I’m starting to do photometry on various targets with a recently purchased Seestar S50. I’ve previously done wide-field DSLR photometry and visual estimates (I will still do both, especially visual).
For a short period pulsator, it’s easy to take and work with a lot of images in a short space of time, e.g. a HADS or SXPHE star. Indeed, take SX Phe as an example. Is it considered preferable from an AID submission perspective, say for a full cycle, to submit many (perhaps even all, e.g. 300) observations or fewer (e.g. 50) representative observations via an appropriate grouping mechanism (e.g. median stacking)?
Hello! I was taught to divide the period by 50 and that gives a good time separation between images. For my short period work, I actually used more like 100 instead of 50.
For my short period work, I had more images than that since the integration time was short enough in each filter to give about double the recommended number of images. As a result, I would take two or three images in each B,V,I filter and then repeat that through out the run. Each subrun would be 10 to 15 minutes.
Lots of images during a run of say 6 hours. For analysis, I would stack the two or three images in each filter for each sub-run. Analysis and stacking was easy with MPO Canopus. I would then upload the several hundred data points. Best regards.
David,
I have done a lot of photometry over the years on delta Scuti stars, mostly HADS, and EBs. In my opinion rather than using any guideline about how many images might be taken or what the cadence should be, it is better to take images often enough to yield precision with your setup that allows accurate determination of the times of the light curve peaks (HADS) or the times of mid eclipse (EBS). The reason for this policy is that one of the main purposes of studying these stars is to obtain precise data for O-C diagrams.
I use a small aperture, short focal length system (camera lens, 200mm focal length) and a 12 bit CMOS astro camera. Many of my fairly bright targets (often 8th to 10th magnitude in V) are captured with an exposure of about 60 seconds, sometimes less and sometimes more, to maximize signal without the risk of saturating stars (12 bits is not much to play with). The eGain is often a few electrons per ADU.
Just the other night I took a light curve of ZZ Mic, a 9th mag HADS with a period 97 minutes. Exposures were 60 seconds. Times of maximum were determined on all four peaks of the light curve captured over several hours. When they were plotted on an O-C diagram. See attached LC and O-C (if it works - my first attempt to attach images), the four points overlapped in the O-C diagram (marked with a black vertical line). That’s pretty good on the y axis scale chosen.
Roy
Here is an idea for the programmer/AI folks out there. A program to type in four corners of sky and say do VSTAR phase plots for all the HADS in that square. Easy to spot the HADS that need more data for better plots. For all HADS that have no data, display " No data, fire up the SeeStar! "
From a professional prospective I measure each frame to get as many data points as possible. That is assuming each frame has a reasonable signal to noise, otherwise stack. I use the entire data set of magnitudes to do a first cut to the analysis. I might average over observations to smooth things out, but want to start with everything I have. The multiperiodic nature of SX Phe is a case I like to see as much detail as possible in the curve.
I would say that having more data points is better. You will not have a cadence issue with more data, but could with less.
There are observations in the AID from which I need to calculate additional times of maximum, then write a paper for the JAAVSO.