[Issues] Sextractor config for SourceList creation
Ewout M. Helmich
helmich at astro.rug.nl
Tue Jan 23 11:45:28 CET 2007
Hi,
I have changed SourceList so it sets the GAIN and MAG_ZEROPOINT
Sextractor configuration parameters correctly for RegriddedFrames.
In the case of ReducedScienceFrame it is left to the user to set these
values in the configuration, because it is not clearly defined which
values to take; a PhotometricParameters object (containing zeropoint)
may or may not exist for the ReducedScienceFrame, it's also possible
that versions newer than the ReducedScienceFrame in question exist of
this object or of GainLinearity.
Gijs has discussed the case of CoaddedRegriddedFrames.
Regards,
Ewout
Gijs Verdoes Kleijn wrote:
> Dear Hans,
>
> An update on the photometry issues:
> -regriddedframes: Ewout will change soon the implementation of
> SourceList such that it will use the appropriate zeropoint for these
> frames which will lead to correct magnitudes/fluxes and their
> associated errors. We will let you know when this has happened.
> -coaddedregriddedframe: as you know SourceList no determines properly
> the magnitudes/fluxes by default. Bow for errors in them: thinking
> more about it myself and talking to Emmanuel further on this I now
> understand there is no good way now to obtain good error estimates
> from a coaddedregriddedframe which is the product of >1
> regriddedframes which each have their own zeropoints (hence FLXSCALE
> values). The information needed to do it is simply missing. A map
> detailing the resulting GAIN for each pixel would be a solution, but
> Emmanuel has no time to implement this in the near future. For the
> time being I therefore reach the conclusion that one has to decide
> oneself how to obtain approximate errors best for
> coaddedregriddedframes made out of >1 regriddedframes. Obviously, I
> can image that one uses some average from the GAIN values of the
> RegriddedFrames.
>
> If I come up with a better solution I'll let you know.
> Best wishes, Gijs
>
> verdoes at astro.rug.nl wrote:
>> Dear Hans,
>>
>> As one step in the process to solve the current issue I have determined
>> how the coaddition is done algorithmically. I have added this
>> information
>> in the how to on coadd.
>>
>> Ewout and I are working further on solving the issue.
>>
>> best wishes, Gijs
>>
>>> Dear Gijs,
>>>
>>> thanks for your response. As far as I can see the FLXSCALE is not
>>> computed
>>> for
>>> CoaddedRegriddedFrames at the moment. The attribute does not even exist
>>> for
>>> this class. I agree that using this number, once it is available and
>>> correctly computed for the CoaddedRegriddedFrames, would solve the
>>> problem.
>>>
>>> Best wishes,
>>> Hans
>>>
>>> On Wednesday 13 December 2006 08:18, Gijs Verdoes Kleijn wrote:
>>>
>>>> Dear Hans,
>>>>
>>>> Thanks for the report. I would like to add information on how
>>>> photometry
>>>> is *handled* in Astro-WISE to the HOW-TOs which currently tell how
>>>> photometry is derived. For this reason I am happy to help out in
>>>> resolving the remaining issue below. Currently the system is down
>>>> here,
>>>> so I cannot check the following statement: aren't the FLXSCALE of the
>>>> RegriddedFrame and of the corresponding CoaddedRegriddedFrame
>>>> identical
>>>> in your test below? If so, I would presume that you should be able to
>>>> use the FLXSCALE of any CoaddedRegriddedFrame (e.g., one composed of
>>>> many RegriddedFrames) to set GAIN. But we should carefully check how
>>>> this FLXSCALE is computed from all the different FLXSCALEs of all the
>>>> RegriddedFrames that went into the CoaddedRegriddedFrame.
>>>>
>>>> Best wishes, Gijs
>>>>
>>>> Hans Vaith wrote
>>>>
>>>>
>>>>> This is a followup to the problem that we (MPE) reported at the
>>>>>
>>>> AstroWise
>>>>
>>>>> workshop concerning some parameters in SourceLists created from
>>>>> ReducedScienceFrames, RegriddedFrames and CoaddedRegriddedFrames:
>>>>>
>>>>> 1) When creating SourceLists from Reduced or Regridded Frames the
>>>>> zeropoint is not applied. This can be fixed easily by setting the
>>>>> sex.conf parameter MAG_ZEROPOINT to the zeropoint as returned by
>>>>> frame.photom_params.get_specific_zeropoint(...)
>>>>>
>>>>> 2) When creating SourceLists from a CoaddedRegriddedFrame, the errors
>>>>>
>>>> of
>>>>
>>>>> the fluxes and magnitudes are way too high. In an email conversation
>>>>>
>>>> E.
>>>>
>>>>> Bertin pointed out that this is caused by not setting the GAIN
>>>>>
>>>> parameter
>>>>
>>>>> in sex.conf.
