GSC-2.2 Guide update
Click here for Italian version of this page
I've just made some small revisions to Guide to enable display of GSC-2.2, the recently-released new version of GSC. (Click here for details about GSC-2.2.) The catalog currently has about a half billion objects, with pretty good positions, magnitudes in up to four different bands (though you rarely get more than two for any given object), and even ellipses indicating the shapes of objects such as galaxies.
To get this to work, you should first download the current 32-bit Windows version of Guide (I've not updated the 16-bit Windows or the 32-bit DOS versions yet, and I'm not sure they'll be able to support this feature anyway). Start it up, and go to "Settings... Toolbar", and look about three-quarters of the way down the list for a "GSC 2.2 data via Internet" button. (Click here if the button doesn't show up at all.) Turn it on, and close the Toolbar dialog. A button labelled "22" should appear in the toolbar.
Now zoom in on an area of interest... zoom to a very small field of view, say, a few arcminutes across. GSC-2.2 is quite detailed, and it sends a lot of bytes per object, and in any case, the server doesn't let you grab data for more than 5000 objects at a time. Now click the "22" toolbar button.
Nothing will happen immediately, though if you have a modem indicator on your screen, you'll see the simulated modem lights flash for anywhere between a few seconds (small area) and a few minutes (large area). You can keep on using Guide while all this is happening; the downloading of the GSC-2.2 data takes place in the background, in a separate process. But when it's done, Guide will redraw the screen and (assuming you haven't changed the screen center) you'll see the GSC-2.2 data, shown as stars covering the field (plus a bit of margin).
You can then zoom in and out on the data, and right-click objects to get information about them. In Guide, GSC-2.2 is actually displayed as a user-added dataset; if you go to "Extras... Toggle User-Added Datasets", you will see listed "GSC-2.2 data (as ellipses)" and "GSC-2.2 data (as stars)", each of which can be toggled separately. (By default, the former is turned off and the latter is turned on.)
If you turn the "(as ellipses)" dataset on, you can see that GSC-2.2 does a passably good job of showing shapes of galaxies and other elliptical galaxies. There are a few cases where the algorithm clearly got confused, turning "patchy" galaxies into a series of smaller blobs. Also, John Greaves pointed out an area where, when combining data from a 1984 plate and a 1998 plate, it appears that almost every star was incorrectly matched, so you get a slew of close "double stars"; and I found one case where the server returned a dog's breakfast of mixed-up magnitude data. But most of the time, GSC-2.2 seems to shows objects quite well.
When you use the toolbar button, Guide gathers the data and appends it to a file called GSC22.DAT. If you accumulate immense amounts of objects and want to go back to scratch, delete this file.
Astrometric accuracy: I've tested GSC-2.2 for astrometric accuracy by comparing it to the ICRF Sources list (a set of 667 objects whose positions are known to within about a thousandth of an arcsecond, measured by VLBI. They're used to define the International Celestial Reference Frame). The result was that 436 objects had matches within 2" of the ICRF location. 227 were "misses", and in four cases, there were multiple sources and no match was possible. For the matched objects, the average separation (ICRF - [GSC-2.2]) was .28 arcseconds. This happens to match, almost exactly, the results of a similar comparison of ICRF and GSC-ACT conducted by Paul Green, except that he found only 61 matches for 43 sources (some sources were measured on different plates). (I am indebted to Paul Green for suggesting use of ICRF for this purpose.)
This actually indicates much better accuracy in GSC-2.2, for two reasons. First, the GSC-2.2 comparison went down to a much fainter mag limit (thus the matching of about 2/3 of the sources, as opposed to less than a tenth). Second, GSC-2.2 has one sterling characteristic not apparent when dealing with ICRF objects: it's based mostly on plates exposed since 1990, whereas GSC-1.x is based mostly on plates from the '70s and '80s. So errors due to proper motions ought to be considerably less than they were in GSC-1.x (and immensely less than they are for USNO-A2.0, with its 1950s-based data).
Photometric accuracy: Watch this space. The same code described in the section on GSC-2.2 astrometric accuracy, which matched ICRF "astrometric standards" to GSC-2.2 counterparts, can be used to match LONEOS photometric 'standards' to their GSC-2.2 counterparts. All I need do is add some code to compute statistics from the result, and we're home and dry, or at least home and vigorously towelling off.
If the '22' toolbar button fails to appear: Several people have been reporting that new toolbar items aren't getting added the way they should be. (This happened, for example, with the LX-200 focus and chart rotation controls, and with the function to simplify movie-making.) One way around this is to edit the file TOOLBAR.DAT, bump up the number of toolbar buttons (given in the top line) by three, and add the following lines to the list of buttons:
2378 !web_gsc.bmp GSC 2.2 data via Internet 2410 !web_dss.bmp Grab DSS image via Internet 2411 !web_a20.bmp A2.0 data via Internet
Do that, and the buttons will appear (without even needing to go into the Settings... Toolbar menu.)
Alternatively, you can just click here to download a completely new TOOLBAR.DAT file (about 20 KBytes). The only downside to this is that your toolbar will return to default settings.
The problem arose because I added a "clever" scheme to Guide to allow it to add/modify buttons listed in TOOLBAR.DAT, without destroying your toolbar settings. When you got an update with, say, the new feature to download GSC-2.2 data, Guide was supposed to recognize that a new toolbar button for that function was to be added between two existing buttons. Unfortunately, this doesn't always work as expected.