"MIA" - MIDI Interactive Analysis


Introduction


On this sites we want to keep you informed about the current status of MIA - the software developed at the MPIA to reduce data obtained with MIDI. Although some of the routines are taken from the software package developed in Leiden (e.g. the I/O-routines), the procedure for analysing the data is different. MIA uses a Fast Fourier Transformation to calculate the Fourier amplitudes of the fringe packets. These amplitudes help to decide, whether a fringe packet can be used to calculate the visibility or not. Please see below what modes can be handled by MIA.
  
Dispersive Elements
 PRISM> 1.0
GRISM> 1.03alpha1 (not fully tested)
 
Channels
 HIGH_SENSE  > 1.0
SCI_PHOT> 1.3
 
Tracking mode
 undispersed> 1.0
dispersed> 1.03alpha1 (optional, not fully tested)
 
Special
 Choppingshould work in principle when header informations are used  

History


  • MIA+EWS version 1.5 (29-Sep-06)
    • second sky subtraction disabled for large dsky
    • store more things in xmdv-object
    • new option /UNMASKED for xmdv::photometry()
    • wavelength scale corrected by ~1%
    • midifile_search without arg does what user means
    • added miaguis
    • fit for spectrum-position with slope/curv. fixed
    • maskgui shows regions used for sky-subtraction
    • Introduced Installation Wizard
    • Fixed and Fitted mask will reset mask-filename
    • "Plot Almost Everything" in GUI
    • Error message if user asks for photometry from window 2 although EWS-photometry was used.
    • button in the mask-GUI to use default EWS-Mask with MIA
    • default is to fit the mask
    • Function xmidispvisi::baseline to compute the mysterious projected baseline
    • get wavelength calibration directly from the input files

  • MIA+EWS version 1.4 (16-Jun-06)
    • "Save xmdv-object" in GUI-menu
    • new function midifile_search
    • added baselines between AT-stations to calibrate_disk.pro
    • cleanup-routine has option /keep_compr
    • support for SCI_PHOT mode
    • adapted to new "single-file" EWS-Photometry
    • reactivated SCI_MODE by FP (under development)
    • improved determination of mask parameters
    • xmdv recognizes OBS_FRINGE_TRACK_DISPERSED as fringe track data when called with one filename only
    • flux in SCI_PHOTometric channels is plottet in GUI
    • When xmdv determines whether a scan was on target or on sky, a number of frames at the beginning and end of the scan is ignored. This number can be set with the parameter TARTYPSKIP
    • removed switch /DONT_WRITE
    • serious errors are logged in "ERRRORS/<MIDI-filename>.log"

  • MIA+EWS version 1.3 (14-Dec-05)
    • fixed bug in mask_utilities that caused xmdv to always use the mask created from the fitted traces (this occured only with the "old" photometry)
    • support for SCI_PHOT in photometric and interferometric files: x->photometry(/Aphot) or /Bphot, x->sci_photometry(/A) or /B
    • get_photometry() replaced by photometry()
    • added small window to set plot-ranges from GUI
    • new routine print_results
    • new menu items in GUI: "Mask GUI", "White Fringe Method", and "Save Visibility"
    • new routine save_visibility (writes IAU-format)
    • miamask uses Walter's oirChopPhotoImages
    • photometry from oirChopPhotoImages and oirMakePhotoSpectra This breaks just about anything with SCI_PHOT mode. We get only one spectrum per photometric file, therefore get_photometry(/A2) and ...(/B2) return only zeroes.
    • added x->make_white_fringe to allow changing dispersedtrack and number of bins on the fly
    • removed SKY- and THRESHOLD-options
    • removed MASKTHENMULTIPLY and MULTIPLYTHENMASK
    • get_photometry returns /AB by default
    • introduced x->set_(frame|opd|flux|fluxdiff|fringe_ampl)_range
    • Cached_MeanRMS recreates cache-file if it can't be loaded
    • renamed xmidispvisi::plot_disp_goodbad to xmidispvisi: :raw_correlflux and added ability to estimate noise from same scan as signal
    • xmidispvisi::calibratedvisi accepts an array of calibrators and computes mean instrumental visibility
    • added option /SSN to x->raw_correlflux to estimate noise from power in same scan as signal, but at different frequencies
    • added oirChopPhotoImages
    • force compilation in 32-bit mode (see makecfitsio) This is tricky, since old gcc versions don't accept -m32, while newer versions (those on 64-bit heaters) require it...
    • idl/cmps_form.pro: The IDL manual states that pickfile is obsolete and should be replaced by dialog_pickfile. Did that.

