Hi Stefano

I've finally found the .idf files with ucac3 stuff, you know people can't look at things unless they know where they are...

You have not tried putting the ; symbol in front of the photo line?

in ucac3.idf find the line marked photo and put a ; as the first character in a text editor.

This may be a red herring, but it might work.

The photo line is more extensive in ucac3, including mixed data. In ucac2 it uses just j, h and k for colour. Bill's, or possibly the raw ucac data, return null 2mass magnitudes with the value 30.000, so the ucac2 .tdf has this line :-

photo J:127,6 H:134,6 K:141,6

Bill must've a trap for the case where all, or some, of those are set to 30.000 in the data.

Now, for ucac3 the .tdf has the line :-

photo J:131,6 H:138,6 K:145,6 B:173,6 R:180,6 I:187,6

a mix of supercosmos photographic magnitudes as well as the 2mass ones.

If some slight mishap has occured in accounting for the null cases of these being represented by large values instead of null values it'll crap out will Guide when trying to draw those stars.

You may think the level 8 size is important, and indirectly it could be. The objects with null 2mass magnitudes are the faintest ucac3 objects. So, on auto magnitudes, as you zoom in, you go fainter, thus more likely to plot such objects. If ucac3 is set on they're plotted anyway. Whether it crashes on the drawing of the stars or whether it is drawing black stars, I know not. From what you say I don't think it gets as far as drawing the other catalogues.

And this could all be a red herring, rather than a rainbow trout. But try it and see.

"and I think also that all
guys doing astrometry are quite happy when they can improve the quality of their
job using the latest astrometric catalogue even if it is not perfect"

but they don't, they just convince themselves they have. It's a science, not a popularity contest. Large numbers of people being happy about something doesn't mean it's right. Most of them just do it because astrometrica includes it, no more no less, same thing happens with use of nomad.

John