Notes from Keith concerning version 3 treatment of AS_Q outliers 8/29/03:

The large AS_Q outliers identified by Gaby in H1
and L1 have been identified by three different
quality flags:

ASQ_LARGEP2P:

  This was first used in version 2 DQ segments and applies to only
  time intervals at the ends of lock stretches. Relevant to only L1 so far.

ASQ_OUTLIER_CLUSTER:

  This was first used in version 3 DQ segments and applies to only
  clusters of outliers in AS_Q, regardless of activity in auxiliary
  channels. Relevant to only H1 so far.

ASQ_OUTLIER_CORRELATED 
                     
  This was first used in version 3 DQ segments applies to outliers
  in AS_Q correlated with outliers in auxiliary IFO channels.            
  This flag is mildly troubling because (as yet) there has been
  no study of the deadtime introduced by vetoing intervals with
  outliers in auxiliary channels regardless of AS_Q behavior.
  See the exchange between PeterSh and Gaby on this subject.
  Relevant to H1 and L1 so far.

  --> Please use with care (or don't use!) if searching for very strong
      gravitational wave transient signals