There are lots of changes, mostly centered around lo-gain pulse shapes and noise production
Addition of lo-gain pulse shape
- can select a hi-gain and a lo-gain pulse shape lookup table for each pixel
- or can select a set of analytic pulseshape parameters for each pixel
- Uses a double-pass through pmt-signals, etc. But, the pe_amp for the high-gain pulses is saved and later used with producing lo-gain pulses.
- uses hi-gain pulseshape only unless specify analytic function parameters for hi-gain - loop creation now very, very fast , even when using 20000 noise loops when using lookup tables since the length of the pulse is much less that the length of the noise loop
- minimized the number of floating point operations in creating the loops.
Random number generation
- now have three independent random number generators (TRandom3 from ROOT), one for electronic noise, one for noise, and one for everything else. Thus, can compare pulses with and without noise where the pulses from the shower photons are the same.
Other details are described in the README... in the GrISU directory and in detector.pilot and in the version6 configuration files.
Many changes went into this version:
1. TRandom3 is now the random-number generator for grisudet and for cherenkf7. These codes now have a ROOT dependency.
2. Reorganized kascade directory structure: removed Kascade3 and Kascade7 directories and renamed Kascade as KascadeSrc. Rewrote the make_them_all script (execute the script to see the new command line options). With a KascadeSrc directory, MAC users no longer have to worry about OSX not distinguishing between Kascade and kascade!
3. Removed lots of older, unused files, especially old configuration files and pilot files.
4. REWROTE make_them_all: EXECUTE make_them_all TO SEE COMMAND LINE OPTIONS. The options let you choose the fortran compiler (gfortran or intel) and the grisudet version. Defaults are gfortran and grisudet_shadow_vbf.
5. added make_grisudet script: the single command line entry is the grisudet compile version. Execute the script to see the options. The default option is grisudet_shadow_vbf.
6. new codes for creating "star" photons from point sources, see makePhotonsPtSrc.pilot and makePhotonsOffset.pilot in GrISU/Simulation/Make_Photons for full documentation. These are very useful for creating test photon input files for grisudet.
---------------- CONFIGURATION FILE CHANGES -----------------
New configuration file options and records also documented in Config/Files/veritas2013Feb1_GrISU_version5.0.0.cfg.
7. Separate ANLOG records, one per telescope, to enable setting individual telescope gains.
8. Separate PULSE records, one per telescope, for producing the pulse shape for each telescope.
9. New feature: ISU model for Leeds Pattern Trigger module efficiency: see NUMNO, PATMO, and MODUL records. Each telescope has its own set of modular efficiencies. If you are not using the ISU model, there is no need to include these records in your configuration file.
------------ new grisudet options --------------------------
all new grisudet options are also documented in detector.pilot
10. New record to fix the telescope azimuth and zenith angles to specific values. Default: do not use a fixed telescope pointing so that telescope pointing determined by primary direction plus wobble and telescope offsets. (Can also leave out the record if you do not wish to use this option.) FXTEL <0/1: 0 do not fix az/zn; 1 use given az/zn>
11. In all previous versions of grisudet, the difference between the az/zn values of the primary and the telescope have been correct; but the absolute values of the az/zn have differed slightly from their correct values. Setting the TRUAZ value to 1 gives the corrected values in the VBF file. A value of 0 gives the usual az/zn values (which are incorrect) in the VBF file. With no asterisk, the default is 0 (use the old values)
Compared to the previous one, this version includes the possibility of setting different trigger parameters for the different telescopes.
This version was recovered from the CSV repository. It includes a number of minor corrections, mainly in the makefiles.
Changes made by Charlie.
* kascade: More control line arguments. To get instructions run kascade -h.
* Luis Reyes encountered a bug when no photons were falling on one of the telescope. This was fixed.
Changes made by Charlie.
* Grisudet.c: pe_out print statements, xtilt to xtiltr and same for ytilt to match S line print statement in output_event procedure and also fixed print statements for several debug prints.
