Re: Analyzing Data

[Tom Droege <tdroege2@earthlink.net>]

Michael,

In general, I would like to "flag" the defects and continue processing - at 
least for the pipeline.  Then in between pipeline steps each of us can 
decide what to discard.  Processing does not take so long that we cannot 
afford to process everything.  I do plan to dump dawn and dusk and bad 
cloud data early on.  But much of the rest can be kept.

I would prefer you to add bits to the existing flag word in the .cal file 
to mark the defect.  Later I will add the "Clear Sky" value in several 
bits, and the Clumpiness in another bit.  This will allow us to later pick 
out the "best data" by the criteria of the day.

Tom Droege

At 04:12 PM 2/8/03 -0500, you wrote:
>Perhaps Tom and Arne will be happy to hear that I have _just_ modified
>the Mark IV pipeline so that one can use methods a, c, or d to flag
>a frame as "bad", thus discarding it from further processing.  I hope
>to release the new version before Monday.  I still need to modify
>the "photom" program to include an option to account for differential
>extinction within each frame using fixed coefficient values (since
>ignoring differential extinction is significant in V-band, at least).