UH8K Data Calibration Service - CFHT
UH8K Data Calibration Service - CFHT
UH8K Data Calibration Service - CFHT Status Report This document proposes informations on the data calibration service provided by CFHT for the UH8K April 1998 observing run. The first part describes the data origin, the second part describes the type of data handled and the third part gives a detailed description of the UH8K camera behavior. The fourth part describes extensively the calibrations images built for the individual observing runs. The fifth part provides various informations on these data (tape contain). The last part provides information on the color equations for the UH8K. This work has provided a lot of interesting results and we already learned a lot from it. But there is certainly still a lot of work to be accomplished to make the whole process faster and more efficient for the coming CFH12K. I used an homemade software (FLIPS) still under (constant) development. Your feedback would be appreciated if you think the data you got could be improved. Please contact me at: Jean-Charles Cuillandre Tel: (808) 885-3128 Canada-France-Hawaii Telescope Corporation Fax: (808) 885-7288 P.O. Box 1597 E-mail: jcc@cfht.hawaii.edu Kamuela, Hawaii 96743 USA You can find a version of this file in HTML format with some useful images at http://www.cfht.hawaii.edu/~jcc/Detectors/UH8K/uh8kcal.html Good luck with your data analyses, Jean-Charles.
Contents:
I Data origin 1 Run description 2 Observing log II Data description 1 Exposure types 2 Filename template III UH8K camera behavior description 1 Mosaic geometry 2 Overscan - Bias level a Overscan shape b Overscan level and standard deviation c Readout noise and gain map 3 Dark current a The "dark jumping" phenomena b Dark current value 4 Domeflats and superflats: Flat-field IV Calibration frames 1 Binary mask 2 Dark frames a Example of a data set b First step: statistics on the overscan c Second step: overscan correction d Third step: statistics on the imaging area e Fourth step: combining the data 3 Domeflats 4 Superflats a Filter selection b Field selection c Data combination d Logfile e Flattening efficiency V Data description 1 Tape contain 2 Tar contain for the logfiles 3 Tar contain for the FITS files 4 Tar size VI Color equations
I Data origin:
1) Run description:
The data come from eight individual observing runs from this April UH8K run (18 nights overall). You can find the description of your run in the CFHT observing schedule, I reproduce here the description of these runs so you know who you are. The runs are indicated sequentially as they happened in time. N is the number of nights per run.-------------------------------------------------------------------------------- PI name ---- RunID N Description -------------------------------------------------------------------------------- Tonry ------ H39 - 2 Search for High Redshift Supernovae Wainscoat -- H17 - 2 The Luminosity Function of Dwarf Galaxies in Virgo Kormendy --- H20 - 2 The Luminosity Function of Dwarf Galaxies in Virgo Creze ------ F01 - 2 Matiere noire et halo stellaire Luppino ---- H37 - 2 Probing dark matter with weak gravitational lensing Jewitt ----- H10 - 3 The Kuiper Belt de Lapparent F03 - 1 Fonctions de luminosite des galaxies a z~`0.1-0.5 Mazure ----- F52 - 2 Fonction de luminosite profonde d'amas de galaxies proches Cuillandre - D04 - 1 Distribution de masse des amas a grande distance du centre --------------------------------------------------------------------------------All generated files are based on your observing run identifier (RunID) From these 18 nights of observation more than 150 dark, 100 domeflat and 110 object exposures have been handled by the pipeline to produce individual calibrations for each run. So from all these files (about 50 Gbytes of data processed), 13 combined darks, 9 combined domeflats and 7 superflats have been built.
