I have asked this question before, but didn’t got any response…
In VPHOT I load variables by clicking Catalogs → Variable Star Index. One can also click Catalogs → GCVS.
I don’t know exactly what the difference are between these two, but I feel that the first option give the most correct stars.
The problem is that both of these options load tens or hundreds of stars that don’t have AUID, which mean that they are not observed before, and for 99.9% are not interesting to report. Most of them are from big surveys like ZTF, ASAS-SN or Gaia.
For me that means that I have to click on all these irrelevant stars to remove them from each image. I know I can save a sequence and use that instead of loading the variabels, but sequences have some other drawbacks, which means that I prefer to load the variables.
Could there not be a third option where only variables with AUID are loaded? Then you only got variables that can be reported.
I have noted that the number of stars without AUID that appear in VPHOT have increased significantly and cause more and more inconvenience.
We recently added all of the GAIA DR3 variable stars to VSX. That took our total number of known/suspected variable stars from ~2M to ~9M.
I believe @SGEO is working on implementing a method to filter out these additional stars from VPhot. With that said, you might consider adding them to your program (if they have sufficient SNR) so we can either confirm or disprove the GAIA VARI team’s claims!
I use those stars all the time. First I weed them out. Get rid of ones that are to dim, have low SNR. Then look the rest up on VSX and weed them some more. If variability is not enough for me to see with my instrument, I delete them. If they are close to the edges of my images they get weeded. Then save a sequence. Then request AUIDs for the ones that look observable with my equipment.
I have never had a problem with a sequence.
Then I make an everything sequence that has all the stars or all the stars in half the image. Maybe 30 to 50 stars in each sequence. I run them through the photometry grinder and look at the std column of the time series results page to spot new variables. These AAVSO report files get saved to drives but almost never reported. It takes considerable work to declare a new variable. If something blows up in the field, a nova gets declared and I have old pre-nova data that I can go back and report.
Different people have different motives for their observations, no right or wrong. I have over a hundred variables that I follow and personally follow up and analyze. I don’t have the energy or time to do more. In addition, all new variables have names that are not memorable in the same way as the GCVS names. They feel so impersonal. You can’t think or talk about a star like Gaia DR3 3398443145556244224 in the same way as SU Tau.
One problem with sequences is that if you load the sequence and make a change, for example adjusting the position of a star or adding a star that VPHOT didn’t find, and then go to the photometry page and back to the image, the changes you made are lost. This doesn’t happen if you load the variables via the menu selection. There are some other disadvantages as well, such as not seeing any new comparison stars.