New PixInsight Process: Variable Stars Observer (VSO)

I am pleased to introduce Variable Stars Observer (VSO), a new PixInsight process dedicated to estimating the magnitude of variable stars. VSO leverages the AASO APIs to retrieve data for variable and constant stars, along with the DynamicPSF process to detect stars and measure their PSF.

With VSO, in just one click, you automatically:

  • Retrieve all the information about variable and constant stars available in your image,
  • Compute the average magnitude for each variable star,
  • Select a variable star near the center of your image,
  • Select a comparison and check star close to the average magnitude of the variable star,
  • Recompute the variable starā€™s magnitude.

Of course, you can easily select a different variable or comparison/check star from a dropdown menu.

A second click generates an AAVSO report, and a third click exports the report using AAVSO Extended Format!

Here is the full documentation:

The PixInsight repository link is:
https://pixinsight.deepskyforge.com/update/variablestarsobserver/
Copy/Past URL in PixInsight (including the trailing slash ā€œ/ā€).
Do not open link in a browser else youā€™ll redirected to my website.

For any questions or issues, please refer to the support page here:

Website VSO process:
https://www.deepskyforge.com/vso/

This is the first version.
Iā€™m seeking feedback from experienced astronomers in variable star analysis.

Joƫl

4 Likes

Hi Joel,

wow this looks really good, the long awaited PSF photometry in PI :star_struck:
I just tried an raw color image (bayer RGGB, correct stated in FITS header).

I only see channel 0, or what is that anyway? I might have expected TG TR TB ? And, I also see the detected stars, but in the B-V column is always ā€˜naā€™, all other columns contain values. The value contained in the magnitude column is the value calculated from the GAIA magnitudes according to the GAIA transformation formula to V ?

CS
Thomas

P.S. I just saw: It must be an RGB image, then you can also see 3 channels, and also the B-V column.
And, I assume (hopeā€¦) there are plans to extend the tool for complete image series/light curves ?

1 Like

Hi Thomas,
Thanks for your feedback.

Yes, process works only on debayer tricolor photos or monochrome photos.
When you use raw format, process interprets it as a monochrome photo, that why index B-V is ā€œnaā€.

Magnitude for constant stars comes from AAVSO VPlot (call API).
Magnitude for variable stars are calculated by the process.

I could implement the processing of a complete serie of photos but I need first confirmation that process gives correct results.

Yes! This looks very good. Iā€™ll give it a few tests but my backlog of many things to do with higher priority still get in the way.

Jim (DEY)

1 Like

Hello!

I have already installed this. Looks very good at first glance. I will give it a try, of course, and give some feedback.

Cheers!

Enrique Boeneker (BETB)

Thanks!

For your information, I will release a new version this weekend.

The process will prompt the user to select which filter should be considered when analyzing monochrome images (currently, the process only considers the DSLR Green filter).

I added some shortcuts to AAVSO Chart, Photometry Submit page and Submitted reports.

Iā€™ll keep you posted.

2 Likes

It stop working/Loading in my Pixinsightā€¦was the process removed?

No.

Please ensure you have PixInsight version 1.9.x.

In some situation PixInsight stop loading process and you have to reset updates.
Process to reset updates:

  1. Menu RESOURCES > Updates > Reset updates
  2. Restart PixInsight
  3. RESOURCES > Updates > Check updates
  4. Confirm upload
  5. Restart PixInsight and confirm updates

I hope this help.

1 Like

Iā€™ve used VSO for several nights now. Iā€™m using Seestar to capture data, followed by analysis using VSO inside Pixinsight. I love the fact that VSO automatically populates all the variable stars and comp stars in the field. Then VSO calculates the observed TB, TG, and TR magnitudes with estimated errors. VSO automatically tags saturated stars using a red flag (actually, a red star) so that they cannot be used for photometry. Iā€™m not quite sure how PSF photometry works, compared to traditional aperture photometry, but the results appear to agree very well with posted AAVSO light curves. This is a great addition to Pixinsight.

2 Likes

This comment implies that caution is needed. First, it is in general not a good idea to rely on a method if the principles behind it are not understood (a ā€˜black boxā€™). Second, while it is essential to check that new measures of variable star magnitudes agree generally with those from other observers, this is not the way to check precision and accuracy.

The latter should be done by testing your own observations using standard stars (preferably photometric standards) of known magnitude and colour, using one as a target and another as a comp star and taking multiple observations to allow statstical analysis.

Does VSO allow the above to be done, with standard stars selected by the observer?

Roy

1 Like

Yes, VariableStarObserver (VSO) allows checking of measured magnitudes with photometric standards of known colors. After successful completion of such checking, VSO could very well become a boon for people who are already accustomed to using Pixinsight (Credit is due Joƫl who developed the new VSO add-in process to Pixinsight. In its initial release, VSO is cost-free.)

In every photometric measurement, VSO employs a comp star and check star. As an indicator of measurement accuracy, VSO calculates the difference between the measured value of the check starā€™s magnitude and its accepted (i.e., standard) magnitude. In the case of one-shot color cameras, VSO calculates three such differences, one for each of the three RGB colors. In the case of monochrome cameras, I believe that VSO assumes the image was taken in the V passband, which in the future may be extended to other photometric passbands.

