Subject: Re: Muon digit container design From: David Adams Date: Tue, 09 Oct 2001 14:13:20 -0400 To: LAPORTE Jean-Francois CC: David Adams Thanks for your comments. I am in the process of updating the document today so thing may be a bit confused for another hour or so. Hits as produced by GEANT are outside the scope of my discussion. My example is one where pileup is simulated using digits rather than hits. Mixing hits is very important in other contexts but not relevant to my discussion. I may have been unclear but I meant that the background was taken from real data, not Monte Carlo. I have replace index with identifier to be consistent with ATLAS terminology. The digits (or digitizations) and their identifiers need to be precisely define by each subdetector group and their may be ambiguities to be resolved. I am only trying to provide an idea of some of the options. An ATLAS digit includes its identifier as well as the data. I suppose in your RPC example, the data might be a bool indicating whether or not this channel was hit. Again, I leave that to the RPC group. da ----------------------- LAPORTE Jean-Francois wrote: Dear David we have read your proposal on which we have few comments. We think that the concept of hit (energy deposition) is missing. This makes the discussion on pile-up in the section "Building a Monte Carlo event use case" somewhat confusing. The digitisation is the process that goes from hits (E dep) to digits (signal). The pile-up corresponds to the mixing of hits before applying this procedure. The key point is how to mix hits, and not how to mix digits as suggested in the above section. In the section "Definitions" index is defined as "a label for a single readout channel or time slice in such a channel". The later (time slice) is indeed important to identify uniquely a digit, for instance for the MDT case (many MDT digits could correspond to an unique readout channel). However, in the section "Indices" only single channel readout indices are discussed. We think that the meaning of "time slice in such a channel" should be clarified. Could a identifier of the digit be for instance "readout channel label" +"Rank Nber"? In addition, it seems to us that a digit is primarily formed by the index itself as for some digits, there is no associated data (yes or no channels, like RPC). So we question the separation between data and index of your proposal. Cheers Marc and Jean-Francois --------------------- On Wed, 3 Oct 2001, David Adams wrote: I have put a first pass at the muon digit container design. The classes are very preliminary. I would especially like to hear comments on the requirements and design choices. Please see http://www.usatlas.bnl.gov/~dladams/mudigi da -- David Adams desk: 631-344-6049 Brookhaven National Lab fax: 631-344-5078 PAS group, Building 510A email: dladams@bnl.gov Upton, NY 11973-5000 http://www.usatlas.bnl.gov/~dladams ___________________________________________________________________ This mail has been sent to all members of the list atlas-reconstruction ___________________________________________________________________ -- David Adams desk: 631-344-6049 Brookhaven National Lab fax: 631-344-5078 PAS group, Building 510A email: dladams@bnl.gov Upton, NY 11973-5000 http://www.usatlas.bnl.gov/~dladams