For POINTED LINE OBSERVATIONS:

EXTENSIONS:
-----------
0:  header
1:  flux  (Jy)
2:  wavelength  (um)
3:  flux, nod a (telescope units)
4:  flux, nod b (telescope units)
5:  RA, nod a
6:  Dec, nod a
7:  RA, nod b
8:  Dec, nod b



For POINTED SED OBSERVATIONS:
EXTENSIONS:
-----------
0:  header
1:  flux  (Jy, os=1, no flatfield)
2:  wavelength  (um)
3:  flux, nod a (telescope units, os=1, no flatfield)
4:  flux, nod b (telescope units, os=1, no flatfield)
5:  RA, nod a 
6:  Dec, nod a 
7:  RA, nod b
8:  Dec, nod b
9:  Flux (Jy, os=2, no flatfield)
10: wavelength (os=2, no flatfield)
11: flux, nod a (os=2, no flatfield)
12: flux, nod b (os=2, no flatfield)
13: Flux (Jy, os=1,  flatfield)
14: wavelength (os=1, flatfield)
15: flux, nod a (os=1, flatfield)
16: flux, nod b (os=1, flatfield)
17:  Flux (Jy, os=2, flatfield)
18: wavelength (os=2,  flatfield)
19: flux, nod a (os=2, flatfield)
20: flux, nod b (os=2,  flatfield)


SETTINGS:
---------
For line scan mode, the spectrum is binned within HIPE such that the spectral resolution is oversampled (os) by about a factor of 4.  

For pointed SED observations, I have included a range of possibilities because different reductions yield different results.  The unflatfielded observations with os=1 is my own default because sometimes the flatfielding can mess up the flux.  However, for some observations the flatfielding significantly improves the S/N.  Using os=2 can improve the resolution, but sometimes introduces NaN that can be difficult to deal with.  You should choose which flux to use based on your science goals.  However, the flux in the flatfielded spectrum may still be unreliable!

upsample=1 for all observations.



FLUX CALCULATIONS:
-----------
The flux (extension 1) is calculated based on the telescopic background as follows:

fx = (f_a+f_b)/2., where f_a and f_b are the flux from the two nods, in telescope units
fy = fx/(1.-fx/2.)
Flux = fy * conv, where the conversion is obtained from the Neptune spectrum.  


FLUX ACCURACY:
-------------
The flux is accurate to 10% throughout most of the spectral range, 20% longward of 180um, and worse at edges where light from other orders contaminates the emission.


POTENTIAL BROAD FEATURES:
------------------
The Neptunian spectrum includes bumps at the following locations:

57um, bump
62.3um, bump:  definitely a detector effect
67um:  very slight bump
78um:  bump+negative bump
133.5um:  very slight bump
183um, 190um-210um:  negative bumps, likely real in the Neptune spectrum

In these regions, the RSRF is slightly uncertain (moreso at >180um).  I have done my best to correct for detector features in these regions so that there will not be any spurious bumps in your spectrum.  However, I cannot promise this.  If you think you have a solid state/ice feature in your spectrum, please contact me before publishing so that I can take a closer look at whether the feature may be explained as an RSRF problem. 

QUESTIONS/PROBLEMS:
--------------
Please send any feedback (questions/problems/suggestions) to gregoryh@mpe.mpg.de