Inconsistency between seed pixel and cluster time
I'm currently looking into the clustering of our Timepix3 data from DESY.
I stumbled upon the following:
- cluster time = earliest pixel in the cluster
- seed pixel = pixel with largest ToT in the cluster
These two are not always the same. That's an issue e.g. when one wants to perform a timewalk correction based on the seed pixel ToT but apply it to the cluster timestamp. Along the same lines, @dannheim mentioned that @fpipper obtained a slightly better time resolution when setting the cluster timestamp to that of the pixel with largest ToT.
Another thing: If no pixel ToT is available (in other words, it's always 0), cluster->getSeedPixel()
will return a random pixel from the cluster as can be seen here:
https://gitlab.cern.ch/corryvreckan/corryvreckan/-/blob/master/src/objects/Cluster.cpp#L52
What do you think about the following:
Clustering4D
:
- new parameter to choose between earliest pixel and highest ToT pixel (with the first being the default)
Cluster
:
-
getSeedPixel()
: if pixel ToT is always zero, return earliest pixel