This page is intended for discussions on the draft: Second Version of the OI-FITS format

---

The Document in its current state:

Draft for version 2 available here: http://www.jmmc.fr/twiki/pub/Jmmc/OIFITSTwoProject/oifits_norm_v2.pdf!*

The historical discussion on this format is here http://www.jmmc.fr/twiki/bin/view/Jmmc/JmmcOldPageOnOifitsTwoProject

This page is free for your comments.

1. You may need to register to enter coments.
2. Do not forget to sign them using the signature button!

Comment by AntoineMerand

• Visibility and correlated flux: do I understand correctly that UNITS (table 1) apply to all data of the HDU? if so, can you handle one OI_VIS holding visibilities and the correlated fluxes? If I understand well, you cannot, so you need 2 distinct OI_VIS tables. In practice, it should be easy since I expect instruments providing correlated fluxes in OI_VIS, and visibilities in OI_VIS2 (no OI_VIS). Maybe would be worth to detail how correlated fluxes should be stored.

• MP and MB: "Ratio of input over output apertures” and "time-varying component of the IFOV”. I agree with Gilles, MB should not be in the table, or only optional. Do we have existing or plan instruments with such feature?

• concerning OI_WAVELENGTH:
  • why should we have EFF_BAND set to NULL in the case monochromatic values? I think there should always be a EFF_BAND, I cannot think of a realistic monochromatic case…
  • "These estimates should include the effect of the earth’s atmosphere, but not the spectrum of the target (the effect of the target spectrum should be taken into account as part of any model-fitting/mapping process, i.e. the target spectrum is part of the model).” what does that mean? do yo mean that if the target is red/blue shifted, the wavelength should not be in the rest referential of the target?

• for differential quantities in OI_VIS: "VISREFMAP must be present if any of VISTYP or PHITYP is "differential". If VISREFMAP is absent, neither can be differential.”: isn’t this too strict? Same could be said in the case visibility (or vis2) are binned.

• concerning OI_SPECTRUM:
  • should be restricted to spectra derived with the interferometric instrument, not to third party instruments! otherwise, why not have an OI_PHOTOMETRY as well? Few interferometric instruments are suitable to extract photometry, but MIDI for instance, is pretty good at it…
  • I do not understand the meaning of “CALIBRATED”: I think the relevant information should be: is it raw, normalized, photometrically calibrated? Another important information is wether or not the telluric lines were removed.
  • I do not understand either the usefulness of the following: "If the boolean keyword header CALIBRATED is FALSE, the STA_INDEX column must be present, and the flux stored in FLUXDATA is the flux measured on the telescope indicated by STA_INDEX."

• some things not covered by the document:
  • Astrometry: GRAVITY will not be able to save its reduce data in the OIFITS format. I am not sure it is really possible, but maybe there should be a foreword stating that astrometry has been excluded.
  • Error correlations due to common calibrators for OI_VIS2. Each calibrated observation should give the name of the 2 calibrators used for calibration, as well as their visibilities. See http://www.aanda.org/articles/aa/pdf/2003/12/aa2990.pdf for a description of how correlations can be used.

Comment by PierreCruzalebes

Dear colleagues, Since optical long-baseline interferometric data can be processed using the instantaneous spectral distribution of the complex triple product, as is it done for instance with the SPIDAST software, I suggest to add to the OI_T3 data table the real and imaginary parts of the triple product with their errors. Staying at your disposal, with my best regards.