>>>>>
>>>>> I have experimented with a CoaddedRegriddedFrame that was made from a
>>>>> single RegriddedFrame. Creating the source catalog by running
>>>>>
>>>> Sextractor
>>>>
>>>>> manually, I set the GAIN in sex.conf to the inverse of the FLXSCALE
>>>>> attribute of the RegriddedFrame. This way one gets the same error for
>>>>>
>>>> the
>>>>
>>>>> magnitude as in the case of creating the catalog from the
>>>>>
>>>> RegriddedFrame
>>>>
>>>>> directly. The flux error is now also consistent with the result from
>>>>>
>>>> the
>>>>
>>>>> RegriddedFrame (consistent meaning that the S/N ratio flux/flux_error
>>>>>
>>>> is
>>>>
>>>>> the same). How to handle this in the case of more than one
>>>>>
>>>> RegriddedFrame
>>>>
>>>>> being coadded I don't know at the moment (how to compute an overall
>>>>> fluxscale from the individual numbers)
>>>>>
>>>>> For reference, I appended the response that I got from Emmanuel
>>>>> Bertin
>>>>> after
>>>>>
>>>>> sending him some data after the workshop:
>>>>>
>>>>>> I had a look at your data, and I
>>>>>> noticed that the co-added image are scaled with a very small number,
>>>>>> something like 10e-11!! That should not cause much trouble except
>>>>>> that
>>>>>> you forgot to scale the GAIN (aka conversion factor) accordingly,
>>>>>>
>>>> that
>>>>
>>>>>> is by something like several billions e-/ADU. If you set the proper
>>>>>> GAIN, you will get correct results, although I suggest you set the
>>>>>>
>>>> image
>>>>
>>>>>> zero-point to a more convenient value (say, 30 or so), and use a
>>>>>> non-zero the MAG_ZEROPOINT SExtractor configuration parameter. Best
>>>>>> regards,
>>>>>> Emmanuel.
>>>>>>
>>>>> Dear Emmanuel,
>>>>>
>>>>> I contacted you last Monday during the AstroWise workshop in Leiden
>>>>> concerning a problem that we have with understanding the output of
>>>>> Sextractor when running it to create a catalog from a
>>>>> CoaddedRegridded
>>>>> image. The problem is that the errors in the fluxes and magnitudes
>>>>>
>>>> appear
>>>>
>>>>> to be way too high. This happens only with the coadded image, whereas
>>>>>
>>>> the
>>>>
>>>>> catalog created from the associated regridded image looks fine.
>>>>>
>>>>> There is a zip file containing the images, weights (regridded and
>>>>>
>>>> coadded
>>>>
>>>>> frames) and configuration files that we have been using when we ran
>>>>> Sextractor. This file is stored on "ftp.mpe.mpg.de" and can be
>>>>> accessed
>>>>> via ftp. The filename is "for_bertin.zip". To retrieve the file
>>>>> please
>>>>> login as user "anonymous" and use your email address for the
>>>>> password.
>>>>> Then change to directory "/pub/edi_esoc_dds/".
>>>>>
>>>>> Our Sextractor version is 2.4.4
>>>>>
>>>>> To give you an idea about the numbers, I listed magnitudes, fluxes
>>>>> and
>>>>> their associated errors for a single source that is contained in the
>>>>> output catalogs:
>>>>>
>>>>> regridded coadded
>>>>> MAG_ISO 19.165 19.165
>>>>> MAGERR_ISO 0.009 7338.7
>>>>> FLUX_ISO 20819 2.16e-8
>>>>> FLUXERR_ISO 164.3 1.46e-4
>>>>>
>>>>>
>>>>> Your help is much appreciated.
>>>>>
>>>>> Best Regards
>>>>> Hans Vaith
>>>>> _______________________________________________
>>>>> Issues mailing list
>>>>> Issues at astro-wise.org
>>>>> http://listman.astro-wise.org/mailman/listinfo/issues
>>>>>
>>> --
>>> _______________________________________________________________
>>> Hans Vaith Max-Planck-Institut fuer extraterrestrische Physik
>>> Postfach 1312, D-85741 Garching, Germany
>>> Phone :(Country Code 49) 89 30000-3831
>>> e-mail: hav at mpe.mpg.de (Fax: 89/30000-3569)
>>>
>>>
>>
>>
>> _______________________________________________
>> Issues mailing list
>> Issues at astro-wise.org
>> http://listman.astro-wise.org/mailman/listinfo/issues
>>
>
> _______________________________________________
> Issues mailing list
> Issues at astro-wise.org
> http://listman.astro-wise.org/mailman/listinfo/issues
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Drs. Ewout Helmich <><>
Kapteyn Astronomical Institute <><> Astro-Wise/OmegaCEN
Landleven 12 <><>
P.O.Box 800 <><> email: helmich at astro.rug.nl
9700 AV Groningen <><> tel : +31(0)503634548
The Netherlands <><>
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
More information about the Issues
mailing list