Open PHD2 Multistar Guiding -Part 1

I have always been interested in improving my astrophotos. My focus has always been on how to improve how I process my astrophotos and how to improve the autoguiding of my telescope mount. Because of my 30 years of experience as an Electrical Engineer, I focused on the software/firmware and hardware involved with autoguiding. With the introduction of PHD2 Multistar guiding, I immediately jumped on this bandwagon. By the end of 2023, I was still looking for improvements, so I decided to take a deep dive into the Open PHD2 Multistar Guiding software. I downloaded the Open source code and was able to build PHD2 Guiding on my computer. I studied how the Multistar is calculated and the decision logic it contains. I created programs that allowed me to study many, many debug logs and to display real-time data specific to both single-star and multistar guiding. Soon, I began to notice issues in the multistar code that either didn’t make sense or didn’t work well.

Over the winter of 2023/2024, I began making changes to the multistar code that I believe addressed the issues I found. I then built my own private version of Open PHD2 Guiding and shared it with a small group of people who served as software testers. By the fall of 2024, my version was fully debugged. I used my version for all of my 2024 astrophotos. I decided that I will put all of my findings into a 3-part series of blogs…you are reading part 1. Part 2 will detail the issues I found. Part 3 will detail the software changes I made.

Part 1 is an introduction to Multistar guiding. This is not an overly complicated algorithm, and hopefully, all readers will understand the concepts mentioned here. The following is an overview of Multistar guiding and its main parts. At the bottom of this blog, you will find links so you can download a program I wrote that will display real-time data from Multistar guiding and single-star guiding.

What is Multistar guiding?

Open PHD2 Guiding now has 2 feedback methods for guiding your telescope mount. The original feedback method is called single-star guiding. This method calculates the position of a single star to guide your telescope mount. The 2nd method is called multi-star guiding. As its name implies, it calculates the positions of multiple stars to guide your telescope mount. The logic behind multi-star guiding is that the quality of your guiding will improve if the information from multiple stars is combined.

How are these stars selected?

When the user clicks the Auto-select Star button, Open PHD2 will use the current exposure from your guide camera to identify a very large list of candidate stars. SNR, HFD, X, Y, and Peak values are determined and used to filter through this large list of candidate stars until a maximum of 12 stars are chosen. In some cases, a smaller list of candidate stars is chosen, typically because there were fewer stars available that met the selection criteria. The best quality star is selected as the Primary star. This is typically the same star that Open PHD2 would select if Single-star guiding were being used. The remaining 11 candidate stars are then identified as Secondary stars. The best 8 secondary stars, along with the Primary star, will be used to calculate the multistar, with the remaining 3 secondary stars available should one of the eight secondary stars be rejected by the multistar code.

How is the multistar calculated?

Open PHD2 uses a star centroid calculation to determine the X and Y pixel location for each star. The centroid calculation can calculate X and Y values with several digits of resolution. The centroid calculation is sensitive to noise, but fortunately, the noise becomes less of an issue as the star’s Signal-to-Noise ratio (SNR) increases. This is why the stars selected for the multistar have high SNR values. For each of the stars (Primary + 8 Secondary), their centroid values are initially used to determine each star’s Reference position. As each new exposure from your guide camera is captured, the centroids are calculated once again for each of these stars, and their differences between their centroid and Reference positions are also calculated. It’s these calculated differences that are used to calculate the multistar. There are two main steps involved in calculating the multistar…

1)The differences for each of these Secondary stars are tested, and if any star has too big a difference value (>2.5 Sigma), it will be flagged and not used to calculate the multistar. If this flagging happens too many times for an individual star, it will be flagged for RESET and assigned a new Reference position based on the current centroid value. This flagging logic can also determine if a star can’t be found. This typically happens when clouds move into the field of view. As new exposures arrive, PHD2 will continue to look for this LOST star. PHD2 can also detect if the difference value never changes for a star. When this happens, this star will be flagged as a hot pixel and ejected from the list of stars. The next available star in the list will take its place.

2)For the Secondary stars that were not flagged, their difference values are then combined with the difference value for the Primary star. The Multistar is calculated using a weighted average function that also takes each star’s SNR value into account.

Are there any special cases?

-If the multistar code determines that the difference value for the Primary star is too large (>5 Sigma), it will abandon the multistar and use the Primary star difference. This continues for the next few exposures until the Primary star difference returns to a small value (<2 Sigma).

-For each exposure, if there are no Secondary stars that haven’t been flagged, the multistar will be abandoned, and the Primary star difference will be used.

-When a guiding session starts, the Primary star difference will initially be used until the difference returns to a small value (<-2 Sigma). This is referred to as settling. A similar sequencing happens after a dither.

-If the Primary star difference is less than the multistar, the multistar will be abandoned, and the Primary star difference will be used.

How can I learn more about multistar guiding?

As part of studying my guiding, I created a few programs that allow me to examine what is happening in multistar guiding on an exposure-by-exposure basis. I decided to put together a program that displays lots of real-time information about multistar guiding. Most of the programs I write are written in a language called LabVIEW by National Instruments. National Instruments’ head office is located in Austin, Texas.

You can use this link to view a PDF document that describes what is being displayed.

I named this program PHD2Trends. It is a 64-bit Windows app designed to run while you are autoguiding. It will connect to PHD2 using the PHD2 Server and will read your PHD2 debug log. Your debug logs need to be created by PHD2 in the default directory, which is C:\Users\…\Documents\PHD2.

WARNING!!! Downloading these files using MS EDGE is an exercise in torture. Chrome doesn’t care, but the MS Edge browser forces you to jump through hoops. I just tried downloading using MS Edge. Edge will pop up a dialog. If you move your mouse over the file name, you will see a small delete icon and three dots. If you hover over the three dots, it will change to “more actions”. Click on “more actions” and select “keep”. MS Edge will still not be happy and will challenge you with yet another pop-up menu offering to only Cancel or Delete. You will see a “Show more” text. Click on “Show more” and click on the “Keep anyway” option. MS Edge will finally give up and let you finish the download. Do yourself a favor…set Chrome as your default browser.

Downloading and installing PHD2Trends requires you to download and install the LabVIEW Runtime Engine, which is a 200Mb library of support functions. Use this link to download the LabVIEW Runtime Engine. Once it is downloaded to your computer, you will need to run it so it can install the Runtime Engine.

Windows will typically pop up a dialog indicating that this file is not safe to run. This is because Windows has not seen this program before, so it is being cautious. I use anti-virus software, and the program is stored on GitHub, so it should be free of viruses. You can always scan the file first using your antivirus software. If you click on the “More info” text, you will be able to convince Windows to let you run the installer.

NOTE: Once this Runtime Engine is installed, you will be asked to reboot your computer so that the final installation steps can be completed.

Once the Runtime Engine is installed, you can use this link to download the installation file for PHD2Trends. As with the Runtime Engine, you will need to run the installer, and it will most likely cause a similar pop-up dialog stating that this file is not safe to run.

NOTE: The PHD2Trends installer will create a shortcut on your desktop. You will not need to reboot your computer.

All that is left is to get PHD2 up and guiding, and then run PHD2Trends. You could, alternatively, first get PHD2Trends up and running and then run PHD2.

I hope you enjoy using PHD2Trends.

Peter

Advanced Methods for comparing your guiding with multistar or single-star guiding…this includes comparing your guiding with my version of PHD2 and the official version of PHD2.

One of the simplest comparisons you can study is to switch between multistar guiding and single-star guiding. This is easily done by navigating to the “Guiding” tab in PHD2 Guiding. This is done by first clicking on the PHD2 brain icon at the bottom of the PHD2 screen. Then, by navigating to the “Guiding” tab, you can switch to single star guiding by removing the check mark on the “Use multiple stars” item and then pressing the OK button. The “Use multiple stars” items can be checked and unchecked while guiding…just remember that the change does not take place until you click the OK button. You will not have to recalibrate if you do this, so it is an instant comparison method.

If you want to compare your guiding between versions of PHD2 Guiding, including mine, you will need to adhere to the following restrictions.

a)Only one version of PHD2 Guiding can be running at any time…this includes any version that has been minimized.

b)As you switch versions of PHD2 Guiding, you may have to restore the calibration data. You can tell if you have to restore the calibration data if the “Cal” text displayed at the bottom-right corner of the PHD2 screen is displayed in red. To restore the calibration data, you should first click on the “Tools” menu at the top of the PHD2 screen. Once there, you should navigate to the “Modify Calibration” item, where you should select the “Restore Calibration Data” item. This will display your latest calibration data. Clicking the Restore button at the bottom of this window will restore your calibration data.