2) Observing log:
To be able to select the data, I had to generate ASCII files from our archive. This log can be an interesting complement to the written observing log. A print of this file will be furnished with the tape. These files are run specific so you will be the only one, as a PI, to get a copy of your run observing log. The format is: ExpNumber Date UT Dec Alpha Epoch ExpType Comment ExpTime Filter Airmass TelFoc Example: 443331o 04-05-98 06:42:58.00 12:09:51.80 18:36:03.00 2000 OBJECT NGC4147 10.141 V 1.032 6133II Data description:
1) Exposure types:
Three different types of exposures have been handled: - Dark exposures (XXXXXXdX.fits): exposure time varies from 0 to 1200 seconds to match the exposure time of scientific frames. I consider the bias as a zero second dark exposure so there is no further mention of the bias. The signal as described below is composed of the CCD offset (measured in the overscan region) the dark signal and the light signal. - Flat exposures (XXXXXXfX.fits): I combined only dome flats (no twilight flats). - Object exposures (XXXXXXoX.fits): used to create the superflats. Combined data filenames are based on the RunID, all frames are composed of 8 individual FITS images as the raw data (CCD number ranges from 0 to 7).2) Filename template:
Combined filenames follow a strict definition: Type - Description - Camera - RunID - CCD ID - .fits Example: Dark20mn-UH8K-F01.0.fits Mask-UH8K-98I.0.fits FlatR-UH8K-H17.0.fits SuperflatR-UH8K-H20+H17+F01.0.fitsIII UH8K camera behavior description:
1) Mosaic geometry
The UH8K mosaic is composed of eight 2K4K CCDs organized in two banks. One bank per CCD controller. To avoid pickup noise during synchronous readouts, banks are read sequentially. First the CCDs {0,1,2,3} and then {4,5,6,7}. The mosaic geometry is: . . . . 2 3 5 4 .0 .1 .7 .6 The output being indicated by a dot sign. Note that when visualized individually, all CCDs have the output on the lower left.2) Overscan - Bias level
The signal offset from the detector plus the video chain can be measured and monitored in the overscan region (from X:2050 to X:2080 and Y:0 to Y:4110). When combining data, the overscan level was measured as well as its standard deviation. An histogram is built to define the mode of the image, the standard deviation is measured as half of the width of the distribution at the 68% level from the peak value. It appears that this offset level is not stable in time neither geometrically, hence requiring a per image bases correction prior any further image processing. a) Overscan shape: The overscan shape evolves along the Y axis, it first starts high then goes down by about 20 ADUs before going back to a higher and stable value. This "unstable" area takes place in the first 500 lines. See figure 1 for a cut of the overscan region before and after correction.b) Overscan level and standard deviation: This level appears to be changing by a a few ADUs (from 3 to 20) from one exposure to another. Check out the ASCII logfiles issued from the reduction pipeline to get an idea of it. These files end with the extension "-ov.dat" and are generated by a statistic program run prior any reduction work (immode-Dark20mn-UH8K-F01-ov.dat). The average value of the overscan and its standard deviation is (in ADUs): 2820/11.9 5700/11.7 5520/15.9 3010/19.0 1360/2.2 2340/2.8 2230/7.2 1260/3.7 c) Readout noise and gain map: From a 1996 UH8K run data, I measured the gain of each CCD. These are not highly accurate values, say at +/-10%. Gain: 2.1 1.4 1.3 1.6 2.3 1.6 2.0 2.5 So the readout noise in electrons is: 25 16 21 30 6 6 14 9
Figure 1 Overscan shape. 3) Dark current
a) The "dark jumping" phenomena: The dark appeared not being stable in time. They are actually two regimes of dark current in the mosaic, one I qualify as normal and one as "low". These two regimes on themselves are very stable, the exposures have either one or the other mode. The problem is that it affects the two CCD banks randomly and independently, thus making the phenomena very difficult to track down since the difference in dark current is not very high. The problem is that even in the V band, you can't detect an offset of about 20 ADUs has you are dominated by sky brightness fluctuations. I found a trick to track down this phenomena: reduce the data with the normal dark map (which holds for at least 85% of the data), along with a proper overscan correction and flat-field correction. Then you strongly bin the whole mosaic in one image (binning by 8 is enough): as you display it, if the right bank is affected then these four right CCDs will appear darker than the left part and vice versa. If both banks have been affected you will see some dark wave shapes on the data showing that the dark has been over-corrected, and if data are fine then it looks just fine and flat. Please refer to the figure 2 for an example of four images which have been corrected with the same dark (the normal one) and illustrating the four cases (these four files happen to be a set taken in a row (Run H37, files 669 to 672). After you have detected in which category the given image fits in, you have to correct it with the proper bias: -> Normal data: then you can use the dark generated for your specific run. Example: Dark20mn-UH8K-H17.0.fits (0 to 7) -> "Low" data: depending on which of the three cases (low right, low left or low full) you correct your data with a composite dark made of part (or complete) of Dark20mn_LOW_LEFT-UH8K-98I.0.fits (0 to 3) and Dark20mn_LOW_RIGHT-UH8K-98I.0.fits (4 to 7). This file (in LOWDARK directory) has been generated from several 20 minutes exposures from the whole observing run as each individual run did not provide enough data to create a high signal-to-noise ratio frame. The dark current appears to be very stable in time anyway.Go through the process of processing and binning your data after proper correction of the dark level, data should be fine then. This phenomena does not affect short exposures (less than a minute) and it is hardly detectable on R an I images due to the high sky background. b) Dark current value: Here is the dark current value for a 20 minutes exposure in both modes (judge by yourself if you want to neglect this effect or not). Normal 20mn dark: 41 37 46 71 26 38 42 36 Low 20mn dark: 14 8 10 24 8 12 15 10 This strange phenomena is recent (less than one year) and not understood. !!!!!!!!! For the run H39 (first two nights of the run), the heater system was failing so the CCD temperature was very low, resulting in a ZERO level dark current even on 20 minutes exposures (data are provided anyway) but with a severe CTE problem on the top four CCDs when the sky background is low.
Figure 2 Illustration of "dark jumping" phenomena. 4) Domeflats and superflats: Flat-field
The optical transmission of the telescope+camera obviously evolves slightly in time. Structures appear at the level of a few percent. This is valid also for superflats. Analyzing high signal to noise ratio data, the pixel to pixel quantum efficiency non-uniformity is about e_rqe=0.9%. The noise equation is then: N=sqrt(readout_noise^2 + S/g + (e_rqe * S)^2), S is the signal level in ADU. The readout_noise is to be expressed in ADU and the gain in electron per ADU, these values are given in this document for each CCD of the mosaic (gain and readout noise).IV Calibration frames:
A pipeline based on the FLIPS software (a homemade software to deal with large amount of big mosaic FITS files, high speed and low memory use) has been used to reduce and generate all the files quoted in this document.1) Binary mask:
A binary mask (1/0) image is available. It sets to 0 the following areas of detectors: - overscan region - bad columns - clusters - big dust (on the CCD) This mask is based on data obtained last fall so some dust might have moved on the CCD in the meantime... The areas were defined from a superflat in the V band (low flux level).2) Dark frames:
To combine several dark images, since the overscan is not stable in time, it has to be corrected prior combination. The program running the whole process first makes a first pass on the data to determine statistics in the overscan area (level and standard deviation). Then the images are corrected from the overscan response by modeling it with a smooth 1D vector (figure 1). This reduction does not affect at all the original signal-to-noise ratio of the raw data. Then a second statistic step gives some statistics on the imaging area (level and standard deviation). Then the combine algorithm uses the overscan corrected data to create the final frame. The combine algorithm is a standard sigma clipping procedure rejecting bad values and then taking the median of the remaining sample. The clipping values adopted for the dark exposures are: -2.5 sigma / +1.5 sigma, then taking the median value. Log files are generated at each step and the final combined image contains all the information about the combining process in its FITS header. Here is an example: a) Example of a data set: a set of thirteen 10 seconds dark exposures. b) First step: statistics on the overscan The ASCII logfile immode-Dark10sc-UH8K-F03-ov.dat is generated. It contains a header telling what has been analyzed and how the results are provided: -> immode -> Xc=2070 Yc=2000 Xs=20 Ys=4000 -> statistics on 60 percent of the area -> Format: name mode mean sigma Then the data table contains all images listed per CCD number so one gets to see all CCD number 0 listed together, etc... through 7: /h/uh8k/images/Darks/10sc/443091d0.fits 1366 1366.1 3.0 /h/uh8k/images/Darks/10sc/443093d0.fits 1363 1362.5 3.0 /h/uh8k/images/Darks/10sc/443094d0.fits 1363 1362.6 3.0 ... Here is a example of a complete file: immode-Dark20mn-UH8K-F01-ov.dat c) Second step: overscan correction The 1D smooth vector is generated from a modeling of the overscan by first taking the mean value of a 10 pixels large vector along the X axis. Then this vector is smoothed with a 40 pixels large median filter. This vector is then subtracted to the file. The output file has a "O" letter extension: 443091d6.fits -> 443091d6O.fits The FITS header of the file 443091d6O.fits then contains: HISTORY = imred: Fri May 1 13:08:00 1998 HISTORY = imred: input image: 443091d6.fits HISTORY = imred: output image: 443091d6O.fits HISTORY = imred: overscan correction: U1 2069 2079 4110 4110 COMMENT = imred: CFHT data calibration service/Waimea 0498/jcc@cfht.hawaii.edu d) Third step: statistics on the imaging area -> immode -> Xc=1000 Yc=2000 Xs=2000 Ys=4000 -> statistics on 40 percent of the area -> Format: name mode mean sigma /h/uh8k/images/Darks/10sc/443091d0O.fits 10 1.6 5.0 /h/uh8k/images/Darks/10sc/443093d0O.fits 10 1.4 5.0 /h/uh8k/images/Darks/10sc/443094d0O.fits 10 1.5 5.0 ... ! When the mode is lower than 10 ADUs, the algorithm has some difficulties to find the right values (to fix...) so I canceled the