high resolution image (Keck NIRC2 Narrow Camera): mb13220_KWMB13220Ks.fits
light curve reference image (MOA-II image):       mb13220_B1447-gb13-R-4.fit

Transformation from Keck image to MOA image:
X_moa = 361.8475989821 + -0.0008463541 * X_keck + 0.0680430967 * Y_keck
Y_moa = 439.1344002918 + -0.0684771861 * X_keck + -0.0001016912 * Y_keck

Transformation from MOA image to Keck image:
X_keck =  6420.6356804087 + -0.0218245538 * X_moa + -14.6031340604 * Y_moa
Y_keck = -5238.0547760899 + 14.6962965708 * X_moa + -0.1816410920 * Y_moa

Keck image event position:
XCEN =     573.7512 +/-   0.2767
YCEN =     569.5419 +/-   0.5892

MOA image event position:
XCEN =     400.1154 +/-   0.0399
YCEN =     399.7876 +/-   0.0189

-----------------------------------------------------------------------------
FRAME1 mb13220_KWMB13220Ks.fits
FRAME2 mb13220_B1447-gb13-R-4.fit
P2 = D + A P1
X2       361.8475989821    -0.0008463541     0.0680430967
Y2       439.1344002918    -0.0684771861    -0.0001016912
P1 = D + A P2
X1      6420.6356804087    -0.0218245538   -14.6031340604
Y1     -5238.0547760899    14.6962965708    -0.1816410920
mb13220_KWMB13220Ks.fits
XCEN =     573.7512 +/-   0.2767
YCEN =     569.5419 +/-   0.5892
mb13220_B1447-gb13-R-4.fit
XCEN =     400.1154 +/-   0.0399
YCEN =     399.7876 +/-   0.0189

-----------------------------------------------------------------------------
Here is an explanation of the "info" files.

The two frames that are being registered - "FRAME2" will be the low resolution
frame
> FRAME1 mb07192_mag03auglgs_ks.fits
> FRAME2 mb07192_B6039-gb14-R-9.reg.fit

A transformation between two frames is implemented as an affine transformation
of the form Pt = D + A P where
   Pt = (yt,yt) coordinates in target frame
   P  = (x,y) coordinates in original frame
   D = (dx,dy) displacement component
   A is a matrix of coefficients, i.e.

   xt = dx + a * x + b * y
   yt = dy + c * x + d * y

   These numbers are given below as:
      dx a b
      dy c d

In all coordinate frames the origin (0,0) is at the centre of the first pixel
(following the programmers convention for array indexing).

Transformation from FRAME1 to FRAME2, i.e. calculate coordinates in FRAME2
given coordinates in FRAME1
> P2 = D + A P1
> X2       343.8346436773     0.0001627449    -0.0170845694
> Y2       348.8814187595     0.0169562852    -0.0000485012

Transformation from FRAME2 to FRAME1
> P1 = D + A P2
> X1    -20518.3379475527    -0.1674285698    58.9767886274
> Y1     19929.9953570887   -58.5339451485     0.5618035400

Position of the target in the high resolution frame
> mb07192_mag03auglgs_ks.fits
XCEN =     778.6002 +/-   4.1408
YCEN =     416.4997 +/-   5.7055

Original measured target position in the low resolution frame
> mb07192_B6039-gb14-R-9.reg.fit
XCEN =     336.8456 +/-   0.0968
YCEN =     362.0634 +/-   0.0699