  • MIA+EWS version 1.2 (16-Jul-05)
    • compute TARTYP per scan, plot it over Flux Diff. (blue = S, red = U), and select only scans on target for signal estimate
    • added PREFIX-option to xmdv
    • improved layout of plots in GUI
    • added CHARSIZE-option to x->gui
    • convinced IDL that short and long wavelength limits are floats This affects the dispersed tracking algoritm so strongly I wonder how it ever worked in the old version...
    • x->lambda() returns only valid datapoints
    • x->set_lambda_bins checks for out-of-range values
    • guesses scanlength if value in header is bogus
    • added parameter "errorlog" to xmdv et al.
    • print wavelength in plots of dispersed powerspectra
    • included the reducenight script by Frank and Thorsten
    • try to repair colortable after, e.g., atv has been used
    • get_photometry now uses lambda binning
    • changed selection of frequencies in power spectrum: use always 4 pixels around mean frequency. This is the Right Thing to do as long as scanlength << coherence length.
    • wavelength range returned by correlflux, visibility etc. Now gives range from left edge of leftmost to right edge of rightmost pixel.
    • changed xmidispvisi::get_trace to use miamask-object. This changes the syntax somewhat, read the friendly manual.
    • reduced size of midigui to make it small enough for XGA displays
    • divide correlated flux by 2 so that visi = corrflux/photflux. This cancels out with the factor 4 in the power determination.

  • MIA+EWS version 1.1 (May 2005, intermediate GTO-release)
    • new version of chop_nod_phot by Thorsten Ratzka: Allows determination of the flux, position, and FWHM of the source in each frame of the acquisition data
    • added tests for chopping throw zero, trace out-of-window, and FWHM < 0
    • moved routines for mask creation etc. into object "miamask"
    • added procedure xmidispvisi::show_traces Now you can plot or print the traces as often as you like
    • store traces in arrays in preparation for sci_phot mode
    • Mask check now plots pos/width of mask on top of RMS img
    • rewrote xmidispvisi::meanmask to use loops over tags to simplify handling of HI_SENS and SCI_PHOT
    • new function make_mask_from_trace
    • removed warnings that this is an experimental version. This does not make the results more reliable, though.
    • added some warnings about stupid parameter settings.
    • replaced "Bad threshold" by "Noise OPD" in GUI
    • fixed sqrt(2)-bug in display.
    • gui uses true- or pseudocolor depending on number of colors

  • MIA+EWS version 1.0 (3-Feb-05)
    • added make_maskname in "MIA/utilities.pro"
    • improved file-type detection in xmdv, now recognizes "*_CHOP" as photometric data
    • added old mask under the name "minrtsRef/config/prismmask_before_2003dec01.fits"
    • added "setup.csh" for setup without IDL

  • MIA+EWS version 1.0-test2 (1-Feb-05)
    • fixed paths to C-programs in "c/bin/disp*" scripts
    • added "README.MIA+EWS" in "cfitsio"
    • added mia-script to simplify startup

  • MIA+EWS version 1.0-ESOtest (28-Jan-05)
    • added code to select predefined maskfile if the maskfile generated from the photometry is not satisfying enough
    • removed assumption that maskfilenames are "PHOTOMETRY/Mask_*", i.e. maskfiles can be specified with full path now

  • Version 1.06 (MIA+EWS version 1.0-MPIAtest (27-Jan-2005)
    • incorporated changes by Frank Przygodda for Sci-Phot-mode according to Walter, these won't work properly since they don't do the mapping between photometric and interferometric channels right.
    • threw in reducenight.pro and an extended version of make_midi_log.pro

