[Oli, Jenne]

The times that we've been able to get locked today (including earlier when Tony was on shift) the violins have been quite high.  Unfortunately, it's that pair on ITMY that have a beatnote of period ~5 mins.  It turned out that the OMC_WHITENING checker saw that we were in the trough of that beatnote, and decided to enable the DCPD whitening so we could get to Observing.  However, as soon as the violins came out of that trough, we started saturating the DCPDs and had to use the OMC_LOCK guardian to REMOVE_WHITENING.

I took a look at the saturation checker (which is in ISC_library), and it has some logic for if the quick check (30 sec lookback) says that we might be close to saturation, go check more data.  However, the way the logic had been set up, rather than looking back at a longer chunk of data to find the max (what I think we want), it was just looking at the same length of data from 5 mins in the past.  Since that beatnote is about 5 mins long, this means that it was just checking the values at the last trough of the beatnote, saying things were going to be fine (narrator: they would not be fine) and sending us along to adding the DCPD whitening.

I changed the lookback for the extra check to be a full 5 mins, to hopefully gather information from a whole cycle of this beatnote.  I also lowered the threshold for which it does the extra check (now if the quick check thinks we are within 60% of saturating, it'll do the extra check.  Previously, it only did the extra check if we were within 80% of saturation). 

We'll have to see if this unneccessarily adds to how long it takes for us to escape the OMC_WHITENING state; I'm hopeful that it won't keep us here unncessesarily, but that it will prevent us from engaging the whitening before we're ready.  ISC_library has been checked in to the svn.