This is probably one for Ken and the VPhot Team, but I thought it was interesting enough to post. I am processing a number of time series analyses for a relatively new variable. The first analysis was fine. The second analysis had three images similar to the many images from the third analysis, which was not uploaded.
The attached Word file shows the third time series results. Note total garbage in the middle of the run in Figure 1. Figure 2 shows the correct placement of apertures for the variable, check and comps. During the middle of the run target apertures were misplaced while check and comp apertures were placed correctly on the image. Figures 2 and 3 show two typical misplacements.
Manually correcting the apertures does not work.
As the check and comp apertures are correctly placed on all images I examined, I don’t think this is a plate-solve problem. As two runs seemed to process fine (with three image exceptions), it does not seem to be a sequence file problem. This is a strange problem. Any suggestions?
Ed
Hi Ed:
Yes, certainly unusual! I'll ask a few questions/comments:
1. What settings do you have for Tools>Centroid Determination, when you are looking at an image?
2. Did you manually click the position of the target when you created your sequence or use the normal VSX catalog selection?
3. Can you share the problem image set to me at MZK?
4. The images certainly look similar (same quality) for those with and without the problem.
5. There is also no nearby companion star that might pull the centroid away from the correct location.
6. Did you plate-solve off-line with pinpoint before uploading?
7. When you look at the small thumbnails of the time series, does anything look unusual for the problem images?
8. Did the 180 degree shift occur after a meridian flip? I would have said no, unless you flipped the images?
That's a start at sleuthing?
Ken
Looks like this problem was correlated with the updates performed yesterday. A new analysis of the same data was successful.
Ed