tab - April-October 2000. .......................... The HIRES reduction program "makee" has been modified to extract ESI spectra. There are a few differences between HIRES and ESI. With ESI: -- Orders run down columns (not rows) in the raw images. -- The setup is fairly fixed compared to HIRES. The same orders appear in roughly the same place in all exposures, although the object may move within the slit. -- The orders are much longer in pixels (~4000 pixels with 1x1). -- The orders are more curved. -- The bad columns run roughly in the dispersion direction (although there are fewer bad areas and no "ink spot"). -- There are fewer orders than in most HIRES exposures. -- There are larger gaps in the arc lamp exposures with no good lines. .......................... Changes for ESI: Global variables Global_HIRES and Global_ESI are now used throughout makee to separate code differences. Some code and data file differences with ESI include: -- In the (internal) pre-processing step (see subroutine BIW) the images are rotated and flipped, to blue end of the bluest order near the 0,0 pixel position. (No rotation or flipping with HIRES). (This effect should not be noticeable to the user.) -- There are two separate parameter files are used: makee.ESI.param and makee.HIRES.param. -- A higher polynomial fitting order is used for tracing echelle orders (15 for ESI and 4 for HIRES, see makee.*.param). -- In the star (or trace) image, the bad columns are left alone in HIRES, but are masked with ESI data. -- For the object profile and object centroid shift procedures (see the profiles-###.ps file), the first two (bluest) orders are ignored for ESI data. -- The arclamp calibration for ESI is done by correlating and aligning the given arc exposures with previously calibrated arc exposures. The accuracy should be less than or about +/- 0.2 pixels assuming the user provides CuAr (~300-600s) and HgNeXe (~2-5s) exposures. ..........................