Hi Jeremy,

Sorry for my slow reply.


> Is the scintillator (PMT?) state/data available in the converted LCSim events?  Or is it currently being ignored now and not converted?  I know in the EVIO to LCIO converter it says the channels don’t exist.  From my very basic understanding of things, the PMT data is needed for cutting events from the sample.

Sho did write a driver to read the PMT channels.  It is likely possible to get by
without them in offline analysis, but we are currently using them to cleanup
the cosmic signals.  How much effect it has in the end I'm not sure; maybe
thresholds and geometric cuts on the calorimeter render offline offline PMT
cuts irrelevant.


> Is the trigger data needed/useful for doing this analysis?  If so, is it being properly converted to LCSim yet?


I do not know the answer the second question, but there are currently two possible
triggers in our cosmic data, a 1Hz pulser and ~5Hz cosmic, in the TI bank (57610):

0x20010003
0x10010003

I've not noticed their "AND" in the data; maybe it would be a 3 in that last hex digit.
I'm not aware of anyone using them currently, and its highly likely that offline cuts
on PMT pulse integrals make the trigger bits not important.  But that was one thing to
do before finalizing cosmic gains: cut on trigger.  I copied below an email from Ben.


> How has the analysis, if any, been done thus far?  If there is any code in C and/or ROOT with algorithms that can be easily ported?


Just a reminder, the C-code I am using to convert evio to ROOT ntuples, and
some followup analysis on the ntuples, is below.  But I guess you are looking for
something else.  Holly has something more sophisticated regarding geometrical
cuts.

/home/baltzell/EvioTool/EvioTool_Test/AnaEcal.cc
/home/baltzell/a.C


> What sigma is a good cut for trying to pick out a MIP/cosmic signal from the pedestal + noise?  I think Nathan said 10 ADC here from my understanding, but would this make more sense as a function of sigma * N, where sigma measures noise fluctuation?


A global ADC threshold of 10 was chosen because it is just over 3-RMS noise for
the majority of channels while still less than cosmic signal.  The two are not much
different, and whether it is "good" is subjective.  But it is definitely sufficient when
combined with other cuts.


> Also, is anyone else working on analysis of this data in the Java framework or planning to do it?  I would be interested in participating/collaborating on this, if so.

Probably everyone intends to analyze HPS data in the Java framework.  Time is tight,
and with all the flux in data format etc, just because some people are working outside
that framework for now should not imply anything otherwise!

Happy Holidays,
Nathan








> For TImaster bit(29:24) represents the TS#6-1 inputs
> 
> So:
>
> bit 24 should be HPS Singles 0
> bit 25 should be HPS Singles 1
> bit 26 should be HPS Pairs 0
> bit 27 should be HPS Pairs 1
> bit 28 should be Cosmic
> bit 29 should be pulser
>
> I think this is consistent with what you're seeing.
>
> Ben
>
>
> ----- Original Message -----
> From: "Nathan Baltzell" <[log in to unmask]>
> To: "Benjamin Raydo" <[log in to unmask]>, "Serguei Boiarinov" <[log in to unmask]>
> Sent: Monday, November 24, 2014 3:03:25 PM
> Subject: trigger in evio
>
> Hi Guys,
>
> I want to know what trigger caused each event. 
> I suspect it is the first one in this bank:
>
> <bank content="uint32" data_type="0x1" tag="57610" num="0">
>            0x10010003        0x27bfa      0x343e90a          0x5e9
> </bank>
>
> About 20% of the events have 0x20010003 and the rest have
> 0x10010003, so that must be 1Hz pulser and cosmics respectively.
>
> Can you confirm?
>
> Thanks,
> Nathan
>


On Nov 26, 2014, at 11:47 AM, "McCormick, Jeremy I." <[log in to unmask]> wrote:

> Hi,
> 
> I have a number of basic questions about doing an analysis in LCSim from the ECAL raw mode cosmic data.
> 
> Is the scintillator (PMT?) state/data available in the converted LCSim events?  Or is it currently being ignored now and not converted?  I know in the EVIO to LCIO converter it says the channels don’t exist.  From my very basic understanding of things, the PMT data is needed for cutting events from the sample.
> 
> Is the trigger data needed/useful for doing this analysis?  If so, is it being properly converted to LCSim yet?
> 
> How has the analysis, if any, been done thus far?  If there is any code in C and/or ROOT with algorithms that can be easily ported?
> 
> What sigma is a good cut for trying to pick out a MIP/cosmic signal from the pedestal + noise?  I think Nathan said 10 ADC here from my understanding, but would this make more sense as a function of sigma * N, where sigma measures noise fluctuation?
> 
> Also, is anyone else working on analysis of this data in the Java framework or planning to do it?  I would be interested in participating/collaborating on this, if so.
> 
> Thanks.
> 
> —Jeremy
> ########################################################################
> Use REPLY-ALL to reply to list
> 
> To unsubscribe from the HPS-SOFTWARE list, click the following link:
> https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=HPS-SOFTWARE&A=1

########################################################################
Use REPLY-ALL to reply to list

To unsubscribe from the HPS-SOFTWARE list, click the following link:
https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=HPS-SOFTWARE&A=1