  • Version 1.05 (31-Oct-2004, "Halloween-version")
    • made white-light tracking the default again
    • removed install-light.sh, pure IDL-version does not work anyway
    • replace spawn, "rm" calls by file_delete
    • removed unused code
    • no longer uses environment variable vltiCbin
    • changed vltimake to recompile packages
    • simplified paths
    • the expert went away to search for his sunglasses

  • Version 1.04 (20-Oct-2004)
    • made dispersedtrack the default, use /whitetrack to disable it
    • giving a directory to xmdv will start midiguis
    • xmdv will search for photometry if no photometry-files are given
    • fringe- and photometry-files can be given in any order
    • removed options that nobody remembered

  • Version 1.04exp1 (18-Okt-2004)
    • set default noiseopd to 4mm
    • added code in xmdv to call midiGuis if argument is a directory, and look for photometry if only fringe-track or fringe-search is given

  • Version 1.03alpha5 (31-Aug-2004)
    • added some interactive elements to 'chop_nod_phot' (orientation & assignment of the beams have to be verified)

  • Version 1.03alpha4 (19-Jul-2004)
    • fixed bug that wrote traces_win[12].ps every time now the files are created if SHOW_TRACES="ps" is given

  • Version 1.03alpha3 (23-Jun-2004)
    • Thorsten changed some Makefile

  • Version 1.03alpha2 (7-Jun-2004)

    Suggestions/Bug fixes by Ch. Hummel:
    • install.sh: "make clean" before "make"
    • miFitsTables: "make clean" rm's libraries
    • MIDIskySubImage cleans up imagedata-object
    • xmidispvisi ptr_free's good- and badfringes
    • missing expert in xmidimovie fixed

  • Version 1.03alpha1 (26-May-2004)
    • fix stupid bug that selected the wrong pixels with Grism (Versions 1.01 and 1.02 did not produce usable results with Grism-data!!!)
    • added simple support for dispersed fringe-tracking: The fringe amplitude is not computed in white light, but in a number of wavelength-bins. Activate it with:

      x = xmdv(files,/dispersedtrack)
      or
      x = xmdv(files,dispersedtrack=20)

      where 20 is the number of bins to use. Default is 10, which should be enough for all data we have so far.

  • Version 1.02 (16-Apr-2004)
    • new, but still preliminary wavelength calib for Grism
  • Version 1.01 (29-Mar-2004)
    • put whole thing under GNU General Public License and added the necessary comments
    • updated, but not final wavelength calib for Grism

  • Version 1.0 (10-Mar-2004)
    • install.sh uses make -k to keep going if a target fails
    • updated chop_nod_phot and chop_nod_disp from Olivier
    • added explanation of /tru and /block to README.HD
    • added xmidispvisi::whitefringeampl()

  • Download


    MIA+EWS-1.5.tar.gz   (the necessary IDL and C-routines 4.9 MByte)
    For the most recent version please visit the MIA+EWS site at Leiden Observatory!

    Please execute after the download has been completed

    gunzip -c MIA+EWS-XXX.tar.gz | tar -xvf -


    to unpack the tarball. A directory MIA+EWS will be created that needs about 20 MByte of disk space. If IDL 5.4 and greater (it was never tested for earlier versions) is available "MIA" runs on Solaris and Linux machines where the software will be constantly tested. A succesful try has also been made under HP-UX. "EWS" requires IDL 6.x.

    Installation


    Change the current working directory to MIA+EWS-XXX by typing

      cd <path>/MIA+EWS-XXX

    where <path> is the path where you unpacked the tarball. Try

      gmake install

    and relax. Problems that can occur and possible solutions are described in the README file. If the installation runs to completion, assign the IDL environment variable:

      export IDL_STARTUP=<path>/MIA+EWS/STARTUP   (within .bashrc)

      setenv IDL_STARTUP <path>/MIA+EWS/STARTUP   (within .cshrc)

    Alternatively, add the code in this file to an existing IDL_STARTUP script.

    If you want to use the software not so often you can specify the startup script after starting IDL by typing

      @<path>/MIA+EWS/STARTUP

    Believe it or not, that's all! Now you should proceed with the tutorial to learn more about the functionality offered by "MIA". To understand the second package named "EWS" read the documentation that can be found here.