-------------------------------
Objects 456.
-------------------------------

Quick verdict: 4 = UMa-like, 5 = too far to tell, 6 = disrupting globular

Coordinates of the centers of 4,5 and 6 used to create the plots:

ra_cs = [186.75, 194.295, 151.775]
dec_cs = [23.92, 34.3, 16.073]

IDL: if smth's missing it's either in Astronomy library, or can be downloaded from Eric's page.

-------------------------------

Fig. 1 
SDSS discovery plots.

PS:	       obj456_fig1.ps
IDL code:      obj456_fig1.pro
Data file:     obj456.idl	  

     Panel 1: SDSS image cut-out (still missing for obj 6)
	      box shows the location of the Subaru field
	      curcle marks the central part of the object same as in Panel 2
	      

     Panel 2: Stellar density
	      15x15 pixels for 1x1 deg field -> 4' x 4' pixels
	      Radii of the 3 circles for 4,5,6 in degrees:
		    rad1 = [0.15, 0.12, 0.12]
		    rad2 = [0.4, 0.3, 0.5]
		    rad3 = [0.43, 0.32, 0.52]
		    
     Panel 3: CMD of stars within rad1

     Panel 4: CMD of stars between rad2 and rad3 (gives the same area as rad1)

-------------------------------

Fig 2.
SDSS and Subaru Hess diagrams.

PS:		obj456_fig2.ps
IDl code:	obj456_fig2.pro
Data file:	obj456_subaru.idl

Top panels:	SDSS data, normalized Hess diagram constructed with stars selected within rad1 - normalized Hess diagram constructed with stars selected between outside rad2.

Lower panels:	Subaru data, normalized Hess diagram constructed with stars selected within rad1 - normalized Hess diagram constructed with stars selected between outside rad1.


-------------------------------

Fig 3.
SDSS and Subaru Hess diagrams. Countours overlaid with M92 ridgeline from J. Clem.

PS:		obj456_fig2_cont.ps
IDl code:	obj456_fig2_cont.pro

Contours show the densities from the previous figure. Contour levels are selected suitably to show low, high and intermediate CMD densities.

M92 ridgeline is shifted to appropriate distance and overlaid. The agreement is STRIKING.
The resulting distance moduli are:

	      dmag = [18.2, 20.9, 16.8]

In principle, from the CMD matching, you can't really shift the ridgeline by more than 0.05 mag either way, but let's be generous and quote 0.1 or 0.2 sigma error (if we incorporate M92 distance error).

Note that both obj 4 and 5 have clear BHB that are matched just as well.



-------------------------------

Fig. 4
Density contours.

PS:	       obj456_density.ps
IDL code:      obj456_density.pro

From Fig 3 it's clear that M92 is a perfect match to the objects CMDs so we are gonna use to construct the mask to selct stars for density plots. The mask is constructed by shifting the ridgeline by 0.25 mag up and down and by 0.075 left and right.

Top panels show CMD-selected SDSS stars with 18 < i < 22.5 
30x30 pixels 2' on a side, smoothed with a gaussian with FWHM=2.75 pixels.
Contour levels are 1.5, 2, 3., 5, 7, 10 sigma above the background.

Bottom panels: the central parts of the objects in Subaru data.
40x40 pixels 0.75 ' on a side, smoothed with FWHM=3 pixels
Contour levels are 1.5, 2, 3., 5, 7, 10 sigma above the background.