—
Plain Text,
12 KB (12511 bytes)
File contents
27th July 2004 Hoyle meeting room 10.30 - 12:30
Present: JRL, PSB, MJI, DWE, WJS, STH, JMI
Apologies: RGM
Agenda
------
1. Actions from last meeting
2. Comments on WFAU minutes
3. VDMT quarterly meeting
4. Q3 work plan
5. Status of WFCAM
- WFCAM ZPTs
- JRL and MJI WFCAM crosstalk data from Morag Hastie
6. VDFS VISTA stuff
- VISTA UK URD and science examples
- dicHEADs
- QFITS
- RIX responses
7. Pipeline Processing update
- COROT/IPHAS
- virtual console server
- APM2/APM3 rack mounting + RAIDs + virtual consoles
8. AOB
- APM building news
- Grant updates
Minutes
-------
1. Actions from last meeting
MJI One correction to the last minutes didn't make it to the online
version. The correct version will be moved to the online location. <<<<
MJI has made some progress on the comparison of the FIRES project reduced
data with ours. The WCS parameters in the headers have been corrected
and pipeline catalogues have been generated. MJI asked STH for the <<<<
location of the FIRES catalogues. There was also a discussion about
different seeing-weighted stacking options. Ongoing. <<<<
JRL have written a document describing requested enhancements of QFITS
PSB (+ some CPL). This should be circulated for all of us comment on. <<<<
Further QFITS discussion was deferred until item 6.
JRL has obtained a copy of the crosstalk data and Morag Hastie's report
(see item 5).
STH has emailed Simon Dye to request the generated MSBs that the survey
heads are responsible for.
JMI has investigated setting up an external (to the APM building)
backup location. 2 GB of storage will be required. Where this would
go was discussed. It was decided to wait until /data/superstore was
set up (soon). rsync will not now be used since it requires the use
of root privileges.
WJS has checked the latest URD and found no major problems.
ALL the latest URD (version 0.7) can be found at
http://www.ast.cam.ac.uk/~wfcam/internal/vdfsdocs/vdfs-ukvurd_07.pdf
Only a few comments in so far. The rest of us should read the new
version and send in comments. Ongoing. <<<<
ALL sent in monthly reports and quarterly data to STH.
PSB have commented on RIX response document.
WJS
MJI
JRL
STH
PSB has collated the RIX responses and sent them on to JPE.
MJI has asked Peter Weeks regarding the storage of office equipment
during the building work. There has been no reply so far since
Peter is on holiday. MJI to chase up when Peter returns. <<<<
2. Comments on WFAU minutes
There were no comments on the last two sets of WFAU minutes.
3. VDMT quarterly meeting
MJI reported that STH had lead a fascinating and lively discussion on
progress metrics. In a nutshell: Earned Value = Weighted Completion
ie. the summation of the current estimated percentage completion of each
task weighted by the prior estimate of the effort required for the task
at the start of the project, normalised by the sum of the weights.
E-Science funding was discussed. The Cambridge part of the VEGA bid for
the period from October 2004 up to October 2007 involves ~4 FTEs.
A JeS application (RG2 equivalent) for the Cambridge element of the VEGA
award has been prepared and sent in by Paul Aslin.
A statement of interest has been made regarding the processing of VST data
which is now being juggled between assorted PPARC committees.
There was some discussion of oversight committees and the possibility was
raised of inviting external agents to attend the quarterly meetings.
MJI said that JPE was keen on having a UK VDFS PDR, roughly scheduled for
January. This would require a complete document set to be delivered
by ~mid December taking into account the additional UK requirements.
There was general concern that this may take a significant amount of work
in an already busy period. It was noted that much of the documentation
required for a UK VDFS PDR existed but would require significant reworking.
The most efficient way of dealing with this would be to issue an extra
set of documents in addition to those submitted (by then) for the ESO FDR.
This all requires the timely completion and issue of the UK VDFS URD
ie. before the end of August. The necessary effort to updated existing
documentation and to draw up and document any new processing strategy
will be included in the September Q4 work plan.
It was noted that WFAU had requested CASU assistance in the selection of
candidates for their 2 posts. This is anticipated to be at the shortlist
stage.
The next VDMT telecon meeting is a week on Friday (6 August).
4. Q3 work plan
All were happy with the Q3 plan except PSB. He said that he would look at
the plan and comment on it by the end of the day. <<<<
5. Status of WFCAM
Steve Warren forwarded AA's WFCAM status update to all UKIDSS consortium
members: all is OK except for the optics. The optics problems (FWHM 2-3x
larger than anticipated) are caused by the corrector plate. Since the
cost of a new blank is a significant part of the overall cost of the
corrector, it has been decided to ship WFCAM immediately without the
corrector so that the corrector can be refigured. Commissioning, mainly
engineering, A/G and some detector characterisation, will go ahead as
planned in September/October. UKIRT will then revert to normal use for
the rest of the semester (November, December, January) in anticipation of
the remaining commissioning taking place shortly thereafter.
The upshot of this is that CASU personnel will not be required to go out
for on sky characterisation until the New Year. A modest series of
relevant commissioning tests can be done in the Autumn but this can be
initiated and dealt with remotely.
- WFCAM ZPTs
The magnitude discrepancy in the zeropoints may have been resolved in
favour of the more optimistic ones used in the CASU simulations and on the
UKIDSS web pages.