mode calculation. The standard deviation appears also to be overestimated so ONLY the mean value should be looked at for these low level images. The noise can be checked anyway from the overscan in the first logfile. Here is a example of a complete file: immode-Dark20mn-UH8K-F01.dat. You can see here the "dark jumping phenomena happening randomly on any of the two CCD banks. Also notice what I just mention before about the mode value if less then 10 ADUs. e) Fourth step: combining the data The FITS header contains: HISTORY = imcombine: Fri May 1 14:11:52 1998 HISTORY = imcombine: output image: Dark10sc-UH8K-F03.6.fits HISTORY = imcombine: mediane of 13 images HISTORY = imcombine: standard sigma clipping [-2.5;+1.5] HISTORY = imcombine: thresholding from 0.00 to 65535.00 ADUs HISTORY = imcombine: no rescaling/offsetting HISTORY = imcombine: 443091d6O.fits 10 0.000 [0.0,0.0]->[0.0,0.0] .... HISTORY = imcombine: 443104d6O.fits 10 0.000 [0.0,0.0]->[0.0,0.0] COMMENT = imcombine: CFHT calibration service/Waimea 0498/jcc@cfht.hawaii.edu The list of images used is provided. The 0.0 values are of no importance here (when combining scientific frames, they contain the flux ratio and the offsets in X and Y). The value after the filename is the mode, hence truncated if lower than 10.0 ADUs, again the ASCII logfiles provides the information if needed. There is no rescaling/offsetting since the dark is an additive component to the signal. I checked the logfiles after each combine to be sure that the set of data used was good.3) Domeflats:
It is the same process as for the dark exposures except that now a dark current correction is required (with a few seconds dark exposure). Also, the combine algorithm parameters are: -2.5/+2.0 sigma taking the mean value after rescaling by the mode of each individual image. Since the flux is high, cosmic rays are lost within the noise so the mean value is adequate. The rescaling is essential to take into account the fact that the dome lamps are not stable in time. The relative gain of each CCD in the mosaic is preserved: there is no normalization. The relative gain is preserved by rescaling each CCD to the mean of the mode of input files, then of course you have to combine on a per mosaic basis, not a per CCD basis otherwise the ratio between CCDs might be truncated if less images are used for one CCD than another. ASCII logfiles are generated the same way as for the dark exposures. Example: HISTORY = imred: Thu Apr 23 14:23:41 1998 HISTORY = imred: input image: 442019f3.fits HISTORY = imred: output image: 442019f3OD.fits HISTORY = imred: overscan correction: U1 2069 2079 4110 4110 HISTORY = imred: dark/bias correction: - (Darks10sc-UH8K-H17.3.fits * 1.00) COMMENT = imred: CFHT data calibration service/Waimea 0498/jcc@cfht.hawaii.edu HISTORY = imcombine: Thu Apr 23 14:41:10 1998 HISTORY = imcombine: output image: FlatV-UH8K-H17.3.fits HISTORY = imcombine: mean of 6 images HISTORY = imcombine: standard sigma clipping [-2.5;+2.0] HISTORY = imcombine: thresholding from 0.00 to 65535.00 ADUs HISTORY = imcombine: mean modes rescaling HISTORY = imcombine: output mode = 12431 = mean(modes in) HISTORY = imcombine: 442019f3OD.fits 12269 0.000 [0.0,0.0]->[0.0,0.0] ... HISTORY = imcombine: 442024f3OD.fits 12566 0.000 [0.0,0.0]->[0.0,0.0] COMMENT = imcombine: CFHT calibration service/Waimea 0498/jcc@cfht.hawaii.edu Here is a example of a complete file FITS header: FlatV-UH8K-F01.1.header4) Superflats:
a) Filter selection The first step is to select the fields that can be used to build a superflat per filter (V, R and I for this run). As the optical transmission evolves in time, superflats were created on a per observing run bases so you get the best flattening of your data. Also filters removal was creating a logical barrier between different superflats: the UH8K can contain only two filters at a time, so there several change of the R and I filters whereas the V filter always remained in place. Here is when filters where mounted in the camera:RunID: H39 H20 H17 F01 H37 H10 F01 F52 D04 V: xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx R: xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxx I: xxxxxxxx xxxxxxxxxxx xxxxxxxxxxxxxxSo