Minutes for the Terameeting of 10th November 2005

Present: Yannick, Henry, Emmanuel, Fred, Chiara, Jean-Christophe, Laurent, Mireille, Anis

Status T0003 (Laurent)

 The QualityFITS-in preprocessing stage has terminated, around 13000 images have been processed

 Fred is making tests to try to resolve the problem related to PGPLOT and LIBSPICA on the nodes which prevents the generation of the completeness graphs
 SPICA has been installed on pix 3,4 independently for testing purposes
 Mail from mix6 works now
 problem with reliability of scp not resolved; several possibilities under investigation. Around 0.5% of files are affected by this problem
 Once the LIBSPICA problem will be fixed, Laurent will start a test stack in ugriz for the D2-deep field so we can test the entire pipeline process

Data transfer to CADC (Jean-Christophe, Fred)
 Jean-Christophe and Laurent have modified the .ldac catalogues names so that they conform to ICD document (.ldac.fits).

 Jean-Christophe will send to JJ at CADC information about several new SCAMP output files that are not yet recognised by the ICD filename CADC checklist

 Following JJ suggestion, we will try to implement BITPIX=16 for weightmaps for the next release (T0004), but this requires extensive changes to Emmanuel's software to implement a double-pass mode; potential gain is a factor of two reduction in storage space
 It will take around one month to transfer all data (weight-maps and QF pages) to CADC
 At times the transfer takes around half of the total available bandpass of the IAP

Status of machines (Fred)
 mix7 should arrive today (it did!)
 Henry will test mix5 with writing coadded files to the system disk

Scamp (Emmanuel)
 Still working on problem of weighting for crowded fields with lots of stars
 Using sloan as a reference catalogue helps a lot for the astrometric solutions
 Would like to implement a "photometric reference" catalogue option in SCAMP
 Also need to add an option to photometrically decouple the CCDs (relaxing the assumption that the zero-points of all the CCDs are the same, which was true for MEGACAM but perhaps is not true for WIRCam, see below)

WIRCam (Chiara)

 Calibration has been re-done with correct scaling, and now magnitudes match with those with the 2MASS

 Still some problem with the counts, this is under investigation
 There seems to be shift of around 0.6 magnitudes between the two (engineering grade) detectors
 note that always the .ahead files files always have priority over the scamp.ahead in a given directory
 Work has started on new data with four detectors and 'science grade' data. There are 20 new images. But they were not acquired in photometric conditions and are only suitable for astrometric solutions.

PI data (Mireille, Henry, Yannick, Emmanuel, Fred)
 Mireille is working on some PI data for Loic/Jean-Charles which will be used to produce publicity images. The value of BACK_MESH however is currently too small
 COSMOS data in u* processed and shipped to Caltech
 ftp server should be started on ftpix and then tests made to see if it is visible from outside the IAP
 Fred will then put Seymour's data processed by Mireille on ftpix
 Still waiting for the old COSMOS data to be shipped from CFHT for future runs
 Skycat cannot properly load the catalogue for some fields which are around 0 hours; seems to be the same kind of problem affects SCAMP.

Skywatcher (Jean-Christophe)
 Packaging in progress

Web-service QF (Jean-Christophe)
 QF web-service disactivated for improvements. Jean-Christophe is implementing new options to take into account Chiara's, Henry's and Yannick's comments/suggestions.
 The web interface should include all credits carefully (Fred, AstroWise, Terapix, etc...).

Visualisation tool (Anis, Emmanuel)
 Learning OpenGL, Qt,
 Some work required to understand the FITS format!
 A decision still has to be made on whether we start from scratch for the software or try to build on PANORAPIX

Quality assessment for T0003 (Yannick)
 Mathias will possibly visit Terapix in the week of 21-25 November (to be decided with Mathias).
 Data products from T0003 will be made available at the beginning of January
 T0003 will be released world-wide in February 2007, so we should get it right!
 Dust and mask files will be created in advance, using the stack list generated by Laurent prior to run SPICA, from the RA,DEC positions of all stacks. Those already created for T0002 fields will be simply duplicated and renamed accordingly.

Other News
 Anthony's Thesis will be co-directed with Albert Bijaoui in Nice
 Emmanuel informs Terapix there will be a workshop organised by Lupton, Gunn, Bertin, Magnier (others?) in Dubrovnic about "Astronomical and Image Processing" (probably between Sept. 4 and 9, 2006).

Next Tera-Meeting: