-last two plot in shows the RMS and high residuals in each echelle order. The solid line shows the RMS values which should all lay below the dashed line (suggested limit). The last plot shows the number of points used in each order. Non-partial orders (first or last orders) usually contain between 20-40 points if only one arclamp was used.

If those plots look ok, we can view the Flux-091.fits and Err-091.fits with spim0 or xplot . You can also view the PostScript file produced by MAKEE. The file shows the same data in digital number units (not flat field corrected). The file shows the signal-to-noise per pixel of the data.

Reducing an entire HIRES run data set using makeepipe.

The program makeepipe can be used to create a script of "makee" commands. In the example below, we copy all of the raw HIRES images into a sub-directory (called "raw") of the working directory. The following steps are executed from the working directory:

  1. Create a listing of all HIRES raw image files:
         cd raw; ls hires*.fits > ../file.list; cd ..

  2. Run the first step of makeepipe:
         makeepipe file.list step=1 raw=raw

  3. You will see a few listing files starting with "image.list*". The file "image.list" is the one that is read by makeepipe. Here is an example: image.list . This run contains bias exposures, although these are not required. Edit image.list and delete any lines for files you wish to ignore. (Save a copy of the result, since if you re-run step=1 you may accidently re-write the image.list file.)
         vi image.list

  4. The next step is optional. Step number 2 will attempt to determine the electrons per digital number and readout noise using the flat fields and bias frames. (Note that makeepipe only reads "file.list" in step #1, the other steps read "image.list", but we still include "file.list" on the command line):
         makeepipe file.list step=2 raw=raw
    makeepipe reports the results of the "eperdn" and "ronoise" calculations. makeepipe does not automatically use these values. If these look ok, then ignore the output, otherwise you may want to include the "eperdn=" and "ronoise=" options on the following makeepipe command lines.

  5. Step 3 is also optional. This step sums up the bias frames you may or may not have and are not required:
         makeepipe file.list step=3 raw=raw
    You will see a file such as "bias1d_1x2.fits" which will be used by MAKEE.

  6. Step 4 sums up the flat fields. This is required, even if you have only one flat exposure in a setup:
         makeepipe file.list step=4 raw=raw
    Note the files such as "flat_0106.fits" which will have been created and added to the "image.list" file.

  7. Step 5 writes the "makee.script" script file. If you want log files (a good idea) include the "-log" option at this stage. If you look at the latest image.list file, image.list , you may notice that observation #128, has a different cross disperser angle (xdang = +1.9420) then the flat field it goes with (flat_0161.fits has xdang=+1.9395). We can still have it accepted by increasing the angle tolerance with "xdtol=":
         makeepipe file.list step=5 raw=raw -log xdtol=0.003

  8. Now you can look at the makee.script file to see if it makes sense. You can run step #6, which just executes the script, or you can run it by hand:
         source makee.script &
    Note that all of the above steps could have been executed by one step:
         makeepipe file.list raw=raw -log xdtol=0.003
    But it usually better to run it one step at a time to edit the image.list file and verify that each step runs properly.

  9. Now we can look at the output. By default, a ".log" file will be created for each reduction. You can check the size of the ".log" files for unusual reductions, and also search for errors and warnings:
         egrep -i error *.log
         egrep -i warning *.log

  10. For each observation, check the PostScript plot output as described above for single object extractions. Note that the "" and "" files are labelled using the observation numbers of the trace stars and arclamps. In particular, if any of the ranges shown in "" files look bad, you may need to re-run makee and manually control the object and background ranges (see Controlling the Object Extraction Boundaries ).

  11. Run logsum to create a tabular and PostScript summary of cosmic ray rejections given in the ".log" files:
         logsum all

Back to top.

Back to makee home page .