the following superflats were built: H39 -> I H20+H17+F01 -> R F01+H37 -> I F01+H37 -> V H10 -> R F01+H37+F03+F52 -> I F01+H37+F03+F52 -> V The runs F03+F52+D04 did not contain enough fields to build a good superflat and after some testing, it appears that the correction for the I band data with the superflat based on the runs F01+H37+F03+F52+D04 was providing good results even with the removal of the filter between the observing runs. b) Field selection A proper selection of the fields is required to get the best gain correction both at small scales (pixel to pixel) and large scales. The set of data must not contain: - data with extended faint astronomical objects - data with no major scattered lights from a bright star - data with abnormal sky background level (avoid moon or twilights) - more than one identical field with an offset lower that the largest objects Fortunately only a few data from this whole run suffer from these artifacts so most fields could be used but the last condition was highly constraining. Here is the number of frames used per superflat:
-------------------------------------------------------------------------------- RunID Filename Filter #images -------------------------------------------------------------------------------- H39 SuperflatI_mean-UH8K-H39.*.fits I 27 H20+H17+F01 SuperflatR-UH8K-H20+H17+F01.*.fits R 15 F01+H37 SuperflatI-UH8K-F01+H37.*.fits I 17 F01+H37 SuperflatV-UH8K-F01+H37.*.fits V 18 H10 SuperflatR-UH8K-H10.*.fits R 22 F01+H37+F03+F52 SuperflatI-UH8K-F01+H37+F03+F52.*.fits I 23 F01+H37+F03+F52 SuperflatV-UH8K-F01+H37+F03+F52.*.fits V 24 --------------------------------------------------------------------------------c) Data combination: The same process as the dome-flats is applied: overscan correction then a dark current correction matching the scientific frames exposure time and then the combination with factor rescaling. The sigma clipping is -2.5/+2.0 sigma and TWO superflats are generated: one with the median value of the remaining sample and one with the mean. So for each superflat the names are: SuperflatR-UH8K-H20+H17+F01.*.fits and SuperflatR_mean-UH8K-H20+H17+F01.*.fits (except H39 which has only the mean). The mean superflat is slightly more affected by remaining objects than the median superflat (first name) but the mean superflat has a higher signal to noise ratio (typically a factor of 1.22 in favor of the mean). The fluctuations of the ratio between these two types of superflats appears to be less than 0.05% so I would advocate the use of the mean superflat. This shows that the sigma-clipping is efficiently rejecting objects from the sample, hence the mean of the remaining value is a good estimator of the sky background. But I leave you the choice, further investigations based on the number of frames in the combined image are required for the future (CFH12K!). The figure 3 shows the ratio between the median and mean V superflats (run F01+H37). Fluctuations are about 0.05%. This small portion (512x512 pixels) is representative of the whole mosaic.
d) Logfile Only the imaging area statistics are generated for the superflats. e) Flattening efficiency: I made some tests with these superflats to check how good they correct large scales structures. Since the baffling is still very poor at the time of this run it is not surprising to notice that it is hard to flatten the 28'x28' field at better than 1%. Even when dividing two consecutive images, some structures are present, proving that improvement have to be made. This is happening right now with the upgrade of the wide-field corrector (internal baffling and lens coating). The large number of frames leads to a very good correction of the pixel to pixel gain, gain in signal to noise ratio is higher than with a high flux level domeflat. You are welcome to test this systematically since I checked this only for data of one run, still, I believe the superflats will give better results.
Figure 3 Ratio between the median and mean V superflats (run F01+H37). Fluctuations are about 0.05%. V Data description:
1) Tape contain:
