ObsCore: o_ucd for uncalibrated data?

[Frederic V. Hessman]

Gentlemen,

ADU’s are not fluxes (ADU/s/pix^2 might be) but simply uncalibrated counts, so I’d go with

	phot.count;stat.uncalib

However, “phot” means “photometry”, not “photons”, and ADU’s are just a different unit of photometric counting, so 

	phot.count

(some number of photometric events) wouldn’t be so bad after all.  More metadata is usually better, but one has to stop somewhere….

RIck

On 28 Jul 2015, at 08:37, Marco Molinaro <molinaro at oats.inaf.it> wrote:

> Hi Alberto,
> not an expert on this, but would
> 
> phot.flux;stat.uncalib
> 
> fit your case?
> I added Semantics to the reply, just in case some relevant people were
> not on the dm list.
> 
> Cheers,
>    Marco
> 
> 
> 2015-07-24 16:35 GMT+02:00 Alberto Micol <amicol.ivoa at googlemail.com>:
>> 
>> Thanks Petr,
>> 
>> Unforunately phot.count does not seem correct...
>> phot.count represents the number of photons counted
>> by the instrument, while what I'm talking about is
>> really instrumental data numbers (ADUs) which are far
>> from being the number of photons.
>> 
>> That is, I cannot take the number of ADUs, multiply
>> it by the energy of a photon (which I can derive from the wavelength info
>> of obscore) to get a representative flux...
>> 
>> That is why I would like to have a phot.uncalibrated,
>> or phot.adu, or similar... any other idea?
>> 
>> Thanks,
>> Alberto
>> 
>> On 24 Jul 2015, at 16:17, Petr Skoda wrote:
>> 
>> On Wed, 22 Jul 2015, Alberto Micol wrote:
>> 
>> 
>> Dear ObsCorers,
>> 
>> 
>> At ESO we are serving spectra very well calibrated in wavelength, but
>> uncalibrated in flux.
>> 
>> What would be the correct o_ucd value for such data?
>> 
>> I cannot find a UCD that could fit… something like: phot.uncalibrated would
>> be nice to have…
>> 
>> Any idea?
>> 
>> 
>> Hi Alberto,
>> 
>> In stellar astronomy we formally describe the vertical axis as integrated
>> flux in ADU or counts ....
>> It is in principle produced by summing all pixels values in given wavelength
>> but modified by various sensitivies of detetror (flat-field etc) . So
>> perhaps something like flux in counts may be used??
>> So UCD="phot.count"
>> ?????
>> 
>> 
>> Than the problem comes with continuum normalization made by some pipelines.
>> For this was already a suggestion by Mirellei for SDM
>> 
>> Char.FluxAxis.CalibrationStatus='NORMALIZED'
>> Char.FluxAxis.unit=NULL  or 'unitless'
>> Char.FluxAxis.ucd=phot.flux;arith.ratio;
>> 
>> 
>> Is there any other suggestion ?
>> 
>> Just my practical view for intended usage of obscore:  I am either
>> interested in getting SOME form of spectra of my object - so I am perfectly
>> happy with the RAW extracted spectrum - in counts (or ADUs) or in a SPECIFIC
>> form for analysis - than I prefer the FLUXCALIB=normalized or ABSOLUTE - if
>> I already know about many spectra of the same object in given form. But more
>> flexible approarch is to use DataLink.
>> 
>> 
>> OTOH - it is not sometimes obvious the spectra are normalized to continum
>> without visual inspection - you see the most values are at level 1.0 and
>> there is not visible skew of the spectra ....
>> 
>> BTW - I am glad that the ESO is still considering the VO as a viable
>> technology for presenting data ....
>> 
>> *************************************************************************
>> *  Petr Skoda                         Phone : +420-323-649201, ext. 361 *
>> *  Stellar Department                         +420-323-620361           *
>> *  Astronomical Institute AS CR       Fax   : +420-323-620250           *
>> *  251 65 Ondrejov                    e-mail: skoda at sunstel.asu.cas.cz  *
>> *  Czech Republic                                                       *
>> *************************************************************************
>> 
>>