Tue, 02/22/2011 - 05:36
Hei Geir,
I'm having trouble saving a check star. I've been trying to process my U Gem run from Feb 17 UT. When I tried the time series, it says something to the effect of "you don't have a check star". So I take a sequence, say u_gem_new3, add a check star, and save it as u_gem_new4.
I try to do the time series, and it says: "you don't have a check star". Sure enough, I open a U Gem image, select sequence u_gem_new4, and the check star is gone!
Do we have to start from scratch if we're trying to set up a new sequence?
Thanks,
--Bill
You're right. I did not convert the oldest sequences in the system (it was a bit complicated), thought they were never to be used again. Then you woke up ...
So yes, you will have to create the sequence again from scratch. Sorry about that.
Geir
Hei Geir,
I'm baaaaacccckkkk. OK, this time I created a new sequence from scratch, starting with a download of AAVSO sequence stars, then tossing the brightest ones out. I also picked a check star. This sequence I saved as u_gem_new5.
Select the 17 Feb run, push Time Serices, select my u_gem_new5 sequence and:
No check stars was identified in the first image. Please check your sequence, and that the first image is OK.
Now this time, when I open an image and select u_gem_new5, I do see the check star. The first image looks fine. So ... do I need multiple check stars for this to work?
--Bill
That is strange.
No, you only need one comp star.
So, knowing that you wouldn't mind, I logged in to your account and ran a time series using u_gem_new5. It worked without problems, you can see the resulting light curve in the analysis log.
Do you have the same problems after restarting your browser?
Geir
Hei Geir,
Yes, same problem, but a clue! I checked your light curve in the analysis log, and you used the latest 22 Feb run. I was trying to do my 17 Feb run. When I tried the 22 Feb data, it ran OK, but when I did the 17 Feb run, same error message!
Give the 17 Feb U Gem run a try.
Thanks!
--Bill
Aha! The problem is that the check star is too faint. There is a setting when you start the time series - Min. SNR - that specifies the minimum SNR a star must have to be included in the analysis. This setting applies to comp and check stars, not targets.
So simply lower the min. SNR to do the time series analysis.
There is a similar setting when you analyse a single image, look under the 'Tools' -> 'Settings'
Geir
Hei Geir,
Tusen takk!
That was the issue. I like having a check star about the same brightness as the target, so I can see quickly what the noise levels are like. When I droppd the SNR, the time series ran fine (might want to update the error message).
Thanks again! --Bill