Hi,

I'm using theSKY6 for plate solving. While it pretty much always solves it occasionally returns an incorrect result. For example around ngc6888 it seems to find a solution about half a degree off every time.

Needless to say unreliable plate solve is a real issue for unattended imaging - especially as the mount generally nails the object and the plate solve is causing CCDAP to move away from the correct place!

So I have two questions?

First is it possible to get reliable plate solves out of theSKY6, or should I give up and just get pinpoint?

Second it is possible to get CCDAP3 to keep the plate solve images even if the plate solve is "successful"? I'd like to explore the issue when it happens and having the exact image that failed would be useful.

Thanks,
Darren.

Hi Darren,

I use CCDSoft/TheSky6 for plate solving and get very good results. I just finished up a series on NGC6888 and the plate solving worked just fine. As always, please post a log file on one of the runs where the plate solving failed as this offers a look at the equipment and events that took place and can supply the info needed for making good suggestions to improve your results. Thanks......

Hi Frank,

Generally it's pretty good for me too , but these occasional miscalculations are getting annoying. I'd appreciate it if you could look at the log.

Last night, for example, it solved M31 correctly, but got the focus star wrong in a seemingly random way - leading to inconsistent focus.

The ngc6888 log is at: http://gbdt.com.au/ccdap20080826_225825.log

Note that the first plate solve after initial goto reports a 33 arc min difference. A later correct M31 run puts the DECerror at around 0 to 3 arc min, but this is an a different part of the sky so the comparison may not be valid.

You can see why it would have been nice to have a way to keep the sync image it 'sucessfully' solved for examination - maybe the mount miss queued?

I've put up the first captured image (#584) at http://gbdt.com.au/87W_ngc6888_00584.jpg

It's a single sub, so there is color noise that I normally down worry about (dithering seems to take care of it pretty well). If you need them stacked let me know.

Thanks,
Darren.

Hi Frank,

If you can find time to give the log a quick look it would still be appreciated, but I may get you to hold off for a day or two.

I was looking through the SB guide for astrometry with theSKY ( http://www.bisque.com/SC/Download?File=Bulk/PDFs/AutoAs...etryAndImageLink.pdf ) and noticed one of their examples explicitly mentions ngc6888 and the UCAC catalog issue. I just checked and I still had it enabled (I've disabled it before, but it is what it is).

The only issue they mention is a plate solve failure if this catalog is enabled (why it is enabled by default is an interesting question!), not a false detect. Still I should give it a try and see if it helps.

One positive sign is that I was able to solve some previously-failed ngc6888 sync images when I disabled UCAC and the coordinates seemed correct.

I'll be able to do some testing tonight.

Thanks,
Darren.

Hi Darren,

I took a look at your log and all the parameters look good. One thing I might suggest is to turn on dithering for your dawn flats. This keeps any stars appearing in the exposure, moved off the same pixel, so a good sigma combine will remove any that happen to get picked up.

I only use the GSC and the Hipparcos/Tycho catalogs most of the time, especially when imageing with a wide FOV. If I'm using my 16" RC then I will turn on the UCAC2 and USNO B as the smaller FOV seems to benefit with the additional catalogs most of the time. But I find that they don't all work everywhere and sometimes you have to experiment to find the right catalog for the scope/camera/location combination. Good luck on your testing and let us know how it come out.

Hi Frank,

I did some testing with NGC6888. It would reliably get the plate solve wrong with the UCAC2 catalog enabled, and reliably get it right with the UCAC catalog disabled.

I'm actually imaging at the moment, but I'll turn on dithering for the flats. Do you suggest leaving the tracking on or turning it off?

I'll keep your catalog tips in mind when I get back to narrower field imaging. BTW how small an FOV do you find you need the extra catalogs for?

Thanks,
Darren.

Hi Darren,

Glad to hear it working better for you.

I leave tracking on, but it is probably a matter of preference. With tracking off, you would get star trails rather than star points, and some may say that combines out more uniformly. I've done it both ways and I never saw a difference in my master flats.

How small a FOV needs the extra catalogs seems to depend on whether you are in a star rich or star poor area. My procedure on a new target is after I have all my parameters setup and the run started, I watch the first plate solve and if it fails, I'll go into TheSky6 and adjust the catalogs I use until it solves.

Hi Frank,

I've been pretty successful with this lately - all my solves seem to be correct and reliable.

In retrospect it seems that the more important parts of getting this to work correctly were:

1. Guidestar catalog selection (as described above)
2. Setting the star extraction parameters in CCSoft to correctly identify stars

Having the plate solve images to analyze after the fact were very important. I noticed in another thread that a "trace" option is mentioned which retains images that were solved where CCDSoft did not return an error. Must hunt that option down in case I get another batch of bad plate solves!

Thanks,
Darren.

Hi Darren,

The Trace option is found under Help | Troubleshooting.