- WFCAM crosstalk data from Morag Hastie
JRL reported on the crosstalk data received. Analysis of the data confirms
what was in the report. The strange +/- profile of the crosstalk images
could be due to a "bias" problem, however no bias frames were available
to check this out. In order to get more information on this we apparently
have to wait until September when the WFCAM PCs have been unpacked (sic!).
Various low level patterns along (readout) rows were seen in the data which
appeared in adjacent channels and even within other quadrants, but did not
persist between frames.
JRL was asked to acquire more data, specifically series of darks and flats
from Derek Ives and to chase up the WFCAM acceptance test data from Paul
Hirst. <<<<
6. VDFS VISTA stuff
- VISTA UK URD and science examples
MJI went through four science examples that had been forwarded by Tim
Naylor to append to the current version of the URD. He also had a few more
detailed comments regarding the 0.7 version which he will email to WJS. <<<<
- dicHEADs
PSB said that the data dictionary is supposed to be sent to Michelle Peron
by 1 August. However, the VISTA-TCS dictionary has still not been issued,
although the interim DIC contains a set of new VISTA-specific TEL keywords
previously agreed. There was some discussion about how best to proceed and
in the end it was unanimoulsy decided that he should proceed by forwarding
the current (interim) DIC to JPE for him to send to Michelle. <<<<
- QFITS
PSB went through the main enhancements CASU would like to see in QFITS:
* access to MEFs via [] notation, also extraction of arbitrary
sub-image sections
* 8-bit BYTE and 16-bit INTEGER data types
* seemless RICE tile compression and decompression
* a WCS tools library
* ugraded error handling
* a more comprehensive User Guide
* CPL should handle more sophisticated weight maps (eg. confidence maps)
- RIX responses
PSB said that these have been sent to JPE for comment. It is unclear
whether these have been passed on to Michelle Peron. STH will check on
the status of this at the next VDMT. <<<<
There was a general discussion of the need for a meeting (with Steven
Beard, Malcolm Stewart, Jim Emerson) to sort out some of the wider-ranging
issues arising from the RIX responses, and agree on the more
straightforward ones. The earlier the but as August can be tricky it
may be more pragmatic to aim for early September. MJI said he would raise
this at the VDMT. <<<<
7. Pipeline Processing update
- COROT/IPHAS
JMI reported that he has used a 250 Gbyte ~1 week chunk of IPHAS data
as a dry run of the processing pipeline and progress monitoring scripts.
The r' and H_alpha data went straight through with no major problems.
The i' data processing fell over due to the usual problem of not having
an alternative master library fringe frame in the correct place. <<<<
Splitting the data by filter to the different processing computers
worked very well in this case since the persistence problem is
negligible.
One problem encountered was that access to USNOB via VIZIER (sic!) was
way too slow (as expected) and that a local copy needs to be used if
available. JRL will be incorporating a local version of the calibration
catalogues to fix this problem. This is crucial for summit processing. <<<<
[It was noted in passing that there was an online disc archive on La Palma
of all ING data since Jul/Aug 2003. So it is now possible to get replacement
files of corrupted or missing data eg. the attempted COROT data dry run
test which was postponed due to tape reading problems.]
The progress monitor worked fine during these trials, highlighting problem
datasets as well as the overall pipeline progress. There is also a monitor
which shows the status of the various computers.
- virtual console server
- APM2/APM3 rack mounting + RAIDs + virtual consoles
JMI said that he had set up a virtual console server so that there was no
need to have a console for each computer. From these virtual consoles you
can access remotely the BIOS and perform various network operations which
before had to be carried out on the computer's console, implying apm2 and
apm3 can therefore now be rack mounted together with their associated
RAID disk systems. <<<<
8. AOB
- APM building news
MJI said that various seemingly random people keep coming in to measure
things, this may represent progress but ...... as there are no tenders
for the work received yet, there are no firm dates.
- Grant updates
Paul Aslin is helping MJI sort out the various grant left-overs
(remaining CASU budget, Data Archives support, VDFS operations bid,
research etc.) in lieu of a September grant bid for such sundry things.
Continuing Actions
------------------
MJI cp correct version of the previous minutes to the online location
MJI compare FIRES project reduced data catalogues with ours
ALL read QFITS wish-list and send in comments
ALL read latest version of URD (0.7)
http://www.ast.cam.ac.uk/~wfcam/internal/vdfsdocs/vdfs-ukvurd_07.pdf
and send in comments
MJI chase up Peter Weeks regarding the storage of office equipment
during the building work
New Actions
-----------
STH provide location of FIRES catalogues to MJI
PSB read Q3 plan and provide comments to STH by end of day
JRL acquire any available darks or flats from Derek Ives
and also acquire the acceptance test data from Paul Hirst
MJI email his URD comments to WJS
PSB forward DIC to JPE and suggest he in turn forwards it to Michelle
Peron even though the DIC is still not complete
STH to check up on status of responses to RIX response document at next
VDMT meeting
MJI raise question of organizing meeting to sort out some of the issues
arising from the RIXs at next VDMT meeting
JRL update pipeline to more gracefully deal with lack of fringe (or
suitable flat) frames.
JRL build in option of using local version of calibration catalogues
in pipeline
JMI rack mount apm2 and apm3 systems
