Photometry

By Cyril Richard

In this tutorial, we will see how to do photometric analysis with Siril, with an applied example on the light curve of a variable star.

DY Pegasi, a short-period variable star #

Plate solving #

I have a sequence of 46 FITS images found some time ago on the Web. They are perfect for a tutorial.

The first challenge is to find the variable star in the image. It can be tricky when we are not familiar with the star field, which is the case here. The Image Plate Solver tool of Siril is in this case a very good friend. For this sequence of images, I don’t know the focal length nor the pixel size, which are input data for the tool. We will have to find them by guessing, trying various values until it works, for example by setting an arbitrary pixel size and modifying the focal length at each try. Quickly we got the result shown below. It was done on the first image of the sequence.

With trial and error, we can find parameters for which the plate solving suceeds. Note that the flip image if needed option that flips the image if it’s in a mirror was disabled to avoid making it different from the others of the sequence. — With trial and error, we can find parameters for which the plate solving suceeds. Note that the `flip image if needed` option that flips the image if it’s in a mirror was disabled to avoid making it different from the others of the sequence.

The console displayed the following astrometric parameters:

20:48:50: Findstar: processing...
20:48:51: Catalog PPMXL size: 411 objects
20:48:51: 27 pair matches.
20:48:51: Inliers:         0.926
20:48:51: Resolution:      1.387 arcsec/px
20:48:51: Rotation:       +0.48 deg 
20:48:51: Focal:        1338.67 mm
20:48:51: Pixel size:      9.00 µm
20:48:51: Field of view:    31' 48.15" x 28' 27.07"
20:48:51: Image center: alpha:  23h08m54s, delta: +17°08'31"

Now that the plate solving succeeded, we have access to a lot of interesting features. Here we will use the object search, available in the context menu of the image (right click on it), the Search Object... entry (also available as control-/). A small dialog box opens, in which we’ll type DY Pega and <ENTER>.

We input again the name of the variable star.

The star DY Pega becomes annotated, and this will help us greatly for the analysis.

Photometry of the sequence #

Now that we know where to find the star, we will start analysing it. First, the images have to be aligned, in case they show some movement between each other, to facilitate the analysis of the same star in all images. This will be very easy in our example since we have many stars visible on images, we can use the Global registration algorithm. We will activate the Translation only option however, for two reasons:

When not checked, rotation is used and Siril will create a new sequence of rotated images, which doubles the space required on the disk. With translation only, the shifts between images are stored in Siril’s sequence file.
Not rotating the images also avoids unwanted interpolation between pixel values, so we keep the images as pristine as possible.

Aligning images is a key step for sequence photometry analysis.

In my example, the registration executes quickly and without error.

21:47:25: Sequence processing succeeded.
21:47:25: Execution time: 8.31 s.
21:47:25: Registration finished.
21:47:25: 46 images processed.
21:47:25: Total: 0 failed, 46 registered.

Once all images aligned, we can actually analyse the photometry of the sequence. Let’s start with the variable star, drawing a selection around it (middle mouse click draws a square centred on the pointer generally large enough for that). Then, right click in the image, and select PSF for the Sequence.

Photometry is done on the variable star for all images of the sequence.

The photometry is run on the whole sequence for the selected star. It is now possible to display various plots in the Plot tab: star roundness, FWHM, Gaussian amplitude, magnitude, background level and X,Y positions in the images. The background level plot is of great interest. With it, we can see how images evolve during the acquisition session. To remove this bias from the measures, we will absolutely need to select more stars in the image, which will act as a reference against which the variable star’s background can be calibrated.

Sky background plot. We can clearly see that the images brighten as we move along the sequence.

Let’s focus on another plot: magnitude. It is given in relative magnitude, only variations mean something. We can see that it is bell-shaped, which often indicates a variable star.

Uncalibrated magnitude plot for the variable star.

Now we want to run the photometric analysis for reference stars. Ideally, stars whose magnitude is stable across the sequence must be selected. If nearby stars are unknown, another feature is very useful:

Select a star
Context menu (right click) -> PSF (not for the sequence this time)
Click on More details... in the new window
This opens a Web browser page with data on the selected star.

Here, for learning purposes, I didn’t push the analysis as far as finding stable stars in the image, and took 3 random stars. The variable star is variable enough for a probable good result any way. When choosing the reference stars, it is also a good idea to select stars whose brightness is alike the variable’s, that are not too far from it, especially if flat field correction is not done (it should be mandatory for this kind of analysis), and not on the borders of images in case there is some shift between them. After having selected each reference star, run the PSF for the Sequence analysis.

Three reference stars have been chosen in the image.

Now when we display the magnitude plot, the button Light curve button becomes active, and clicking on it allows, as its name suggests, the light curve of the variable star to be plotted. A data file is saved, useful if the analysis is to be shared or for further analysis. If gnuplot is installed, this will automatically display the light curve plot.

Light curve of DY Pega. — Light curve of `DY Pega`.