[daip] FARS issue

[Eric Greisen]

On 01/01/2015 04:06 PM, Lawrence Rudnick wrote:
> Perhaps have isolated the problem.
>
> I haven't checked output in detail yet, but looks like the procedure has
> worked with only changing blc,trc before running DOFARS.
> in the run that failed, i set blc, trc to 3341 3361; 5421 5601 before
> running DOFARS.  when it actually gets to the FARS stage, DOFARS
> apparently has set the BLC, TRC to 1 0 0, 1 0 0.   the output maps have
> the correct dimensions
> 301,2081,1941 , but the output is garbage.
>
> if instead i set original blc, trc to 0,0, then when DOFARS gets to the
> FARS stage, it has blc,trc 0,0

The Inputs to DOFARS would be interesting.  I believe they were
blc = 3341,3361,1
trc = 5421,5601,1
and the 1's were used to set set BLC(1), TRC(1) at the FARS step.  And 
of course FARS cannot work with only 1 frequency.  The BLC(3), TRC(3) in 
DOFARS is used to restrict the frequencies at the FARS step.  That is 
wrong however because the frequencies were already restricted at the 
initial TRANS which uses the full (all 7) values of BLC and TRC.

I will fix DOFARS to not use BLC(3), TRC(3) after TRANS because they 
have already been applied.  First patch of 2015...

sigh

Eric