GEMINI (July2009) Gemini IRAF package v1.10 GEMINI (July2009) Release date: July 28, 2009 This document describes the revisions relative to Gemini IRAF package v1.9.1 The Gemini IRAF package v1.10 is compatible with IRAF 2.14 and IRAF patch release V2.12.2a. This release will run under PyRAF or IRAF CL environments. The binaries distributed with the package were built on Red Hat 9, Solaris 7, and Mac OS 10.4 PPC and Intel. PRIMARY REVISIONS: 1) PyRAF compatibility 2) Near-IR and mid-IR: nstransform split into nsfitcoords and nstransform 3) MIDIR: a) New tasks: msslice, miclean, mipql, mistdflux b) Improved msreduce wavelength calibration 4) GMOS: a) Improved support for GMOS nod-and-shuffle modes b) Improved GMOS variance and data quality calculations c) Added GFEXTRACT IFU single slit extraction feature d) Update overscan subtraction e) No-pre-imaging GMOS MOS mask design software now fully integrated f) Add B600 grating and OIII, He, Ha filters to gmos$data/ files g) Other GMOS improvements - GIFLAT, GSTRANSFORM, GIFRINGE, GFCUBE 5) NIFS: a) Moved tasks NFCKCAL and NFCHECK to GEMTOOLS and renamed CKCAL and CKINPUT respectively. b) Critical update to NFPAD to compensate for recent H-band grating shift. c) Critical bug fix for GEMTOOLS.GEMCUBE (only used in the NIFS package as of this release) d) Remove tasks NFFITCOORDS and NFTRANSFORM as GNIRS.NSFITCOORDS and GNIRS.NSTRANSFORM are used for all instruments. DETAILS OF THE REVISIONS: 1) PyRAF compatibility ------------------ Instruments: All Extensive testing and improvements ensure the Gemini IRAF package works under PyRAF. 2) Near-IR and Mid-IR ------------------ Instruments: NIFS, GNIRS, NIRI, Michelle, T-ReCS The "fitcoords" step that was embedded in NSTRANSFORM has been taken out. A new task, NSFITCOORDS was created for the coordinate fitting step. Now the NSFITCOORDS step is run before NSTRANSFORM. a) New tasks in the GNIRS package - gnirs.nsfitcoords -- takes arc and s-distortion calibration images, which have previously been traced, and fits dispersion and s-distortion surface functions. b) New parameters/functionalities in the GNIRS package gnirs.nstransform: - Parameters removed: * The coordinate fitting parameters: lamptransf, sdisttransf, fl_inter, fl_align, function, xorder, yorder, force gnirs$doc/ - All examples are now updated with calls to both NSFITCOORDS and NSTRANSFORM - All gnirsinfo (ls, xd, and ifu) are updated with NSFITCOORDS descriptions. c) Bug fixes gnirs.nsreduce: Output now sorted 3) MIDIR ------------------ Instruments: Michelle, T-ReCS a) New tasks in MIDIR - midir.msslice -- Produce multiple T-ReCS or Michelle spectra across an extended object, with or without calibration of the slices - midir.miclean -- Remove certain types of pattern noise from Michelle or T-ReCS images. - midir.mipql -- Produce Stokes I, U, and Q images plus four more extensions containing the unpolarized flux, the polarized flux, the % polarization, and the polarization angle in degrees. - midir.mipsf -- Measure FWHM and Strehl value in Michelle or T-ReCS images - midir.mistdflux -- Find the in-band flux density of a mid-IR standard star for either Michelle or T-ReCS b) Improved msreduce wavelength calibration midir.msreduce: - New functionality: * Option to subtract bias from the sky frame before wavelength calibration. Significantly improves wavelength calibration. - New parameters: * fl_skybiassub: subtract bias from sky frame * fl_fitcoords : replaces fl_transform, reflects split NSTRANSFORM - Parameter removed: * fl_transform 4) GMOS ------------------ Instruments: GMOS North and South a) Improved support for GMOS nod-and-shuffle modes. Please read the updated nod-and-shuffle data reduction script example (gm> gmosexamples NS) and the info page (gm> gmosinfospec). (i) New tasks in GMOS - gmos.gnsdark -- process and combine GMOS dark images (ii) New functionalities/parameters