Minutes for the teraminutes of February 12th 2007

Present: Gregory, Emmanuel, Chiara, Yannick, Mireille, Fred, Mathias.

Emmanuel's summary document

Processing T04

 Yannick has begun to grade the new CFHTLS QualityFITSin data (2300 exposures). Laurent will reprocess the data graded "D".
 Mireille will grade L01 and L04 (Deep), Yannick L02 and L05 (Wide). L03 and L06 will be graded later on.
 Yannick will check whether astrometric offsets and correlation coefficients are reported in QualityFITS.
 The pipeline is interrupted whenever a flat is not remotely available for instance; it would be much more efficient to put the images that cause problems aside and go on on the next images. The sorting of file processing seems unpredictable. How is the order in which data are processed determined?
 Fred mentions an incompleteness problem on the new CADC proxy; He will contact them about the problem.
 According to Fred, all the QualityFITSin for T04 will be completed this week.
 The T04 release will contain many more Wide fields (including the new W4). Two kinds of Deep stacks will be produced, with different thresholds in quality (25% and 85% of the best seeing images)
 The measurement of elongations is lacking in PSFEx. Emmanuel will try to add it when he finds the time.
 It is decided that all the QualityFITS will have to be redone for T05, to benefit from improvements in diagnostics and masking.

PI data processing

 Mireille is surprised by the presence of particularly bad data in the list of files downloaded by JC's tool. Yannick explains that the PIs are not necessarily aware of the existence of these bad quality data (they are filtered out by CFHT on the metadata DVDs the PIs receive).
 Some discussion goes on about the content of Datatransfer: Mireille has looked into the status of PI data. Yannick will also have a look for non-PI data and tell Laurent what must be kept.

Photometry

 Mathias mentions that the WIRCAM curves still lack proper information about the detector and optical transmission curves. This point should be raised later this week during the teleconference with Loic.
 Mathias will compute the components of typical extinction vectors with the new filter curves.

Quality Assessment

 Yannick, Gregory, Fred and Emmanuel will meet with Laurent Monday afternoon and/or Tuesday morning to discuss about automatic statistics for quality assessment.
 The WIRCAM teleconference will occur on Thursday evening. Chiara is not sure to be able to have enough statistics ready because of all the work she has to do until then [edit: she has already completed some statistics on PSF quality].

Other matters

 Emmanuel will contact Edwin Valentijn to propose him to hire Anthony on Panorapix for 4 months.
 CSA: 15kEuro for TERAPIX and 15kEuro for WIRCAM in top priority; PNC: 7kEURO for TERAPIX.
 MDA (Strasbourg): nothing new.
 Ultravista: Fred will send to Henry a quotation for 2 dual core machines, a UPS and a rack.

WIRCAM

 Chiara asks for a background computation option with object masking in QualityFITS. Fred will do it.
 Some WIRCAM PIs ask for many more tests than other, and cause more work for Chiara. Chiara mentions that up to now this extra work proved to be relevant for the improvement of processing quality, but this may not always be the case in the future. In case of priority conflicts Chiara may refer to Yannick to decide what task shall be completed first.

The next terameeting is planned for Monday 19th at 10:30am (not confirmed yet).