/* ======================================================================== *\ ! ! * ! * This file is part of MARS, the MAGIC Analysis and Reconstruction ! * Software. It is distributed to you in the hope that it can be a useful ! * and timesaving tool in analysing Data of imaging Cerenkov telescopes. ! * It is distributed WITHOUT ANY WARRANTY. ! * ! * Permission to use, copy, modify and distribute this software and its ! * documentation for any purpose is hereby granted without fee, ! * provided that the above copyright notice appear in all copies and ! * that both that copyright notice and this permission notice appear ! * in supporting documentation. It is provided "as is" without express ! * or implied warranty. ! * ! ! ! Author(s): Thomas Bretz 12/2000 ! ! Copyright: MAGIC Software Development, 2000-2004 ! ! \* ======================================================================== */ ///////////////////////////////////////////////////////////////////////////// // // MRawEvtHeader // // One Event is a sample of FADC measurements of different Pixels // (Photomultipliers) from the Camera of MAGIC. So all data (FADC) of the // interesting pixels are the modules of an event. To describe pixels the // class MRawPixel is used and the class MRawCrate to describe Crates. // To define a single events some other data members are needed // (Time of the events, tirgger pattern of event..) // // To describe one event on the level of FADC values the Class MRawEvtHeader is // created. It has the following data members: // // UInt_t fDAQEvtNumber // ----------------------- // This it the number of the Event in one // data run. The first event in this run get // the number zero. The next one is one bigger. // // Assuming that one run takes 1 hour and a // triggerrate of 1kHz the number must be able // to reach 3.6e6 Events. To reach this number // you need at least 22 bits. This is the reason // why we use an integer (of root type UInt_t) // with a range to 4.2e9. // // MTime fRawEvtTime // ------------------- // Time of the event. // The start point of the time determination can be // the millenium. From that start point the time is // measured in 200ns-count. One day for example // contains 432.e9 counts. An array of two unsigned Int is able to // contain 1.8e19 200ns-counts. This corresponds to 41.e6 // days. This should be more than the livetime of MAGIC. // Private member of MTime.h // // UInt_t fNumTrigLvl1 // -------------------- // // Number of first level trigger // This member counts the number of First Level Trigger // between the last and this event. May be that due to // dead time of the DAQ this number is different from 1. // If the DAQ is fast enough, this value should be 1. // This may be usefull in GammaRayBursts and if we // apply a data reduction in the DAQ-chain, which selects // only good events. // // UInt_t fNumTrigLvl2 // -------------------- // // Number of second level trigger // This member counts the number of Second Level Trigger // between the last and this event. // // UInt_t fTrigPattern[0] // ----------------------- // Trigger Pattern used for this event // Each event triggers for a particular configuration and each // configuration should have an ID (which is not fixed yet). // // UInt_t fCalibPattern == fTrigPattern[1] // ----------------------- // Calibration Pattern used for this event // Each (calibration) event uses a particular LEDs configuration and // particular strength and colour of the continunous light. // Bits 1-16: Pulser slot pattern: 16 LEDs slots. // Bits 17: CT1 Pulser // Bits 21-24: Colour of Continous light source: // Bits 25-32: Strength of continuous light source: 256 level // // UShort_t fAllLowGainOn // ---------------------- // Type of Trigger. // This is a Byte (8 bit) to indicated if any of the pixels // have a non-negligible