Changeset 9077 for trunk

Timestamp:

08/04/08 11:44:11 (16 years ago)

Author:

tbretz

Message:

*** empty log message ***

Location:

trunk/MagicSoft/Mars

Files:

• Changelog (modified) (1 diff)
• NEWS (modified) (1 diff)
• mjobs/MJCalibrateSignal.cc (modified) (3 diffs)

trunk/MagicSoft/Mars/Changelog

@@ -18 +18 @@
 
                                                  -*-*- END OF LINE -*-*-
+
+ 2008/08/04 Thomas Bretz
+
+   * added a filter against very bright events cleaning the 
+     bad pixel plot
+
+
 
  2008/08/04 Daniela Dorner

trunk/MagicSoft/Mars/NEWS

@@ -79 +79 @@
      Instead of the sequence number also sequence files are accepted.
 
+ ;callisto
+
+   * added a filter again which was was removed in the last release.
+     it removed the most brightest events. This should not change
+     any analysis result because these events are so bright that
+     they cannot be analysed anyway, but the bad-pixel plots doesn't
+     show outliers.
+
  ;star
 
    * The camera currents are now displayed with the rate of the camera
      reports instead of the caco reports
+
+ ;ganymed
+
+   * if mc camera files are new enough (>=8) the source position in the
+     camera is now calculated from the stored telscope and shower
+     orientation (Note, that this might not be suitable in all cases
+     depending on the aim of the study. These case still needs
+     implementation)
 
 

trunk/MagicSoft/Mars/mjobs/MJCalibrateSignal.cc

@@ -540 +540 @@
     // an error is raised if more than 50% of the events were skipped.
     //
+    // Since unsuitable pixels are not counted in the ideal case
+    // there is no empty pixel at all.
+    //
     MFCosmics fcosmicscal;
     fcosmicscal.SetNamePedestalCam("MPedestalFundamental"); //CORRECT?
@@ -548 +551 @@
 
     //------------------------------
-    /*
-     MFCosmics fcosmicsped; // Def=20% empty pixels
-     fcosmicsped.SetNamePedestalCam("MPedestalFundamental"); //CORRECT?
-     MContinue contcosped(&fcosmicsped, "ContCosmicsPed");
-     contcosped.SetInverted();
-    */
+    //
+    // Remove events which are too bright. These events could either
+    // be calibration events or events which produce many many
+    // saturating low-gains and thus spoil the dead pixel plot.
+    // These events don't seem to be physical events so we can fairly
+    // remove them (they are no analysable gammas anyway)
+    // So we remove all events which have less than 5% empty pixels.
+    //
+    MFCosmics fcosmicsbright;
+    fcosmicsbright.SetNamePedestalCam("MPedestalFundamental");
+    fcosmicsbright.SetMaxEmptyPixels(0.10);
+    fcosmicsbright.SetMinAcceptedFraction(0.9);
+
+    MContinue contbright(&fcosmicsbright, "ContBrightEvts");
+    contbright.SetInverted();
 
     //------------------------------
@@ -845 +857 @@
         tlist2.AddToList(&taskenv2);
 
+    // Check if we have an extremely bright event (remove them,
+    // they are presumably errornous events or calibration events)
+    tlist2.AddToList(&contbright);
+
     /*
     if (fIsHiLoCalibration)