I have some Palomar LFC (6-CCD mosaic) data taken in 2x2 binned mode. I also have (thanks to civic-minded previous users) a set of MSCRED database files for unbinned data, and would like to avoid having to redo those from scratch. Which quantities in the database definitions need to be divided by 2 as a starting point? Some are obvious - like the X and Y dimensions in the surface1 and surface2 blocks - but I'm pretty sure there are more that are not so obvious... For reference, an individual chip's entry in the unbinned files looks like this (well, looked like this before being reformatted by phpBB, anyway):

begin lfc0
xrefmean 961.4971133333337
yrefmean 2121.314713333333
lngmean 3.995125574074072
latmean 49.9048992592593
pixsystem physical
coosystem j2000
projection tnx
lngref 4.
latref 50.
lngunits hours
latunits degrees
xpixref -23.70970409636232
ypixref 4094.283578521415
geometry general
function polynomial
xishift 1.972238299237671
etashift -711.2898986551909
xmag 0.1750637928015179
ymag 0.1737369533338654
xrotation 180.1670760104552
yrotation 359.5068065879138
wcsxirms 0.1829651392935669
wcsetarms 0.1856276962379241
xirms 0.1829651392935679
etarms 0.1856276962379235
surface1 11
3. 3.
2. 2.
2. 2.
0. 0.
1. 1.
2048. 2048.
1. 1.
4096. 4096.
1.972238299237671 -711.2898986551909
-0.1750630484999089 5.104899326502185E-4
-0.001495482972771719 0.1737305168483991
surface2 18
3. 3.
4. 4.
4. 4.
2. 2.
1. 1.
2048. 2048.
1. 1.
4096. 4096.
-1.461525216193809 4.592737247546287
0.001348359139103658 -8.687011904145938E-4
4.670001546449614E-7 8.629493665600824E-7
9.556348621651057E-11 9.485365477211111E-12
7.978445232091946E-4 -0.007697462633647413
-1.867505734772753E-6 -1.166988743247384E-7
-1.104574881659228E-11 -2.072858385056247E-10
6.943183391533112E-8 2.822483305635657E-6
2.316151701468996E-10 1.162260003233772E-11
-1.160920427029734E-11 -2.430066894981070E-10

Thanks,

Bill Keel

Hi Bill,

Since the database file has pixsystem=physical it should be as easy as setting up the IRAF "physical" coordinate system. These are

LTVi and LTMi_j

The first are an offset and the second are a scale. The transformation is

unbinned1 = (binned1 - LTV1) / LTM1_2
unbinned2 = (binned2 - LTV2) / LTM2_2

For 2x2 binning the LTM values are 0.5. If the first binned pixels corresponds to CCD pixel 1.5 (that is the middle of the sum of the first two unbinned pixels) then the LTV values are like 0.25. The meaning of the origin is the one that may be slightly open to interpretation.

One way to check this out is to use some image, say dev$pix, and use blkavg to bin and image; i.e. blkavg dev$pix foo 2 2.

You can then apply the unbinned database with MSCSETWCS.

Yours,
Frank