low gain (1) or not (0) // // Class Version 2: // --------------- // - added fCalibPattern // ///////////////////////////////////////////////////////////////////////////// #include "MRawEvtHeader.h" #include #include #include "MLog.h" #include "MLogManip.h" #include "MTime.h" #include "MArrayB.h" #include "MRawRunHeader.h" ClassImp(MRawEvtHeader); using namespace std; // -------------------------------------------------------------------------- // // Default constructor. Create the array to store the data. // MRawEvtHeader::MRawEvtHeader(const char *name, const char *title) { fName = name ? name : "MRawEvtHeader"; fTitle = title ? title : "Raw Event Header Information"; // // set all member to zero, init the pointer to ClonesArray, // fPixLoGainOn = new MArrayB; Clear(); } // -------------------------------------------------------------------------- // // Destructor. Deletes the array to store pixlogainon // MRawEvtHeader::~MRawEvtHeader() { delete fPixLoGainOn; } // -------------------------------------------------------------------------- // // you have to init the conatainer before you can read from // a raw binary file // void MRawEvtHeader::InitRead(MRawRunHeader *rh, MTime *t) { // // this is the number of entries in the array like specification // UInt_t fN = (rh->GetNumCrates() * rh->GetNumPixInCrate() + 7) / 8; // // initialize the array // fPixLoGainOn->Set(fN); // // this is the conatiner where we have to store the time of the event we // read from the input stream // fTime = t; } // -------------------------------------------------------------------------- // // Implementation of the Clear function // // Resets all members to zero, clear the list of Pixels // void MRawEvtHeader::Clear(Option_t *) { fDAQEvtNumber = 0; fNumTrigLvl1 = 0; fNumTrigLvl2 = 0; fTrigPattern[0] = 0; fTrigPattern[1] = 0; fTrigType = 0; fNumLoGainOn = 0; } // -------------------------------------------------------------------------- // // This member function prints all Data of one Event to *fLog. // void MRawEvtHeader::Print(Option_t *o) const { *fLog << all; *fLog << "DAQEvtNr: " << dec << fDAQEvtNumber << " ("; *fLog << "Trigger: "; *fLog << "NumLvl1=" << fNumTrigLvl1 << " "; *fLog << "NumLvl2=" << fNumTrigLvl2 << " "; *fLog << "Pattern=" << hex << setfill('0'); *fLog << setw(8) << fTrigPattern[0]; *fLog << setw(8) << fTrigPattern[1] << " " << dec; *fLog << "Type="; switch (fTrigType) { case 0: *fLog << "Trigger"; break; case 1: *fLog << "Pedestal"; break; case 2: *fLog << "Calibration"; break; case 3: *fLog << "PinDiode"; break; } *fLog << ")" << endl; *fLog << "Number of Lo Gains On: " << fNumLoGainOn << endl; TString str(o); str.ToLower(); if (str.Contains("nogains")) return; for (unsigned int i=0; iGetSize(); i++) { for (int j=0; j<8; j++) { const UInt_t on = (*fPixLoGainOn)[i]&(1<GetSize()) *fLog << endl; } // -------------------------------------------------------------------------- // // Return the size in bytes of the event header. // Int_t MRawEvtHeader::GetNumBytes() const { return 36+fPixLoGainOn->GetSize(); } // -------------------------------------------------------------------------- // // Used to set the header information. This is for MC only. NEVER, NEVER // use this somewhere else! // void MRawEvtHeader::FillHeader(UInt_t uiN, Float_t ulTP) { fDAQEvtNumber = uiN; fTrigPattern[0] = (UInt_t)(ulTP/4294967296.0) ; fTrigPattern[1] = (UInt_t)(ulTP-fTrigPattern[0]*4294967296.0); } // -------------------------------------------------------------------------- // // Decode the binary Time Stamp. For more detailed information see the // source code. // Bool_t MRawEvtHeader::DecodeTime(UInt_t abstime[2], UShort_t ver) const { // // SuperSecond (20 bits giving hh:mm:ss) // ------------ // // Reading the hours: // Swap bits: 23->16, 22->17, 21->16, 20->19 // abstime[0] = abstime[0]>>7 & 0x00010000 | abstime[0]>>5 & 0x00020000 | abstime[0]>>3 & 0x00040000 | abstime[0]>>1 & 0x00080000 | abstime[0] & 0xff00ffff; // // SubSecond (24 bits giving number of clock ticks of a 5Mhz signal since // the beginning of last second, i.e., number of ns with a precision of to // 200 ns) // ---------- // // The last 8 bits must be flipped. // abstime[1] ^= 0x000000ff; // // Due to a problem with one Digital Module, three of the less significant // eight bits of the subsecond are corrupted. So, until new DM's arrive to // La Palma, we won't use the eight first bits of the subsecond. // This reduces the precision from 200 ns to of 51.2 us. (ver<5) // if (ver<5) abstime[1] &= 0xffffff00; // // Decode Time Stamp // const Byte_t h = (abstime[0]>>18 & 0x3)*10 + (abstime[0]>>14 & 0xf); const Byte_t m = (abstime[0]>>11 & 0x7)*10 + (abstime[0]>> 7 & 0xf); const Byte_t s = (abstime[0]>> 4 & 0x7)*10 + (abstime[0]>> 0 & 0xf); const UInt_t ns = abstime[1]*200; // // Update the time stamp with the current event time. // Make sure, that the time stamp was initialized correctly // with the start-date/time of the run (after reading the run header) // if (fTime->UpdMagicTime(h, m, s, ns)) return kTRUE; *fLog << warn << "WARNING - Time (" << Form("%2d:%02d:%02d,%09d", h, m, s, ns); *fLog << ") in header of event #" << dec << fDAQEvtNumber << " invalid..." << endl; return kFALSE; } // -------------------------------------------------------------------------- // // read the EVENT HEADER information from the input stream // return FALSE if there is now header anymore, else TRUE // // For version>2 we expect to have a valid time-stamp in the files. // // Updates the time stamp with the current event time. // Make sure, that the time stamp was initialized correctly // with the start-date/time of the run (after reading the run header) // // Remark: This 'feature' disallows single runs of more than 11h! // Int_t MRawEvtHeader::ReadEvt(istream &fin, UShort_t ver) { Bool_t rc = kTRUE; fin.read((char*)&fDAQEvtNumber, 4); // Total=4 if (!fin) return kFALSE; UInt_t abstime[2]; fin.read((char*)abstime, 8); // Total=12 if (ver>2) if (!DecodeTime(abstime, ver)) rc = kCONTINUE; Byte_t dummy[4]; fin.read((char*)&fNumTrigLvl1, 4); // Total=16 fin.read((char*)&fNumTrigLvl2, 4); // Total=20 fin.read((char*)fTrigPattern, 8); // Total=28 fin.read((char*)&fTrigType, 2); // Total=30 if (ver<=6) { fin.read((char*)dummy, 2); // Total=32, was fAllLoGainOn fin.read((char*)fPixLoGainOn->GetArray(), fPixLoGainOn->GetSize()); fNumLoGainOn = 0; for (unsigned int i=0; iGetSize(); i++) for (int j=0; j<8; j++) if ((*fPixLoGainOn)[i] & (1<Reset(); fNumLoGainOn = 0; } return fin.eof() ? kFALSE : rc; } void MRawEvtHeader::SkipEvt(istream &fin, UShort_t ver) { fin.seekg(GetNumBytes(), ios::cur); } // -------------------------------------------------------------------------- // // Low level decoding of the trigger pattern. // The trigger pattern consists of 16 bits (8+8 bits) generated by the // trigger system. // The first 8 bits correspond to the trigger configuration before the // prescaling, the others after prescaling. // The meaning of the configuration depends on the chosen trigger table // (that is how the trigger has been programmed) and must be interpreted // at higher level by the analysis. // Bit structure: // not prscd | prscaled // xxxx xxxx xxxx xxxx <-- pattern (x=0,1) // bit 7654 3210 7654 3210 // H L // // e.g. 1000 0000 1000 1000 (hex: 8080) is the pattern when no // L2 trigger selection and no prescaling is applied. // // Up to now only fTrigPattern[0] is used. // UInt_t MRawEvtHeader::GetTriggerID() const { return fTrigPattern[0]; } UInt_t MRawEvtHeader::GetCalibrationPattern() const { return fTrigPattern[1]; } UInt_t MRawEvtHeader::GetPulserSlotPattern() const { return (fTrigPattern[1] >> 16) & 0x1ffff; }