[Heig] HEIG Obscore Note - ready for final review/ comments on the Note #part 1

[Karl Kosack]

> On Aug 25, 2026, at 21:22, Dr. Ian N. Evans via heig <heig at ivoa.net> wrote:
> 
> Hi Mireille,
> 
> I provided some comments/answers in-line below.  Let me know if you need more information.
> 
> 
>> On Aug 24, 2026, at 12:52, Mireille Louys via heig <heig at ivoa.net> wrote:
>> 
>> Hi Janet , Hi HEIG members,
>> Here are my comments about the last version of the note from August 03. 
>> I just want to mention that july-august is the summer leave period for people working at the university, 
>> that is why I answer now. 
>> I will try to put my remarks and suggestions in Github , but I cannot put comments on the current version via the web interface currently. 
>> Here are my remarks: 
>> I believe some features need to be improved before submission :
>> Terms for data products : 
>> we have used advanced data products and analysis data products to express, I suppose new elaborated products that are more easy 
>> to the user to take a decision, in data selection or in data interpretation. 
>> 
>> I think we must provide a clear definition for them. 
>> I believe both terms are fuzzy: 
>> Why "advanced"? with respect to today’s available data products? 
>>  What will it mean in 2 years ? 
>> Are some advanced products not used for science analysis? dedicated to data selection only? 
> 
> We use the term “advanced data products” throughout the document, simply because that is the term used in the existing ObsCore Recommendation.  See for example, footnote 8 on page 11 of the latest draft.  We don’t in general use the term “analysis data products” as the usage of the latter term imputes a calib_level value that may not apply.  The one exception to this is in Table 2, where we reference a proposed vocabulary name (and even here we again note in a footnote that “advanced data products” is more appropriate).  

That might also come indirectly from CTAO nomenclature, where we call data products in our data level DL5 "advanced science products", which are just data products that are at a high calibration level such that they are no longer in instrumental units (e.g. images/cubes/spectra/lightcurves in astrophysical photon flux units, rather than detector counts), and can thus be interpreted or compared with results from other instruments without having to think about IRFs (mostly).  But that is somewhat specific to our data levels, so dropping the qualifier as Ian suggests is probably ok.