iraf.net

Hi,
yesterday I was happily displaying 16 images on DS9, one in each of the 16 available frames (using IRAF task display). All well and good.
However, when I tried imexam on frame n. 12, I realized that the output from imexam (statistics, line/column plots, contours etc.) bore no relation whatsoever with the image being examined. Careful investigations showed that imexam was actually reading the data off the image loaded in frame number 2.I wrote a very dumb script to test the behavior of imexam:[code:1:718ec9a080]mknoise test0 ncols=512 nlines=512 back=0
imarith test0 + 1.0 test1
imarith test0 + 2.0 test2
imarith test0 + 3.0 test3
imarith test0 + 4.0 test4
imarith test0 + 5.0 test5
imarith test0 + 6.0 test6
imarith test0 + 7.0 test7
imarith test0 + 8.0 test8
imarith test0 + 9.0 test9
imarith test0 + 10.0 test10
imarith test0 + 11.0 test11
imarith test0 + 12.0 test12
imarith test0 + 13.0 test13
imarith test0 + 14.0 test14
imarith test0 + 15.0 test15
imarith test0 + 16.0 test16displa test1 1 zs+ zr+
displa test2 2 zs+ zr+
displa test3 3 zs+ zr+
displa test4 4 zs+ zr+
displa test5 5 zs+ zr+
displa test6 6 zs+ zr+
displa test7 7 zs+ zr+
displa test8 8 zs+ zr+
displa test9 9 zs+ zr+
displa test10 10 zs+ zr+
displa test11 11 zs+ zr+
displa test12 12 zs+ zr+
displa test13 13 zs+ zr+
displa test14 14 zs+ zr+
displa test15 15 zs+ zr+
displa test16 16 zs+ zr+
[/code:1:718ec9a080]As far as I could see, for frames N>10 imexam actually operates on the image loaded on frame N-10 (for N=10, frame 1).I discovered this problem in a RHEL 4 box with iraf 2.12.2a and DS9 4.0b9, and found it also present in a MacIntel with Iraf 2.13 and DS9 4.0b10. Can anyone confirm or explain this peculiar imexam behavior?Thanks
Nicola

Nicola,This seems to be a DS9 problem in that it isn't returning the cursor WCS properly. Basically, for a display to frame 12 the cursor return should look something like100. 100. 1201 mwhere these are the (x,y,wcs,key) values respectively. From DS9 what I get back is100. 100. 200 mThe Frame number is encoded in the wcs as "wcs = (frame*100) + wcsnum", but apparently DS9 is truncating this to only 3 chars and so the wcs comes back as 200 ==> frame=2.You can verify this with a simple "cl> =imcur" after displaying to frame 12. XImtool handles this properly but you may want to report this to SAO as well. I couldn't find any workaround.Cheers,
-Mike

Mike, Nicola,This problem has been fixed and will be availalble with the next release of ds9, version 5.0Regards,bill joye