Please note that the intervals flagged as showing strongly
coherent noise between H1 and H2 should be treated with caution.
These intervals should be excluded in the mainstream stochastic analysis
search because their source is inconsistent with ordinary, stationary
stochastic GW background. But in other searches, one should be careful
that a putative source could not produce the coherence seen in 
H1 and H2. As indicated in Vuk's accompanying document, most of
these periods are associated with the last few seconds before a
lock is lost, but some are associated with hardware signal injections!

Use with care!

K. Riles 4/28/05