Post Reply 
 
Thread Rating:
  • 0 Votes - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Wrong computed magnitudes by SkyMaker ??
02-14-2012, 21:12 (This post was last modified: 02-16-2012 14:17 by JVarela.)
Post: #1
Wrong computed magnitudes by SkyMaker ??
[Update: There is no problem with SkyMaker. It was found that there was a problem with the input list of objects.... so you can continue trusting SkyMaker :-) ]

Hi Bertin,

I was happily enjoying the power of SkyMaker to produce artificial images when I came across with a weird unexpected behaviour.

This behaviour consists on producing wrong magnitudes (or using different zero points) in some areas of the image. To be more specific, I have constructed an image with stars with the same input magnitude. Then I've run SExtractor to measure the output magnitude and I've found that the output magnitudes change across the image, with similar values in some regions but quite different (>0.5mag) from region to region (I've checked with IRAF that the magnitudes are really different and consistent with the output of SExtractor)

I have attached a picture of the image with red circles around stars with magnitude ~14.0 and in blue stars with m~14.8.

   

I list the configuration file of SkyMaker that I've used:

IMAGE_NAME Test5.fits
IMAGE_SIZE 10500,10500
IMAGE_TYPE SKY
GRID_SIZE 64
IMAGE_HEADER INTERNAL
GAIN 1.0
WELL_CAPACITY 80000
SATUR_LEVEL 65000
READOUT_NOISE 4
EXPOSURE_TIME 35.0
MAG_ZEROPOINT 23.3829496198
PIXEL_SIZE 0.5
MICROSCAN_NSTEP 1
PSF_TYPE INTERNAL
PSF_NAME psf.fits
SEEING_TYPE SHORT_EXPOSURE
SEEING_FWHM 1.0
AUREOLE_RADIUS 200
AUREOLE_SB 16.0
PSF_OVERSAMP 20
PSF_MAPSIZE 256
TRACKERROR_TYPE NONE
TRACKERROR_MAJ 0.0
TRACKERROR_MIN 0.0
TRACKERROR_ANG 0.0
M1_DIAMETER 0.84
M2_DIAMETER 0.3029
#ARM_COUNT 4
ARM_COUNT 0
ARM_THICKNESS 2
ARM_POSANGLE 0.0
DEFOC_D80 0.0
SPHER_D80 0.0
COMAX_D80 0.0
COMAY_D80 0.0
AST00_D80 0.0
AST45_D80 0.0
TRI00_D80 0.0
TRI30_D80 0.0
QUA00_D80 0.0
QUA22_D80 0.0
WAVELENGTH 0.7
BACK_MAG 21.0
STARCOUNT_ZP 0
STARCOUNT_SLOPE 0.2
MAG_LIMITS 30,31
SEED_MOTION 0
SEED_STARPOS 0
VERBOSE_TYPE NORMAL
NTHREADS 0

I think with this information it could be possible to produce a similar image and test if this is a bogus behaviour connected with my installation of SkyMaker (I've tried the stable version 3.3.3. and the last development version 3.8.2 (2012-02-14) ) or whether I'm doing something wrong or, more unlikely, there is a bug in SkyMaker.

Cheers,

Jesús
Find all posts by this user
Quote this message in a reply
02-15-2012, 13:34
Post: #2
RE: Wrong computed magnitudes by SkyMaker ??
OK. I've tried modifying several of the configuration parameters (M2_DIAMETER=0;ARM_THICKNESS=0;SEEING_TYPE=LONG_EXPOSURE) with no change.

I have shrink the size of the image to 5000x5000 and there is still an area in the bottom left of the image with magnitudes >0.5m brighter than in the rest of the image.

I post a plot of the distribution of this points. The boundary is quite straight and you can see that outside the area there are some other "offending objects" distributed in a quite "regular" way (~straight lines separated ~1500 pixels along the Y axis).

I've no idea where the problem can be.

Can someone reading the post and knowing SkyMaker do a test with the files that I'm posting? Having SkyMake and SExtractor installed it should be quite straight forward doing:

-------------------------------------------------
tar xvzf Test.tar.gz
sky -c default.sky input.lis
---------------------------------------------------
(here it is needed to defined an environment variable called SEX_CONFIG with the path to the directory containing the "default.nnw" and "default.conv", eg. /usr/local/sextractor-2.8.6/config/)
---------------------------------------------------
sex -c default.sex Test5.fits
---------------------------------------------------

I've tried in two machines running linux, one with Ubuntu 10.04 and the other with openSUSE 11.3 (both 64-bits machines) with same results.

Thanks in advance,

Jesús


Attached File(s) Thumbnail(s)
   

.gz  Test.tar.gz (Size: 27.1 KB / Downloads: 691)
Find all posts by this user
Quote this message in a reply
02-16-2012, 13:24 (This post was last modified: 02-16-2012 13:25 by Emmanuel Bertin.)
Post: #3
RE: Wrong computed magnitudes by SkyMaker ??
Hi Jesús,

I noticed that in the input.lis file from your archive some stars have the same position within a fraction of a pixel: a sort -n -k 2 input.lis will return something like:
100 104.483 498.359 17
100 104.546 498.433 17
100 104.643 62.995 17
100 104.672 63.099 17


The end effect is of course to create stars that are twice (about 0.7 mag) brighter in some places.
Is this list generated by SkyMaker? If yes, this would mean that there is a problem with the random number generator.
Cheers
- Emmanuel.
Visit this user's website Find all posts by this user
Quote this message in a reply
02-16-2012, 14:15
Post: #4
RE: Wrong computed magnitudes by SkyMaker ??
Hi Emmanuel,

you got it!! Thanks!!

I have downloaded the input list from SDSS and I haven't checked for these double identifications. I have checked that the "offending" objects indeed corresponds with these double identifications and I would say that the pattern is connected with the SDSS tiles.

Thanks again for the fast answer and, by the way, for all the software that you are making available in Astromatic,

Jesús
Find all posts by this user
Quote this message in a reply
Post Reply 


Forum Jump:


User(s) browsing this thread: 1 Guest(s)