Date: Tue, 17 Oct 2006 (Typos corrected on Jun 5, 2007)

We have recalibrated the IRS Peak-Up Imaging (PUI) mode after the
first year of its operation.  The new calibration is based on
observations of five stars.  Changes from the previous calibration
include the use of IRS spectra to tie down the photometric calibration, a new
definition of the effective wavelength of the filters, and a new fixed
aperture for measurement of the zeropoint.  We provide new flux
zeropoints which are 8% and 16% lower than the previous calibration
in blue and red, respectively.  The new calibrations will be applied
when data is processed through the S15 pipeline.  We also provide new
aperture corrections based on STinyTim 2.0, and new PRFs for use in
APEX.  We plan future observations to improve the red PRF.

Typically, observers will measure lower fluxes in the reprocessed 
data.

The new flux zeropoints are: 0.0124 +/- 5% for blue and 0.0136 +/-
5% for red in units of MJy/sr/elec/sec.  These values are lower than 
the previously adopted values
of 0.01375 and 0.01617, respectively.  These difference arise
from several factors:

  The new flux calibration was pegged to a monochromatic wavelength,
  at the effective wavelength of each filter.  These effective
  wavelengths were defined following the Color Correction memo.
  Effective wavelength values were 15.6 microns for blue and 22.0
  microns for red. The effective wavelength for blue was initially
  taken to be 16.0 microns, but is now correctly replaced with 15.6
  microns.  This change raises the zeropoint by ~2%.

  PUI observations of each standard star were taken in the blue and
  red filters.  The counts (in electrons per second) were measured in
  standard apertures of 12 pixels radius in blue and 13 pixels radius in red. 
  The width of the sky annulus varied but was typically about 5 pixels 
  i.e. radius of 12-17 pixels for blue and 13-18 pixels for red. By
  choosing finite apertures in which to measure standard star fluxes,
  we peg the flux scale to that aperture.  Extended sources larger
  than that aperture will need to know the aperture correction out to
  larger apertures.  The aperture correction was calculated using
  STinyTim version 2.0.  The correction values were validated using a
  deep PUI observation of a star over an extended (3 arcminute radius)
  area.  Using the new 12 and 13 pixel radii lowers the zeropoint by a
  few percent.

  Actual IRS spectra were used for
  4 of the 5 stars, instead of models, to cross calibrate the fluxes. 
  It is not known in every case in which direction
  this correction would go.

---------------------
APERTURE CORRECTIONS
---------------------

  RADIUS    APCOR BLUE   APCOR RED
       2     1.661        1.885 
       3     1.418        1.561 
       4     1.169        1.412 
       5     1.085        1.200 
       6     1.068        1.086 
       7     1.057        1.050 
       8     1.048        1.039 
       9     1.033        1.032 
      10     1.019        1.025 
      11     1.007        1.020 
      12     1.0          1.011 
      13     0.9960       1.0   
      14     0.9918       0.9906
      15     0.9873       0.9828
      16     0.9831       0.9765
      17     0.9791       0.9723
      18     0.9757       0.9699
      19     0.9726       0.9672
      20     0.9699       0.9640
      21     0.9675       0.9609
      22     0.9652       0.9581
      23     0.9632       0.9553
      24     0.9613       0.9529