--- Release Notes on IRC_SPECTRED v20101025 for phase 1 and 2 data reduction

Corrections made in this version:

1. Correction on the error calculation for all IRC slit spectroscopy data
2. Correction on the bad pixel propagation
3. Correction on the Nan data in the split spectroscopy

Descriptions

1. Erroneous estimate of errors in the slit spectroscopy

1.1 Symptom:
When one reduces the IRC slit spectroscopy data (Ns, Nh, Ms, and Ls) and plots the spectra with the plot_spec_with_image command, the errors have been erroneously estimated. The errors were by a factor of sqrt(number_of_frames) larger than the correct value for faint spectra, but the factor varies according to the brightness of the spectra. This occurs only for the slit spectrum processing.


1.2 Corrections made:
The factor of 1/sqrt(number_of_frames), which was not taken into account in the estimation, has been added in the estimation.  Since for bright sources, other errors become dominant and thus the estimated errors were not 
significantly affected by this problem.

1.3 Updated subroutines:
IRC_UTIL/w_mean.pro
IRC_UTIL/calc_mir_refimage.pro
IRC_UTIL/meanclip_mod.pro
IRC_UTIL/mean_sigma.pro
IRC_UTIL/weighted_mean.pro     ... new program
IRC_SPECRED/irc_specred.pro
IRC_SPECRED/do_skysub_n.pro

1.4 Instructions for users:
You must redo the data reduction from the beginning in irc_specred. Staring from the plot_spec_with_image WILL NOT correct the errors.  The source table file ('***.source_table.tbl') and the IDL save file ('***_IRC_SPECRED_OUT.sav') will be modified by the revised version.  For slit-less data, the errors are correctly estimated and you do not have to redo the reduction.


1.5. Usage of the new pipeline:
Usage is the same as for the original version.  The corrected software will not modify the FITS output files.


2. Correction of bad pixel propagation
Erroneous propagation of bad pixels (mostly affecting phase 3 data reduction) has been corrected and the number of masked pixels decreases.  

2.1 Updated subroutine:
IRC_SPECRED/irc_distcalib.pro

2.2. For users
You have to redo the reduction from the beginning. 

3. Correction on the Nan data in the slit spectroscopy
In the diffuse slit spectroscopy, treatment of Nan data has been corrected.

3.1 Updated subroutine:
IRC_SPECRED/make_spec1d.pro

3.2 For users
You can restart from plot_spec routines.