You will receive a tape which contain is organized in individual tar files. To get to one of the tar, use the command "mt fsf n" where n is the index indicated before the tar name. This is the case of a "rewind device". Data are provided on an EXABYTE tape. -------------------------------------------------------------------------------- tar number tar name -------------------------------------------------------------------------------- 0 Statistics/Superflats/H39 1 Statistics/Superflats/F01+H37+F03+F52 2 Statistics/Superflats/H10 3 Statistics/Superflats/H20+H17+F01 4 Statistics/Superflats/F01+H37 5 Statistics/Darks+Flats/F01 6 Statistics/Darks+Flats/F03 7 Statistics/Darks+Flats/F52 8 Statistics/Darks+Flats/H10 9 Statistics/Darks+Flats/H17 10 Statistics/Darks+Flats/H20 11 Statistics/Darks+Flats/H37 12 Statistics/Darks+Flats/H39 13 Statistics/Darks+Flats/LOWDARK 14 Mask 15 Darks/LOWDARK 16 Darks/H20+H17/20mn 17 Darks/H20+H17/00sc 18 Darks/H20+H17/10sc 19 Darks/F01/20mn 20 Darks/H10/00sc 21 Darks/H10/10mn 22 Darks/H37/20mn 23 Darks/H37/25sc 24 Darks/F03+F52/10sc 25 Darks/F03+F52/20mn 26 Darks/H39/20mn 27 Darks/H39/25sc 28 Flats/H20/R 29 Flats/H17/V 30 Flats/H17/R 31 Flats/F01/V 32 Flats/F01/R 33 Flats/F01/I 34 Flats/H10/R 35 Flats/F03+F52/I 36 Flats/F03+F52/V 37 Superflats/H39/I 38 Superflats/H10/R 39 Superflats/H20+H17+F01/R 40 Superflats/F01+H37/I 41 Superflats/F01+H37/V 42 Superflats/F01+H37+F03+F52/I 43 Superflats/F01+H37+F03+F52/V --------------------------------------------------------------------------------2) tar contain for the logfiles:
Here is the name of the files in individual tars: -------------------------------------------------------------------------------- Statistics/Superflats/H39/immode-SuperflatI-UH8K-H39.dat -------------------------------------------------------------------------------- Statistics/Superflats/F01+H37+F03+F52/immode-SuperflatI-UH8K-F01+H37+F03+F52.dat Statistics/Superflats/F01+H37+F03+F52/immode-SuperflatV-UH8K-F01+H37+F03+F52.dat -------------------------------------------------------------------------------- Statistics/Superflats/H10/immode-SuperflatR-UH8K-H10.dat -------------------------------------------------------------------------------- Statistics/Superflats/H20+H17+F01/immode-SuperflatR-UH8K-H20+H17+F01.dat -------------------------------------------------------------------------------- Statistics/Superflats/F01+H37/immode-SuperflatI-UH8K-F01+H37.dat Statistics/Superflats/F01+H37/immode-SuperflatV-UH8K-F01+H37.dat -------------------------------------------------------------------------------- Statistics/Darks+Flats/F01/immode-FlatR-UH8K-F01-ov.dat Statistics/Darks+Flats/F01/immode-FlatR-UH8K-F01.dat Statistics/Darks+Flats/F01/immode-FlatI-UH8K-F01-ov.dat Statistics/Darks+Flats/F01/immode-FlatI-UH8K-F01.dat Statistics/Darks+Flats/F01/immode-FlatV-UH8K-F01-ov.dat Statistics/Darks+Flats/F01/immode-FlatV-UH8K-F01.dat Statistics/Darks+Flats/F01/immode-Dark20mn-UH8K-F01-ov.dat Statistics/Darks+Flats/F01/immode-Dark20mn-UH8K-F01.dat -------------------------------------------------------------------------------- Statistics/Darks+Flats/F03/immode-Dark10sc-UH8K-F03-ov.dat Statistics/Darks+Flats/F03/immode-Dark10sc-UH8K-F03.dat -------------------------------------------------------------------------------- Statistics/Darks+Flats/F52/immode-Dark20mn-UH8K-F52-ov.dat Statistics/Darks+Flats/F52/immode-Dark20mn-UH8K-F52.dat Statistics/Darks+Flats/F52/immode-FlatI-UH8K-F52-ov.dat Statistics/Darks+Flats/F52/immode-FlatI-UH8K-F52.dat Statistics/Darks+Flats/F52/immode-FlatV-UH8K-F52-ov.dat Statistics/Darks+Flats/F52/immode-FlatV-UH8K-F52.dat -------------------------------------------------------------------------------- Statistics/Darks+Flats/H10/immode-Dark00sc-UH8K-H10-ov.dat Statistics/Darks+Flats/H10/immode-Dark00sc-UH8K-H10.dat Statistics/Darks+Flats/H10/immode-Dark10mn-UH8K-H10-ov.dat Statistics/Darks+Flats/H10/immode-Dark10mn-UH8K-H10.dat Statistics/Darks+Flats/H10/immode-FlatR-UH8K-H10-ov.dat Statistics/Darks+Flats/H10/immode-FlatR-UH8K-H10.dat Statistics/Darks+Flats/H10/immode-SuperflatR-UH8K-H10-ov.dat -------------------------------------------------------------------------------- Statistics/Darks+Flats/H17/immode-Dark10sc-UH8K-H17.dat Statistics/Darks+Flats/H17/immode-FlatV-UH8K-H17-ov.dat Statistics/Darks+Flats/H17/immode-FlatV-UH8K-H17.dat