From bennett@hpmvd.lanl.gov Thu Dec 3 09:06:42 MST 1998 Received: from hpmvd.lanl.gov (bennett@hpmvd.lanl.gov [128.165.86.56]) by p2hp4.lanl.gov (8.8.6 (PHNE_14041)/8.8.6) with ESMTP id JAA00482 for ; Thu, 3 Dec 1998 09:06:41 -0700 (MST) Received: (from bennett@localhost) by hpmvd.lanl.gov (8.8.6 (PHNE_14041)/8.8.6) id JAA27612; Thu, 3 Dec 1998 09:06:42 -0700 (MST) Date: Thu, 3 Dec 1998 09:06:42 -0700 (MST) From: Michael Bennett Message-Id: <199812031606.JAA27612@hpmvd.lanl.gov> To: wohn@iastate.edu, sullivan@p2hp4.lanl.gov, young@mail.phy.ornl.gov Subject: Re: mvd and level 1 Cc: petridis@iastate.edu, schlei@lanl.gov, jsimon@lanl.gov, hubert@lanl.gov, sullivan@p2hp4.lanl.gov Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-MD5: Ef4bwZZrF/9LInx1BJkJTQ== Status: RO Hello all, In response to Glenn's earlier mail on offline corrections, attached below, I agree that an MVD multiplicity trigger uncorrected for vertex leads to a messy smearing of actual accepted multiplicity. In terms of normalizing cross sections from this trigger, I don't see having to make an offline correction as such a big deal. Since the acceptance for various processes almost certainly varies with vertex position, one would necessarily be applying a vertex correction anyway and adding a cut on an offline-corrected multiplicity is probably not so difficult. In the end, one can choose a corrected set of 10% central events which contain the process of interest, and arrive at a reliable cross section. However, I would be more concerned about a vertex-biased MVD LVL1 as an input to higher level triggers. Say, for example, one wanted to trigger on a high energy hit in the EMCal; presumably this trigger would normally also require a LVL1 trigger to have been formed. Now the EMCal trigger also carries a vertex bias due to acceptance; if the input LVL1 trigger from MVD also carries a (different) vertex bias, it seems to me that unfolding all of that to arrive at a reliable cross section becomes much more difficult. I would think that anything we can do to make the LVL1 input to higher triggers as unbiased as possible would be worthwhile. So, I would agree with John S.'s earlier message, that while an uncorrected MVD LVL1 is probably OK to start, it would be a mistake to design away the ability to make a vertex correction to the MVD LVL1 in the long run. Regards, Mike > > Dear MVD/LVL1 folks, > A question arising from a sudden bout of nervousness, upon reading > JohnS' note to Fred: > > If John finds a multiplicity signal that is 13 +-4% smaller for a Z > vertex at +-40cm compared to one at the nominal Z=0, then if > we implemented a 'centrality' trigger on such information that took > the supposed 10% most central events, wouldn't we end up with a rather > messy smearing of the resultant event sample in terms of actual multiplicity? > > Said differently, not having a z-vertex correction to take out such > "10%-level" effects sounds like a considerable worry, yes? > > If I can get the correction down to 1-2%, then I'm happy to make > an off-line cleanup cut. The choice of "1-2%" is subject to the usual > debate, of course. > > Regards, Glenn ************************************************************ Mike Bennett P-25 Mail Stop H846 Los Alamos National Laboratory Los Alamos, NM 87545 phone: 505-665-2760 fax: 505-665-7920 email: mjbennett@lanl.gov