Because I have recently started using the Seestar S50, which is like a DSLR with 250mm focal length, my main interest is how well Seestarā€™s observed magnitude (TG filter) compares to the standard (V) magnitude.

As we know, the spectral bandpasses of TG and V are not exactly the same. Nevertheless, I did a spot check of the measured (TG) magnitudes versus AAVSOā€™s standard (V) magnitudes of nine comparison stars in the R Leo Minoris field. All nine AAVSO comparison stars were of moderate color with 0.5 < B-V < 1.0. The Seestar S50 integration was 30x10s for a total of 300 seconds. The new VariableStarObserver process was used to reduce Seestarā€™s multi-color FIT image.

The calculated differences between VSOā€™s measured magnitudes and the accepted standard magnitudes are presented in the table below, for these nine comparison stars in the field of R Leo Minoris. The differential expresses the measurement error relative to the comparison starā€™s standard V magnitude.

Comp Star Label Differential
000-BBQ-809 97 -0.003
000-BBQ-753 102 -0.005
000-BBQ-781 107 0.009
000-BBQ-737 111 -0.074
000-BBQ-726 116 -0.115
000-BNY-485 119 0.066
000-BBQ-749 120 -0.039
000-BBQ-743 127 0.074
000-BBQ-736 129 0.034

In the case of these nine comparison stars, the computed mean differential is -0.006 magnitude. This result suggests that VariableStarObserver and Seestar, working in concert with each other, can achieve very good accuracy.

This small snapshot could (and should) be extended to other AAVSO fields, to include comparison stars that are bluer than B-V = 0.5 and redder than B-V = 1.0.

I hope this post sheds some light into the VSO ā€œblack box,ā€ in connection with photometric analysis of multi-color RGB images taken by Seestar and similar imaging setups. The VSO add-in certainly shows promise for variable star observing, especially by observers who already have a Pixinsight license.

Andy,

Thanks for the information.

For each star in your list, was your measured magnitude from only one image, or from an average of measuements of several images? Each star in your list was a target. What paired comp star(s) did you use?

Roy

Roy,

The Seestar image was one image, i.e., a stack of 30x10s exposures that Seestar registered and saved as a single ā€œenhancedā€ image that is an average of the thirty individual exposures.

Working late last night, I wasnā€™t careful about using the same comp star throughout. So I redid the table with the following results, using comp star 000-BBQ-737 to make the magnitude estimate of each of the listed check stars. 000-BBQ-737 is the one that VSO suggested to use when estimating R Leo Minoris.

TABLE: VSO measured magnitudes of AAVSO check stars in the field of R Leo Minoris, using Comp Star 000-BBQ-737 as the reference.

Check Star Label Differential
000-BBQ-726 116 -0.042
000-BBQ-736 129 0.107
000-BBQ-737 111 na
000-BBQ-743 127 0.074
000-BBQ-749 120 0.034
000-BBQ-753 102 0.069
000-BBQ-781 107 0.082
000-BBQ-809 97 0.071
000-BNY-485 119 0.137

Now this works out to an average differential, Standard Mag - Measured Mag = +0.0665. The measurements make the check stars look about 6 or 7 hundredths of a magnitude brighter, on average, than their AAVSO standard magnitudes.

Best, Andy

Thanks Andy.

Iā€™ve read the documentation. I think it implies that only one image can be analysed. Is that correct?

Roy

Roy,

At the moment, it looks like images must be analyzed one-by-one. Hopefully, there will be a way to batch images in the future.

Best, Andy

Andy,

Thanks. OK, so canā€™t be used for time series at present.

Roy

Iā€™m currently working on the possibility to compute variable star magnitude based on a list of compare stars.
I also added the possibility to easly verify the quality of the estimate using any other constant star in the field.
After that, Iā€™ll add the support of time series :wink:
Iā€™ll keep you posted here.

Variable Stars Observer 2.0 is now available.

This major version allows the calculation of magnitude based on a set of constant stars (Select ENSEMBLE in Compare dropdown selector).

By default, all constant stars in your image will be considered, but you can easily reduce the number by removing them from the displayed list. It is still possible to define a check star, in which case it will not be considered as a constant star. You can still create and submit an AAVSO report. The list of constant stars used during the calculation will be included in the notes.

This version brings improvements to the user interface. Exporting a report is now possible directly from the main window. Many new pieces of information have been added, such as the local signal-to-noise ratio, the instrumental magnitude, AAVSO filter reference used.

If you want to check the quality of the estimates, you can load only the constant stars and select each one as a check star. The process will automatically display the deviation between the calculated magnitude and the actual magnitude. In the future, I should be able to identify the constant stars that provide the best estimates through a dedicated algorithm.

Finally, the CSV export has been updated to provide all necessary details for further analysis.

As reminder my website: DeepSkyForge Repository

Feel free to give me feedback.

Iā€™ll now work on the 3.0 that will make possible to process a batch of photos.