Here's my experience with catalogs from the 8 GPC1 images Gene sent us a note about last week.

Summary

DVO seems basically OK if you have your configuration files right. Neither magnitude nor positional uncertainties appear in DVO catalog. relastro crashes. relphot doesn't do anything.

Methods

 [wwoodvas@serenity gpc1] addstar -accept-astrom -D CATDIR test o4732g0168o.34857/o4732g0168o.34857.ch.2676.XY55.cmf
 no data in db test/Photcodes.dat
 directory test does not exist, creating...
 SUCCESS: elapsed time    1.5099 sec for   459 stars (    0 matches),    459 average,     459 measure

And you can load all of them with a foreach loop

 [wwoodvas@serenity gpc1] ls gpc1
 o4732g0168o.34857/
 o4732g0169o.34858/
 o4732g0170o.34859/
 o4732g0171o.34860/
 o4732g0172o.34861/
 o4732g0179o.34868/
 o4732g0190o.34879/
 o4732g0212o.34901/
 [wwoodvas@serenity gpc1] # Clear the directory and start again
 [wwoodvas@serenity gpc1] rm -rf gpc1
 [wwoodvas@serenity gpc1] foreach cmf ( */*.cmf )
 foreach? addstar -accept-astrom -D CATDIR gpc1 -D CAMERA gpc1 $cmf
 foreach? end
 no data in db gpc1/Photcodes.dat
 directory gpc1 does not exist, creating...
 SUCCESS: elapsed time    0.7603 sec for   320 stars (    0 matches),    320 average,     320 measure
 SUCCESS: elapsed time    0.3961 sec for   448 stars (    0 matches),    768 average,     768 measure
 SUCCESS: elapsed time    0.3962 sec for   379 stars (    0 matches),   1147 average,    1147 measure
 SUCCESS: elapsed time    0.3992 sec for   359 stars (    0 matches),   1506 average,    1506 measure
 [...]
 SUCCESS: elapsed time    0.6230 sec for   160 stars (   22 matches),  16016 average,   16452 measure
 SUCCESS: elapsed time    0.6211 sec for    69 stars (    2 matches),  16083 average,   16521 measure

Then we can look at things in DVO:

 [wwoodvas@serenity gpc1] dvo -D CATDIR gpc1
 no file /Users/wwoodvas/.dvorc

 Welcome to DVO - desktop virtual observatory

 dvo: region 360.5 0.25 1.5
 dvo: images
 current style (0): -x 2 -c black -pt 2 -lt 0 -lw 0.500000 -sz 1.000000
 reading images (mode DVO_MODE_MEF)
 plotting 477 images
 dvo: cgrid
 current style (0): -x 2 -c black -pt 2 -lt 0 -lw 0.500000 -sz 1.000000
 dvo: mextract -region ra, dec, i, i:err
 dvo: czplot ra dec i 17 12 -c red -pt 7

And here's where we see the first problem. 'i:err' is NaN for every value:

 dvo: limits i i:err
 dvo: limits
 limits: 6.623180 29.526000 nan nan [-a] [-n device]

Here we've asked the "limits" function to automatically figure out and set the limits for "i" and "i:err" (for example because we were going to plot them). It finds that the min and max values for "i:err" are NaN, i.e., all of the values of "i:err" are NaN. This holds true for all of the filters (griy). A bit of further investigation reveals that all of the ra:err and dec:err values are 0, but I don't think the inputs to these values from the .cmf value are reasonable in the first place since they tend to be in the 1000s or 0 (are the units on X_PSF_SIG thousandths of a pixel or something?).

Discussion

I don't know what's going on here? "PSF_INST_MAG" and "PSF_INST_MAG_SIG" are both defined and reasonable in the .cmf files. I can only assume that they're not getting set right by 'addstar' in the load into the DVO database.

Tips

A couple of housekeeping items before you try this at home. These are trivial issues which barely seem worthy of bug/usability reports (which I'll do) but since they'll immediately prevent people from doing stuff I'll make suggestions to have them updated to help get more people using IPP and I'll post these (and this whole note) on the Wiki short.

  • Change the "CAMERA" default in your "~/.ptolemyrc" file (which will be a copy of/link to the ippconfig/dvo.site) to "gpc1". This will save you a bit of confusion later on. 'addstar' does not appear to be smart enough to figure the CAMERA out based on the image information (unlike psphot, e.g.), and relphot/relastro and friends try to load all of their configuration information before they actually look at the CATDIR specified (which would theoretically have the necessary information to figure out what CAMERA you were using; although that could be ambiguous if several different cameras were loaded into one DVO dir).
  • Similarly, after doing this you'll think "Ah, great the 'CATDIR' specification in your '~/.ptolemyrc' is already set at 'default'", but that definition actually gets overwritten when the CAMERA config file is loaded because the default "ippconfig/gpc1/dvo.config" file define "CATDIR" to be "/data/alala.0/eugene/isp/catdir"*

Alternatively, you can always feed each command '-D CAMERA gpc1'. On a related note, I found that while the default value of CATDIR in the ~/.ptolemyrc file was "default" addstar seemed to think that it was creating a catalog somewhere else so I always specified '-D CATDIR default' on the command line.

- Michael

[*] This seems to effectively be a bug. The camera-specific configuration and recipes should not make site-specific definitions. Does this actually need to be redefined here or should it just rely on "dvo.site"? I'll file a bug report on this.