Statistics/Darks+Flats/H17/immode-FlatR-UH8K-H17-ov.dat Statistics/Darks+Flats/H17/immode-FlatR-UH8K-H17.dat Statistics/Darks+Flats/H17/immode-Dark10sc-UH8K-H17-ov.dat Statistics/Darks+Flats/H17/immode-Dark20mn-UH8K-H20H17-ov.dat Statistics/Darks+Flats/H17/immode-Dark20mn-UH8K-H20H17.dat -------------------------------------------------------------------------------- Statistics/Darks+Flats/H20/immode-Dark00sc-UH8K-H20-ov.dat Statistics/Darks+Flats/H20/immode-FlatR-UH8K-H20-ov.dat Statistics/Darks+Flats/H20/immode-Dark00sc-UH8K-H20.dat Statistics/Darks+Flats/H20/immode-FlatR-UH8K-H20.dat -------------------------------------------------------------------------------- Statistics/Darks+Flats/H37/immode-Dark20mn-UH8K-H37-ov.dat Statistics/Darks+Flats/H37/immode-Dark20mn-UH8K-H37.dat Statistics/Darks+Flats/H37/immode-Dark25sc-UH8K-H37-ov.dat Statistics/Darks+Flats/H37/immode-Dark25sc-UH8K-H37.dat -------------------------------------------------------------------------------- Statistics/Darks+Flats/H39/immode-Dark25sc-UH8K-H39.dat Statistics/Darks+Flats/H39/immode-Dark20mn-UH8K-H39.dat Statistics/Darks+Flats/H39/immode-Dark25sc-UH8K-H39-ov.dat Statistics/Darks+Flats/H39/immode-SuperflatI-UH8K-H39-ov.dat Statistics/Darks+Flats/H39/immode-Dark20mn-UH8K-H39-ov.dat -------------------------------------------------------------------------------- Statistics/Darks+Flats/LOWDARK/immode-Dark20mn_LOW_LEFT-UH8K-98I-ov.dat Statistics/Darks+Flats/LOWDARK/immode-Dark20mn_LOW_LEFT-UH8K-98I.dat Statistics/Darks+Flats/LOWDARK/immode-Dark20mn_LOW_RIGHT-UH8K-98I-ov.dat Statistics/Darks+Flats/LOWDARK/immode-Dark20mn_LOW_RIGHT-UH8K-98I.dat --------------------------------------------------------------------------------3) tar contain for the FITS files:
Here is a list of the files in individual tars: -------------------------------------------------------------------------------- Mask/Mask-UH8K-98I.*.fits -------------------------------------------------------------------------------- Darks/LOWDARK/Dark20mn_LOW_LEFT-UH8K-98I.*.fits Darks/LOWDARK/Dark20mn_LOW_RIGHT-UH8K-98I.*.fits -------------------------------------------------------------------------------- Darks/H20+H17/20mn/Dark20mn-UH8K-H20H17.*.fits -------------------------------------------------------------------------------- Darks/H20+H17/00sc/Dark00sc-UH8K-H20.*.fits -------------------------------------------------------------------------------- Darks/H20+H17/10sc/Dark10sc-UH8K-H17.*.fits -------------------------------------------------------------------------------- Darks/F01/20mn/Dark20mn-UH8K-F01.*.fits -------------------------------------------------------------------------------- Darks/H10/00sc/Dark00sc-UH8K-H1*.*.fits -------------------------------------------------------------------------------- Darks/H10/10mn/Dark10mn-UH8K-H1*.*.fits -------------------------------------------------------------------------------- Darks/H37/20mn/Dark20mn-UH8K-H37.*.fits -------------------------------------------------------------------------------- Darks/H37/25sc/Dark25sc-UH8K-H37.*.fits -------------------------------------------------------------------------------- Darks/F03+F52/10sc/Dark10sc-UH8K-F03.*.fits -------------------------------------------------------------------------------- Darks/F03+F52/20mn/Dark20mn-UH8K-F52.*.fits -------------------------------------------------------------------------------- Darks/H39/20mn/Dark20mn-UH8K-H39.*.fits -------------------------------------------------------------------------------- Darks/H39/25sc/Dark25sc-UH8K-H39.*.fits -------------------------------------------------------------------------------- Flats/H20/R/FlatR-UH8K-H20.*.fits -------------------------------------------------------------------------------- Flats/H17/V/FlatV-UH8K-H17.*.fits -------------------------------------------------------------------------------- Flats/H17/R/FlatR-UH8K-H17.*.fits -------------------------------------------------------------------------------- Flats/F01/V/FlatV-UH8K-F01.*.fits -------------------------------------------------------------------------------- Flats/F01/R/FlatR-UH8K-F01.*.fits -------------------------------------------------------------------------------- Flats/F01/I/FlatI-UH8K-F01.