Changeset 30335

Timestamp:

Jan 21, 2011, 11:32:52 AM (15 years ago)

Author:

eugene

Message:

CHIP.SEEING (used by ppSub and ppStack) should probably be based on moments, not PSF model

File:

• branches/eam_branches/ipp-20101205/psphot/src/psphotChoosePSF.c (modified) (1 diff)

branches/eam_branches/ipp-20101205/psphot/src/psphotChoosePSF.c

@@ -469 +469 @@
 
     float fwhmMin = stats->sampleMean;  // FWHM on minor axis
+
     if (readout->parent) {
+
+        // we now have 2 possible measurements of the seeing : the PSF model based version and
+        // the Moments based value we need to define a definitive "CHIP.SEEING" value.  This
+        // value is used by the PSF-matching programs (ppSub and ppStack).  The moments-based
+        // value is probably more representative of the value needed by those tools (it is also
+        // more stable?)
+        float fwhmMajorMoments = psMetadataLookupF32(&status, readout->analysis, "IQ_FW1");
+        float fwhmMinorMoments = psMetadataLookupF32(&status, readout->analysis, "IQ_FW1");
+
         pmChip *chip = readout->parent->parent; // Parent chip
         psAssert(chip, "Cell should be attached to a chip.");
         psMetadataItem *item = psMetadataLookup(chip->concepts, "CHIP.SEEING"); // Item with chip
-        item->data.F32 = 0.5 * (fwhmMaj + fwhmMin);
+        item->data.F32 = 0.5 * (fwhmMajMoments + fwhmMinMoments);
     }