TPC forum post 425:

I am sending a short description of the
TPX cluster finder flags. This is more an Offline issue
but I guess this is a better group since it is a bit
technical.

1) The flags are obtained in the same way other
variables such as timebin and pad are obtained.
I understand that this is not exported to StHit *shrug*.

See i.e. StRoot/RTS/src/RTS_EXAMPLE/rts_example.C

2) They are defined in:

StRoot/RTS/src/DAQ_TPX/tpxFCF.h

The ones pertaining to Offline are:

FCF_ONEPAD           This cluster only had 1 pad.
Generally, this cluster should be ignored unless
you are interested in the prompt hits where this
might be valid. The pad resolution is poor, naturally.

FCF_MERGED           This is a dedconvoluted cluster.
The position and charge have far larger errors
than normal clusters.

FCF_BIG_CHARGE    The charge was larger than 0x7FFF so
the charge precision is lost. The value is OK but the precision is
1024 ADC counts. Good for tracking, not good for dE/dx.

FCF_BROKEN_EDGE       This is the famous row8 cluster.
Flag will disappear from valid clusters
once I have the afterburner running.

FCF_DEAD_EDGE          Garbage and should be IGNORED!
This cluster touches either a bad pad or an end
of row or is somehow suspect. I need this flag for internal debugging
but the users should IGNORE those clusters!

        -- Tonko

TPC forum post 426:

I commited the "padrow 8" afterburner to the DAQ_TPX
CVS directory. If all runs well, you should see no more
peaks on padrow 8. The afteburner runs during the
DAQ_READER unpacking.

However, please pay attention to the cluster finder flags
which I mentioned in an email ago. Specifically:

"FCF_DEAD_EDGE          Garbage and should be IGNORED!
This cluster touches either a bad pad or an end
of row or is somehow suspect. I need this flag for internal debugging
but the users should IGNORE those clusters!"

         -- Tonko