* Optics.h: fixed zcos variable only used for debug statements and fixed debug print statement.
* Cherenk/readpilot.h: added GRSAP read statement from pilot file. The pilot flag is at the bottom of Cherenkov.pilot and is used to produce a more detailed output for use with a root plotting package that produces really cool 3-D graphs of the photon-emitting points in the shower and 2-D plots of the photon-hits on the ground.
|GrISU2009July10.tar.gz||With respect to GrISU2009Jul07.tar.gz, Charlie made a few changes in Detector,Cherenkf and Analysis which we had failed to integrate. In the meant time Gary fixed a bug in stage6 within the Datareader suite.|
Charlie found a bug in detector.c and in analysis.c. We convinced ourselves both bugs did nod affect the results of simulations and analysis but this would be worth testing.
Datareader has been improved by Gary and Paul. Main updates are:
1. Extension Analysis code has been updated and improved for both RA/Dec and Camera coords
2. T4 bad bias alignment code has been removed
3. No more MAXCUTEVT - Seg fault problem fixed and MAXNBRUN=600!
|GrISU2008Dec05.tar.gz||Datareader was modified by Paul Nunez so as to eliminate the negative ring suronding bright sources in sky map obtained with the ring background method. The ring is segmented in bins and the bins with the highest and lowest contents are not used for background estimation.|
|GrISU2008Oct14.tar.gz||Most changes in analysis.cpp. Events to be displayed can now be specified in the pilot file and a movie option is also available. The for the analysis of FADC data, the pulse integration window is now tracking the position of the telescope camera average pulse. This makes the analysis more stable against timing issues. Datareader was also modified by Gary Finnegan who implemented the possibiliy of fitting source position and extension.|
|GrISU2008June05.tar.gz||This version is being tested at the moment by the group in Utah. It will be the current version when test will have been done.|
|GrISU2008May06.tar.gz||Problems related to the simulation and analysis of QADC pedestal events were fixed. Pierre made some improvement for VERITAS in data reader. A few not new problems that we should fixe soon are: a bad handling of negative determinant in rotation matrix functions with a very rare occurance, the encounter by detector of an inidentified tag with a very rare occurance and a memory leak problem identified by Slava. Charlie is about to do some changes regardin the pointing of the telescopes.|
|GrISU2008Apr09.tar.gz||A little bug involving the maximal number of telescopes was fixed in cherenk7 and Charlie made is possible to specify the direction and random seed as command line arguments to kascade|
|GrISU2008Apr02.tar.gz||Minor changes in the analysis were made. The camera average pedestal with was replaced by the median and pixels that seem too noisy (greater pedestal width than the boudary threshold) are removed from the analysis. Minor changes also to datareader.|
|GrISU2008Mar27.tar.gz||Many changes were made to the analysis so that pedestal events can be used from the real data and from simulated data as well. This implied small changes in the simulations as well but these were minor. We have also implemented wobble observations in the detector model so the offset can be specified in a NS-EW coordinate system. Pierre has made several improvements in datareader. Improvements in the detector (with in particular the fixing of a bug found by Viatcheslav Bugayov) and in the analysis (output of relative gain and pedestal widths in the parameter file for diagnostic display and improvement in the setting of cleaning threholds when they are given in units of pedestal standard deviation) are in the making already and another version will be released soon.|
|grisu31_10_2007.tar.gz||Most of the changes are VERITAS specific. The detector model can now produce output in the VERITAS data format. The analysis can read the VERITAS dataformat and access the VERITAS database for tracking information. A new program found in Datareader produces maps and other analysis final products.|
|grisu07_02_2006.tar.gz||In the detector model we have introduced the possibility of defining an arbitrary pulse shape. A bug in the grid search procedure was fixed. Analysis is now capable of reading veritas data. For that vbf and mysql have to be installed. The data base is accessed to get the tracking informations. Some of the analysis options might not be working with VERITAS data (in particular we do not identify pedestal records yet) as we only use the GrISU analysis for VERITAS as a work environment we are familiar with, to develop analysis. The makefile has also been improved.|
||Charlie has introduced important changes to grisudet
(the detector code). With the addition of a grid search method, execution
time is now from 6 to 7 times faster than in previous versions. Compilation
is now possible with the gcc/g++ optimization options (after replacing the
deprecated fgets function with getline). There are options for circular, hexagonal,
or square facets or pixels (actually, any regular polygon is possible). And,
grisudet now accepts cherenkf7 output into standard input. These changes are
documented in detector.pilot, the configuration files, and in the grisudet
html documentation (follow the links from here). There are additions in the
facet and pixel records in the configuration file for pixel shape and facet/pixel
rotation angle. If you wish to add these to an older configuration file, you
will find a handy edit_config.cpp code in the Config/Files directory. Just
look at the code and follow instructions (the code inserts a column in configuration
file flag records)
Kascade (Glenn and Mary's latest version) now works properly for gamma showers from zenith. We encourage everyone to use this version of kascade (it requires the absoft or Intel compiler - follow the kascade links to find instructions for installing the free Intel fortran compiler).
