Recent WIRCam News

UNDEFINED lines in FITS headers

From March to October 2016, about 500 images had an UNDEFINED line mixed into the FITS header. It was sometimes inserted between two lines or replaced one valid line. We traced it back to the implementation of the Dome Vents FITS keywords, and due to a race condition in the header capture. It has been corrected on October 11, 2016.

Changes in the astrometry of WIRCam images

On September 28, 2016, the two-step astrometry of WIRCam images was switched off and reduced to a simple mosaic-level astrometry. The second step consisted in refining the astrometry at the chip level, but it was sometimes failing in crowded fields, resulting in badly displaced chips with respect to the mosaic. At the same time, a small error introduced with the new flat-scaling scheme was corrected. This error generated small steps between the chips and was detected by Loic Albert.

Vertical noise bands

During the April 2016 run 16AQ10=16Aw03, a vertical noise pattern appeared due to grounding problems. We implemented on May 26, 2016 a noise filter in both axes, and this removed the pattern. It has however been noticed that this correction affects large objects.

Charging overheads of WDP mode

Starting with Semester 17A, WDP mode overheads will be charged to the PI, at 30sec per exposure (vs 10sec previously). This will allow to better reproduce the overheads at the telescope and improve the planning of the observations.

New CO and W filters

On May 15, 2015, a new CO filter was added to the WIRCam collection. In July 2015, a new W filter was also added, intermediate between J and H. Both filters have been characterized, and zero-points are measured on a regular basis.

Change of zero-points definition

On December 10, 2014, the definition of zero-points has been modified. Rather than using an average value over the 4 chips, we have decided to adopt the value corresponding to chip #4 (detector #60). The average value was contaminated by the poor knowledge of the chip-to-chip offsets. All the calibrations since 2011 were re-reduced to ensure an homogeneous definition of zero-points. The new flat-scaling scheme was implemented on September 18, 2015.

New non-linearity correction scheme

During 2014, An-Yi Bu and Wei-Hao Wang reduced new and better observations (using a stable lamp rather than the dome floor) to determine the non-linearity corrections. A new scheme was implemented on July 4, 2015, using pixel-to-pixel corrections rather than a median correction.

New reduction machines

On May 30, 2014, new reduction machines (polena and konane, from 2009) replaced older machines (halea and ula, from 2005), previously in use.

New WIRCam instrument scientist

After 3 months without an official astronomer in charge of WIRCam, a new WIRCam instrument scientist finally arrived at CFHT on Dec 7, 2013.

WIRCam webpages updated (Oct 31 2011)

WIRCam webpages have been updated with recent additions and improvements to the instrument as well as the processing pipeline, `I`iwi. Documentation regarding the current version of the production pipeline, `I`iwi ver 2.2 have also been updated and are available online.

`I`iwi Version 1 Documentation Online (Mar 25 2008)

The `I`iwi Documentation and recipes can now be found on the WIRCam web page. It describes most steps of version 1 of the pipeline.

Massive preprocessing (Sep 1 2007)

Preprocessing of all past semesters data with the first stable release of the IDL Interpretor of the WIRCam images (`I`iwi) has started. More details about the `I`iwi will be posted on the main WIRCam page soon. Priority is to preprocess all programs of the 06B and 07A semesters first then move back to the older data.

Camera News and Non-Linearity Correction (July 23 2007)

As expected, the new hardware setup for WIRCam is producing very encouraging results. No apparent positive crosstalk can be seen and the negative crosstalk has disappeared entirely. The edge crosstalk is still present but its effects are not as strong as that of negative crosstalk and we cannot seem to reproduce its effects in the lab. So more work will be required to understand it.

Meanwhile, shallow stacks (1000 sec) have shown that the technique to remove edge crosstalk effects in software is effective with very little residuals. Deep stacks will be produced to see what the residual levels are for fields with long integration time.

Good news, the non-linearity characterization for data taken since semesters 06B has now been finalized. The correction has also been included in the pipeline. At high flux, the linearization process increases the flux by up to 10%. We are in the process of reprocessing the standard stars measurements to see what will be the impact on zero-points.

Data preprocessing of the 06B/07A data will start as soon as some last astrometry bugs get fixed. Distribution of preprocessed data should happen before the end of August.

Crosstalk and Data Processing (June 29 2007) Many developments have occured since April both on the hardware side and on the preprocessing pipeline side. See the presentation given at the last CFHT User's meeting in Marseille in May 2007:

A hardware solution has been found and implemented to entirely get rid of the worst crosstalk: the "negative" crosstalk (one of three different crosstalks affecting WIRCam images). The coming July 6-16 observing run will be the first with the "new" instrument.

The processing pipeline, Iiwi - version 1, is a few weeks from being completed. The final touches to correct for the non-linearity are still being implemented. Work has focussed lately on a good crosstalk correction, a better star masking technique for the sky construction, automatic handling of dark/flats processing as well as a more robust astrometry determination.

Semesters 06B and 07A data will start being processed by the end of July and distribution should start before the end of August 2007. Documentation will then be added on this web page with this first Iiwi release.

Data Processing and Distribution (April 2 2007) Our latest test to remove crosstalk on shallow stacks (1000 sec of integration) yielded promising results. It seems that source masking during sky subtraction, rather than the crosstalk, is now what dominates the unwanted background strutures. Optimizing the masking is on the menu this week in order to convingly confirm that our latest crosstalk removal technique works for relatively shallow stacks.

Data Processing and Distribution (March 19 2007) Update on data processing and data distribution. We are experiencing problems with crosstalk features at a ~0.5% level. We are actively developping methods to remove the crosstalk in our processing pipeline and Terapix is helping. We also are finalizing the standard stars analysis to obtain zero points of narrow-band filters. Distribution of the processed 06B data will not start before we complete our work on both aspects. More news at the end the week.

Data Processing A quick description of the current data processing pipeline is now found in the Preprocessing section of the WIRCam webpage. The current Preprocessing queue is also available.

Filters Following a recommandation by the Scientific Advisory Council (SAC), all WIRCam filters are offered for use in semester 2006B and on. The limitation to at most one filter exchange during the semester remains in force. The SAC will arbitrate filter conflicts between highly ranked proposals, should any occur.

QSO snapshot program are particularly encouraged with both WIRCam and MegaCam. Those programs are not charged to their respective Agency, and therefore "come for free". They should request straightforward observations for seeing > 1.2". See the QSO Phase 1 Tutorial for details.

Starting semester 2006A, microdithering will be offered only for the broad-band filters (Y, J, H, Ks). Based on our experience to date, the guiding quality through the narrow band filters is not currently sufficient for microdithering to work well. This may perhaps change after some planned upgrades to the readout software will allow increased numbers of readouts on the guider.