*.fits -------------------------------------------------------------------------------- Flats/H10/R/FlatR-UH8K-H1*.*.fits -------------------------------------------------------------------------------- Flats/F03+F52/I/FlatI-UH8K-F52.*.fits -------------------------------------------------------------------------------- Flats/F03+F52/V/FlatV-UH8K-F52.*.fits -------------------------------------------------------------------------------- Superflats/H39/I/SuperflatI_mean-UH8K-H39.*.fits -------------------------------------------------------------------------------- Superflats/H10/R/SuperflatR-UH8K-H10.*.fits Superflats/H10/R/SuperflatR_mean-UH8K-H10.*.fits -------------------------------------------------------------------------------- Superflats/H20+H17+F01/R/SuperflatR-UH8K-H20+H17+F01.*.fits Superflats/H20+H17+F01/R/SuperflatR_mean-UH8K-H20+H17+F01.*.fits -------------------------------------------------------------------------------- Superflats/F01+H37/I/SuperflatI-UH8K-F01+H37.*.fits Superflats/F01+H37/I/SuperflatI_mean-UH8K-F01+H37.*.fits -------------------------------------------------------------------------------- Superflats/F01+H37/V/SuperflatV-UH8K-F01+H37.*.fits Superflats/F01+H37/V/SuperflatV_mean-UH8K-F01+H37.*.fits -------------------------------------------------------------------------------- Superflats/F01+H37+F03+F52/I/SuperflatI-UH8K-F01+H37+F03+F52.*.fits Superflats/F01+H37+F03+F52/I/SuperflatI_mean-UH8K-F01+H37+F03+F52.*.fits -------------------------------------------------------------------------------- Superflats/F01+H37+F03+F52/V/SuperflatV-UH8K-F01+H37+F03+F52.*.fits Superflats/F01+H37+F03+F52/V/SuperflatV_mean-UH8K-F01+H37+F03+F52.*.fits --------------------------------------------------------------------------------4) tar size:
Here is the size of each tar: (total size is 4,844,248,000 bytes ~ 5 Gbytes) 16 ./Statistics/Superflats/H39 31 ./Statistics/Superflats/F01+H37+F03+F52 12 ./Statistics/Superflats/H10 10 ./Statistics/Superflats/H20+H17+F01 23 ./Statistics/Superflats/F01+H37 55 ./Statistics/Darks+Flats/F01 13 ./Statistics/Darks+Flats/F03 59 ./Statistics/Darks+Flats/F52 46 ./Statistics/Darks+Flats/H10 70 ./Statistics/Darks+Flats/H17 19 ./Statistics/Darks+Flats/H20 17 ./Statistics/Darks+Flats/H37 44 ./Statistics/Darks+Flats/H39 25 ./Statistics/Darks+Flats/LOWDARK 134465 ./Mask 134465 ./Darks/LOWDARK 134465 ./Darks/H20+H17/20mn 134465 ./Darks/H20+H17/00sc 134465 ./Darks/H20+H17/10sc 134465 ./Darks/F01/20mn 134465 ./Darks/H10/00sc 134465 ./Darks/H10/10mn 134465 ./Darks/H37/20mn 134465 ./Darks/H37/25sc 134465 ./Darks/F03+F52/10sc 134465 ./Darks/F03+F52/20mn 134465 ./Darks/H39/20mn 134465 ./Darks/H39/25sc 135369 ./Flats/H20/R 134465 ./Flats/H17/V 134465 ./Flats/H17/R 134465 ./Flats/F01/V 134465 ./Flats/F01/R 134465 ./Flats/F01/I 134465 ./Flats/H10/R 134465 ./Flats/F03+F52/I 134465 ./Flats/F03+F52/V 134474 ./Superflats/H39/I 268930 ./Superflats/H10/R 268930 ./Superflats/H20+H17+F01/R 268930 ./Superflats/F01+H37/I 268930 ./Superflats/F01+H37/V 268930 ./Superflats/F01+H37+F03+F52/I 271050 ./Superflats/F01+H37+F03+F52/VVI Color equations:
During the D04 night, I observed several standard fields, including one that I put sequentially on each CCD of the mosaic (excepted chip 4 which is bad). I derived color equations but I don't have the airmass term since the fields were observed at about the same airmass. Please refer to standard values for the CFHT: Appendix F, photometric coefficients at http://www.cfht.hawaii.edu/manuals/focam/appen.html#F, a table to download. This table shows that the airmass term whatever CCD used at the prime is : V: -0.12 R: -0.10 I: -0.05 So this should hold also for the UH8K. The color equation happens to be similar for all the CCDs so I averaged all the data to obtain the general solution. The color equations derived from UH8K observations on SA104/SA110 and NGC4147 are: (I only made this for the V and I filter) Vc = Vinst - 0.031*(Vinst - Iinst) Ic = Iinst - 0.066*(Vinst - Iinst) A former analyses by Jerome Bouvier based on UH8K data gives a color equation for the R band with a significant color term: Rc = Rinst - 0.37*(Rinst - Iinst) + 0.39