gmos.gsflat: - New functionality: * Create nod-and-shuffle flats - New parameters: * fl_double : make double flats for use before sky subtraction * nshuffle : specify number of unbinned shuffle pixels to shift and double gmos.gireduce: gmos.gsreduce: - New functionality: * Support subtraction of nod-and-shuffle darks produced by the new task gnsdark - New parameters: * key_nodcount * key_nodpix b) Improved GMOS variance and data quality calculations Overhaul of the GMOS package to improve variance calculations: number_of_photons **2 + ron/gain The noise algorithm depends on the number of photons; therefore, it is correct to determine the variance only after the artificial bias level has been removed. All variance and data quality creation was removed from the task GPREPARE and was inserted in GIREDUCE after bias subtraction. This fix applies to spectroscopic data as well as imaging as GSREDUCE and GFREDUCE rely on GIREDUCE to subtract bias frames. (i) Changes to parameter lists gmos.gprepare: - Removed parameters: * fl_vardq * var_ext * dq_ext * bpm * sat * key_dsec gmos.gscut: - New functionality: * Added new approximation of variance if it does not already exist. - New parameters: * key_gain * key_ron * ron * gain gmos.gireduce: - New functionality: * Option to multiply by the gain to make the units of the output data be electrons rather than ADU. - New parameter: * fl_mult : The default is 'yes', fl_mult is not usually modified by the user. c) GFEXTRACT IFU 2-slit mode single slit extraction feature For IFU 2-slit mode data, it is now possible to extract the spectral information one slit at a time. This is *not* the recommended procedure, but it can be a useful feature for some observations taken with a filter not fully compatible with the central wavelength. (i) New functionalities/parameters gmos.gfextract: gmos.gfreduce: - New parameter: * exslits : specifies slit(s) to extract. For 2-slit mode, this option allows each slit to be extracted separately over its full, clean spectral range, instead of just the range common to both slits. It allows for non-standard grating and filter configurations. This parameter should only be used when the center of the filter does not match the desired central wavelength. This only occurs in extreme cases. gmos.gfreduce: - New functionality: * Use information in the header to select the MDF. - Modified parameter: * slits : new option, "header" gmos.gsfquick: - New parameter: * sslit : to specify 'red' or 'blue' slit for single slit extraction d) Update to overscan subtraction Because of a change in the behavior of the GMOS CCDs, it is now recommended to correct the overscan (it was not recommended before). gmos.gbias gmos.gfreduce gmos.giflat gmos.gireduce gmos.gsflat gmos.gsreduce: - New parameters: * nbiascontam : Number of columns removed from the overscan region before overscan calculation. * biasrows: set rows used in overscan region gmos.gfreduce: - New parameter: * low_reject * high_reject * niterate e) No-pre-imaging GMOS MOS mask design software now fully integrated (i) New tasks in gmos$mostools/ - gmos.mostools.gmskcreate -- Produce the required inputs for gmmps from a non-Gemini image with good astrometry - gmos.mostools.gmskimg -- Produce dummy image for use with gmmps - gmos.mostools.gmskxy -- Generate x,y MOS mask coordinates from a list of RA and Dec. f) Add B600 grating and OIII, He, Ha filters to gmos$data/ files gmos$data/gmosGRATINGS.dat: - add rows for B600 grating gmos$data/gmosFILTERS.dat: - update all starting and ending wavelengths - add filters: * GMOS-N: HeII_G0320, HeIIC_G0321, OIII_G0318, OIIIC_G0319, Ha_G0310, HaC_G0311, SII_G0317, DS920_G0312, u_G0308 * GMOS-S: u_G0332, HeII_G0340, HeIIC_G0341, OIII_G0338, OIIIC_G0339, Ha_G0336, HaC_G0337, SII_G0335 gmos.data : new transmission files - CaT_G0309.dat, Ha_G0310.dat, r_G0326.dat, CaT_G0333.dat, Ha_G0336.dat, i_G0327.dat, SII_G0335.dat, u_G0308.dat, OIIIC_G0339.dat, u_G0332.dat, HaC_G0311.dat, OIII_G0338.dat, g_G0325.dat, HaC_G0337.dat, z_G0328.dat g) Other GMOS improvements gmos.giflat: - New functionality: * Supports 'stamp' images - New parameter: * fl_stamp gmos.gstransform: - New functionality: * Allow user to specify the beginning and ending wavelength, pixel ratio, and number of pixels to transform. - New parameters: * lambda1 * lambda2 * dx * nx - Modified parameter: * xlog -> lambdalog gmos.gifringe: - New functionalities: * Much more flexible method of determining zero sky level to optimize fringe frame - New parameters: * typezero: operation to use to determine or specify zero level values. This can be mean, midpt, skyfile, or keyword * skysec: zero point statistics section * skyfile: file with zero point values for each input image * keyzero: header keyword for zero level * glogpars - Removed parameters: * sctype * title * key_ron, * key_gain * ron * gain * snoise * pclip * grow * bpm * fl_dqprop - Modified parameters: * input --> inimages * output --> outimage gmos.gfcube: - New functionalities: * Compensate for atmospheric dispersion - New parameters: * fl_atmdisp * fl_dq * logfile 5) NIFS ------------------ Instrument: NIFS a) Tasks renamed and relocated: * nifs.nfcheck -> gemtools.ckinput, relocated to GEMTOOLS * nifs.nfckcal -> gemtools.ckcal, relocated to GEMTOOLS b) nifs.nfpad: Pads more to account for small shifts in the slices. This is critical to H-band data taken after June 2007 as there was a shift in the slices causing slice 29 to be truncated. c) gemtools.gemcube: Critical bug fix from F. Valdes. GEMCUBE is used to generate a NIFS cube. The weighting was applied incorrectly causing a shift of one pixel in the dispersion features. The geometry of the drizzle weighting now knows which output axis corresponds to the non-spatial axis. d) Remove tasks NFFITCOORDS and NFTRANSFORM: GNIRS.NSFITCOORDS and GNIRS.NSTRANSFORM are now used for all instruments. The NIFS examples and help pages are updated. Use NSTRANSFORM/NSFITCOORDS in the same way that NFTRANSFORM/NFFITCOORDS were used. SUMMARY OF ALL CHANGED PARAMETERS gmos.gbias: - New parameters: nbiascontam, biasrows gmos.gfcube: - New parameters: fl_atmdisp, fl_dq, logfile gmos.gfextract: - New parameter: exslits gmos.gfreduce: - New parameters: low_reject, high_reject, niterate, exslits, nbiascontam, biasrows - New default value: slits = "header" (was slits = "both") gmos.giflat: - New parameters: fl_stamp, nbiascontam, biasrows gmos.gifringe: - New parameters: typezero, skysec, skyfile, keyzero, glogpars - Removed parameters: sctype, title, key_ron, key_gain, ron, gain, snoise, pclip, grow, fl_dqprop, masktype, maskvalue, zero - Changed parameters: * input --> inimages * output --> outimage gmos.gireduce: - New parameters: key_nodcount, key_nodpix, fl_mult, nbiascontam, biasrows gmos.gprepare: - Removed parameters: fl_vardq, var_ext, dq_ext, bpm, sat, key_dsec gmos.gscut: - New parameters: key_gain, key_ron, ron, gain gmos.gsflat: - New parameters: nbiascontam, fl_double, nshuffle, biasrows gmos.gsreduce: - New parameters: nbiascontam, key_nodpix, key_nodcount, biasrows gmos.gstransform: - New parameters: lambda1, lambda2, dx, nx - Changed parameter: * xlog --> lambdalog gnirs.nsprepare: - New default value: fl_forcewcs=yes (was fl_forcewcs=no) gnirs.nstransform: - Removed parameters: lamptransf, sdisttransf, fl_inter, fl_align, function, xorder, yorder, force midir.mireduce: - New parameter: frame midir.msabsflux: - New parameter: fl_plots midir msreduce: - New parameter: fl_skybiassub - Changed parameter: * fl_transform --> fl_fitcoords nifs.nfimage: - New parameter: fl_imexam - Changed parameters: * rootname -> prefix * datadir -> rawpath - New default value: fl_inter=no (was fl_inter=yes) niri.nisky: - New default value: statsec="default (was statsec="[100:924,100:924]")