Is the reason that you recommend using at least twice as many pixels for background as for spot definition simply statistical?
Yes.
Is it possible to create spots masks of differing ellipticity for the same image plate?
No.
When I redefine the box parameters, the shape of the spot ellipse changes also. Why?
If you change box parameters you may change parity (even of odd number) of pixels in the box. Spot ellipse has the same parity as box, so there has to be some adjustment to its shape.
Is the definition of the x,y directions in the averaged profile boxes printed by Denzo the same as in INST_HAMBURGD? I have the impression that the axes are interchanged.
Yes you are correct. The printout definition of axes are arbitrary and are different from display direction definition.
The question I have concerns the choosing of box size and background size. I have just completed processing data (taken from a CCD detector) where the box size was .75 .75, the spot elliptical was .3 .3, and the spot background was .35 .35. Data from 2 crystals was processed using these parameters and yielded Rmerge values that were good. However, I have recently read in documentation of Denzo that the background area should be twice that of the spot area. Why does this ratio of background area to spot area have to be so high ? Is it possible to get by with a lower ratio of background to spot area? Should I go back and reprocess my data with a larger background area even though it has processed well? For imaging plates does the background area still have to be twice that of the spot area?
Your background area was about 5 times the spot area. You had a large safety margin. You may get by with lower ratio at the cost of data quality.
Comment: XdisplayF: You should click Integr box in zoom window. Now, if you will type (in Denzo) command: spot elliptical 0.4 0.4 20 or command: background elliptical 0.5 0.5 20 you will see integration area on the screen. The spot region should include whole Bragg peak - in the case of kalpha1 kalpha2 split, box should include both peaks. The spot shape is defined by statement such as spot elliptical 0.6 0.5 2.0 go