We continue to include analysis.c, primarily for use with single telescope Whipple data. The analysis routines for telescope arrays are not fully functional nor are they fully documented and we anticipate moving to other code for this analysis (such as the analysis package supported by the Leeds group or the Veritas standard analysis package currently under development).
||New version of kascade from Glen and Marie plus
a few fix suggested mostly by Henric
|| The pattern trigger algorithm was accelerated.
The analysis.c program was modified so it does not require GSL by default.
Type make stereo to get the reconstruction capability using GSL. The
reading of segment by cherenk7 was modified so it works even with
showers that do not contain any track (this was a problem for very
low energy simulations for SGARFACE for example).
||The pattern sensitive trigger was introduced
by Jamie Holder. It is also possible to offset the source by different angles
in each telescope independently. The analysis now contains several stereoscopic
reconstruction procedure. The documentation has to be updated to account for
||The sky and electronics noise injection were
moved to after the telescope trigger model in order to avoid problems with
triggering on the noise. Parameters in the veritas 4 configuration file were
updated to more sensible values.
||A small bug was found by S.
Fegan in kaskade. It had no effect on gamma rays shower simulations. An
other one was found in analysis. It was affecting the event display. Both
||Lots of changes, especially
to analysis which was cleaned up. The possibility of simulating calibration
data was introduced. The telescope numbering now starts at 1 in the data
as it always did in the configuration files. A version number was introduced
in the configuration. It is used to minimize risks of confusion between
different versions. It was set to 3.0.0
|grisu03_17_2004.tar.gz||In the previous versions, the Cherenkov pulse widths were too large at low elevation. This was resulting from insufficient precision in the output format of kascade. This was fixed.|
|grisu02_07_2004.tar.gz||An bug in the local trigger model was fixed as well as in the analysis image cleaning. The 3 strongest pixels are now search for within the pixels passing the cleaning.|
|grisu30_07_2003.tar.gz||Various improvements. In grisudet, photons are now stored in a link list. The mirror reflectivity can be entered independently, ...|
|grisu12_06_2002.tar.gz||The analysis program was misbehaving when some N records giving the number of photoelectrons on each tube were in the input file. This was fixed.|
|grisu11_15_2002.tar.gz||All the numerical recipes procedure have been removed to avoid copyrights problems. Older versions are not available anymore.|
|grisu11_13_2002.tar.gz||An error has been fixed. Segmentation fault was occurring when a logbook was not requested.|
|grisu10_29_2002.tar.gz||In the output of Cherenk the telescope identifier use to be a negative number It is now a positive number. This was causing problems in the simulation of arrays. Correction to procedure on_telescope in ./GrISU/Simulation/Detector/optics.h Minor modifications in the pilot files. In the configuration file the QADC record now allow to activate a special output giving the total number of Cherenkov photons on each tube.|
|grisu10_23_2002.tar.gz||First Grisu Version available on the web age|