Recent TileCal software activities and work to be done

Release 10.0.4 and 10.3.0

Release 10.0.4 was built last week and patched on 31-May. Main change with respect to 10.0.2 for TileCal is ability to run Calibration hits simulation for CTB with GeoModel as input, so now it's possible to run this simulation on the grid. Unfortunately, not all bug fixes for LAr Calibration hits were backported from main branch to this 10.0.x branch and situation with LAr calib hit identifiers is the same as in 10.0.2. Some other problems reported by Tancredi (e.g. muon deposits a lot of energy in dead material somewhere). So it is not recommended to use 10.0.4 for massive production with Calibration Hits. Looks like 10.3.0 (due to next week) will be better.

Serious problem related to reading TileRawChannels from POOL discovered after migration to Gaudi 16. A lot of error messages and even crashes. (see bug report at https://savannah.cern.ch/bugs/index.php?func=detailitem&item_id=8739 ) Reading from ByteStream is not affected. Still hope to solve this problem before 10.3.0.

COOL integration in Athena is almost done. It'll be possible to retreive constants from DB in Athena job using old interface provided by TileInfo class by the end of the week. Everything will go to release 10.3.0 and this release will be used for COOL-related tests during June commissioning

DC3 calibration questionnaire

Planning for DC3 has started. DC3 calibration questionnaire is attached here, and TileCal answer is given attached here

TODO: decide which physics samples we want to simulate in DC3

TODO: decide what we can do with miscalibrated samples provided in DC3, which algorithms can be applied to physics samples to recover calibration.

Detector Description

Anna Lupi is working on realistic description of the calorimeter installed in the pit. This means that instead of symmetric calorimeter with 64 identical phi-slices we want to describe all special modules as well as dead material in the gap between barrel and extended barrel modules. First step - to describe special ITC modules - is almost done and DetDescr database will be updates with new numbers related to special ITC modules in June. Second step - to describe special extended barrel modules and finally last step - special barrel modules. Full non-symmetric TileCal is expected by the end of the year.

TODO: description of dead material in the barrel/ext.barrel gap is not started yet. Need somebody to work on this: to make inventory of various crates, cables, pipes and to put all this in GeoModel

Gia Khoriauli is working on GeoModel description of active and inactive material for simulataion with Calibration Hits. The database update and all the code is expected in June

Simulation

Several bugs were fixed by Gia Khoriaulli in simulation with Calibration Hits. Final tag to use - TileG4Calib-00-00-19, which is available in 10.3.0 and 10.0.3 Many tests done by Gia demonstrated perfect energy conservation across full eta range and correctness of energy stored in TileHit object. Next step is to start massive production with Calibration Hits for the CTB setup - contact Francesco Spano and Tancredi Carli for details.

Famous TileCal sampling fraction, which was 1/40 in TileCal software since release 1.0.0 needs to be updated. Andrea Dotti reported that correct factor to be used for CTB simulation is 1/38.1 Gia Khoriauli is checking now what is the correct factor using "truth" information from Calibration Hits

Stanislav Tokar started to work with Joe Boudreau from LAr on fast simulation for TileCal. Status report will be done during June TileCal week.

Event Display

There are two different Event Displays in ATLAS: Atlantis and v-atlas (former GeoModel hit display). Both will be used for commissioning.

People from TDAQ group and Iacopo Vivarelli from TileCal are looking at Atlantis now.

TODO: Need somebody to tune v-atlas for commissioning. Workshop on v-atlas event display will be organized by Vakhtang Tsulaia in June. Everybody is welcome!

Reconstruction

Valencia people are comparing different opt filter methods, report is expected at June TileCal week

Calibration hit reconstruction - Gia Khoriauli wrote code CaloCalibHitRec package, mostly algorithms to store Calibration hits info in CBNT as well as algorithms to check correctness of Calibration Hits (e.g. invalid identifiers). Please, contact him if you feel that something is still missing there.

Common LAr+Tile reconstruction - need more people from TileCal testing common Calorimeter reconstruction, testing and tuning TopoClusters, working on hadronic calibration.

Databases

Conditions DB for commissioning: Karl Gellerstedt is working hard to provide new working DB to store CIS calibration constants by June 20th. Still, no COOL support in core Athena software, so all the coding is done outside Athena. Standalone program exists to store calibration constants in POOL file and to register those constants with proper IOV in COOL. Integration with Athena will be done when releases 10.3.0 and 10.0.4 are available (COOL support is expected by that time)

Trigger software

Giulio Usai is trying to use TileMuID algorithm in Level 2. It is not possible right now because TileByteStream converter is too slow. It takes 2 ms to unpack one barrel drawer and 8 ms to unpack ext.barrel drawer.

Sasha Solodkov found the reason why it is so slow. This is because C++ exceptions are used to handle non-connected channels. Just by avoiding exceptions time to unpack one drawer can go down to 0.2 ms. TileCabling service is updated already (doesn't generate exceptions anymore) as well as 8 other packages which use this cabling service. New code is in CVS repository, but not in the nightlies. Tests needed to make sure that every algorithm which uses online ID <-> offline ID mapping is working correctly.

Pedro Amaral updated BS conveter and now Digits->RawChannel conversion in BS converter is calling TileRawChannelBuilder, i.e. the there is only one place in CVS where all optimail filtering is done. This modification was tested on testbeam data with Level2 people. Sasha Solodkov done the same for RawChannel>->Cell conversion: TileCellBuilder is called from BS conveter (not tested yet).

Pedro Amaral provided code to create TileTTL1 object from TileRawChannel or TileDigit - needed for comparison of Level1 output and detector output at the testbeam. Code is available in 10.x.0 and 10.0.x. TTL1 shape provided by TileConditions was updated, but not yet used in the release (will switch to new TTL1 shape after Rome)


Major updates:
-- Main.solodkov - 17 May 2005

Topic attachments
I Attachment History Action Size Date Who Comment
Texttxt DC3_calibration_questionnaire.txt r1 manage 5.3 K 2005-05-19 - 16:10 SanyaSolodkov DC3 calibration questionnaire
Texttxt DC3_calibration_tilecal_answer.txt r1 manage 9.7 K 2005-05-19 - 16:12 SanyaSolodkov TileCal answer on DC3 calibration questionnaire
Edit | Attach | Watch | Print version | History: r7 < r6 < r5 < r4 < r3 | Backlinks | Raw View | Raw edit | More topic actions...
Topic revision: r4 - 2005-06-01 - SanyaSolodkov
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    Atlas All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2020 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback