… after this ambiguous headline, it seems only fair to clarify what this night with “N.I.N.A.” will really be about: not romance, but a fully automated astrophotography session planned with precision and unexpected possibilities using the Advanced Sequencer tool.

A very good German-language introduction to N.I.N.A.`s Advanced Sequencer is provided by Frank Sackenheim in his Astrophotocast episode from February 19, 2023. I can really recommend Frank’s videos without reservation—they are professionally produced and offer an excellent, well-structured entry into the topic.

For the first time, I planned a complete imaging session using the Advanced Sequencer, and the level of automation it enables I do find really impressive. From conditional logic and automated error handling to real-time notifications via Pushover and a Telegram BOT, virtually every aspect of the imaging night can be defined in advance and then launched with a single click—right down to data storage in the desired file format and a fully customized directory structure written directly to my QNAP NAS.

The following article summarizes the key settings, prerequisites, and configuration steps required to achieve this level of automation within N.I.N.A.’s Advanced Sequencer.

Why All This Effort? 

The objective of this preparatory work was to achieve full automation of the entire observatory setup including:

• Pulsar Observatory Dome (rotation drive and shutter drive)
• OTA on the Skywatcher EQ8-R mount
• imaging cameras (actually ASI 2600 MM Pro and ASI 294 MC Pro)
• EFW 2″ electronic filter wheel
• EAF electronical Autofocusser
• CAA ZWO field rotator
• guiding equipment (Lacerta`s MGen-3 Autoguider + Skywatcher Evoguide 50 guide scope)
• the RB Focus Excalibur automated flat panel
• AAGCloudwatcher as safety guard
• and even auxiliary systems such as the heating of the guide scope or the primary mirror cooling fans …

As this list alone already makes clear, a considerable amount of equipment is involved, all of which must be perfectly coordinated and capable of operating automatically and reliably under a wide range of conditions.

Once all devices operate reliably and in dependency on one another, the sequencer software can execute a fully automated, predefined chain of processes. At the core of this workflow are the actual image acquisitions, which—often using multiple filters—typically occupy the majority of the night.

General Thoughts on using N.I.N.A.

One crucial advantage of N.I.N.A. is its manufacturer independence, allowing hardware from different vendors to be combined. This was the limitation of my previous setup based on the ASIAIR, which is restricted to ZWO accessories and therefore not ideal as a long-term solution. I have not completely set aside my ASIAIR, as the idea of a future side-by-side setup—where two telescopes operate in parallel—remains in the back of my mind. In addition, for possible mobile use of a smaller setup, the simpler ASIAIR-based solution still offers its advantages.

That said, switching to N.I.N.A. required a significant amount of effort, as the entire system essentially had to be set up again from the beginning. Solutions like ASIAIR, with their intuitive user interface and high level of built-in automation, are clearly more accessible for beginners. However, if full control over every step of the imaging process is the goal, there is ultimately no way around a more powerful and complex sequencer-based approach.

Safety First

One of the most important aspects to consider when planning a remotely operated observatory is operational safety. This primarily involves the ability to respond automatically to changing weather conditions. In practice, this means that whenever certain thresholds are exceeded—such as rain or strong wind—the shutter must close as quickly and reliably as possible in order to protect the sensitive and expensive equipment inside the dome.
Within N.I.N.A., this functionality is provided by the so-called Safety Monitor, which in my case is connected to the AAG CloudWatcher from Lunatico Astro. Every 60 seconds, it delivers current measurements for rain rate, wind speed, humidity, sky brightness, temperature, dew point, and air pressure. If predefined threshold values are exceeded, the shutter is closed automatically to protect the observatory equipment.With the rain sensor of the AAG CloudWatcher, just two raindrops are sufficient to trigger an alert, prompting N.I.N.A. to command the shutter to close. The entire process takes approximately 25 seconds.

Whenever the Safety Monitor in N.I.N.A. switches to an unsafe state, I immediately receive a notification via Pushover forwarded to Telegram. This ensures that I am informed in real time about critical changes in weather conditions, even when I am not actively monitoring the system. As a result, I can be confident that the observatory has reacted automatically and safely to protect the equipment.

Pushover Notifications Setup

I set up push notifications by first creating a Telegram BOT via „BotFather“ and linking it to N.I.N.A. using the generated BOT token. This BOT was then connected to Pushover, allowing N.I.N.A. to forward status messages directly to my personal Telegram chat. In addition to Safety Monitor alerts, notifications are also triggered whenever any selected sequencer processes fail—such as unsuccessful guiding calibration or autofocus errors—and it is precisely this level of feedback that makes true remote operation possible, allowing me to go to sleep with confidence that any serious issue will be handled as planned and reported immediately.

How does an Advanced Sequence now look like & work?

Put simply, a sequence can be understood as a predefined workflow for capturing multiple individual frames. Within a sequence, all devices are also activated in a predefined and controlled order. In principle, every sequence is divided into three parts:

• a start sequence,
• the actual imaging sequence,
• and a finally an end sequence.

Start Sequence

In my specific case, the start sequence is configured so that, after the automated launch of N.I.N.A.—handled via a PowerShell script—the first step is the automatic connection of all devices within the software. This is followed by opening the flat-panel cover, activating the primary mirror cooling fans to speed up thermal equilibration.

This now may sound a bit childish, but every time I watch the flat panel cover open automatically via the video surveillance system installed in the dome, it still feels oddly satisfying to me … 🙈😂🚀

After that, the mount is then moved out of its park position and the home position is confirmed as the starting point. The synchronization between the observatory dome and the mount is initialized, and the shutter is automatically opened once safe conditions are confirmed through the Cloudwatcher.
I usually start the sequence at least two hours before the beginning of nautical twilight. Within the sequence, it is configured so that cooling of the main camera only starts once nautical twilight begins (the Sun is then 12 degrees below the horizon). In winter, I cool the camera down to −20 °C and specify a minimum cooling duration of at least 10 minutes to ensure a stable and gentle temperature drop.

Until astronomical twilight—when the Sun is 18 degrees below the horizon—the entire system continues to cool down. As the final steps of the preparation process, the primary mirror cooling fans are finally switched off to avoid turbulent air currents inside the tube that could negatively affect image quality. In the final preparation step, the Mgen-3 guiding is set up by slewing the mount close to the ecliptic, where guiding calibration is performed and guiding is started. In my case, I chose an azimuth of 190 degrees and an altitude of 45 degrees for this calibration position.

Imaging Sequence

The sequence shown in the screenshot follows a clear, logical preparation and start-up workflow: First, the target (the wonderful galaxy PGC 8961 or UGC 1810) is defined with its coordinates, and the altitude graph on the right visualizes the object’s visibility over the night, including culmination and optimal imaging windows. The sequence then begins with a Wait for Time instruction, set to Astronomical Dusk, ensuring that the imaging run does not start until the sky is sufficiently dark. Once this condition is met, the mount performs a slew to the target, followed by centering and rotation, guaranteeing correct framing and camera orientation. Next, the filter is switched to the selected filter (in this case Luminance), preparing the optical train for imaging. An autofocus run is then executed to achieve precise focus under current temperature and seeing conditions. Finally, guiding is started, completing the preparation phase and allowing the actual imaging sequence to proceed with (hopefully!) stable tracking.

For this test sequence, I chose an L-RGB imaging plan, starting with two hours of exposures through the luminance filter, followed by one hour each in the red, green, and blue filters.

To monitor and control the exposures, so-called trigger conditions are defined in advance within the sequence. These include, for example, the execution of a meridian flip, during which the mount switches from the east side to the west side—or vice versa—when the target crosses the meridian. In addition, an automatic autofocus run is triggered after every filter change and again after every twelve completed exposures. Each individual exposure is set to 300 seconds, using a standard gain of 100 and an offset of 50.

After approximately five hours, once all light-frame exposures in this sequence are completed, guiding is stopped.

Flat Wizard and End Sequence

This part of the sequence defines the fully automated acquisition of flat-field calibration frames at the end of the imaging session.

First, the flat-panel cover is closed automatically, positioning the flat-field light source (RBFocus Excalibur) in front of the telescope aperture. This ensures a uniform and repeatable illumination of the optical system. Next, Auto Exposure Flats are executed separately for each filter (L, R, G, and B). For every filter, N.I.N.A. automatically determines the correct exposure time within the defined limits (0 to 10 seconds) in order to reach a target histogram mean of 50 percent, with a tolerance of 10 percent. The flat panel brightness is fixed at a defined level, while exposure time is adjusted dynamically.

For each filter, 30 flat frames are captured using 1×1 binning, a gain of 100, and an offset of 50, ensuring that the calibration data precisely matches the settings used for the light frames. Once all flats are completed, the system can proceed safely to the shutdown sequence, having produced a complete and consistent calibration dataset without any manual intervention.

Once the flat frames are completed—more precisely, once they have been transferred via Wi-Fi to the network storage—the mount is moved to its park position and the observatory dome follows accordingly. The camera is then gradually warmed back to ambient temperature, with a minimum warm-up time of 15 minutes defined. Finally, depending on temperatureিৱ humidity conditions, the observatory shutter can be closed immediately to safely conclude the session.

File Management

Another excellent feature of N.I.N.A., in addition to selecting the storage location for all data, is the ability to define custom directory structures for each target directly at the chosen destination. In my setup, all image data are written to the network storage, where a top-level folder is first created using the object name, followed by subfolders organized by image type (lights, flats, etc.). File names can also be fully customized within N.I.N.A.; in my case, they include the acquisition date and time, exposure duration, sensor temperature, HFR value as an indicator of image quality, and finally a running sequence number.

The next clear night—whenever this increasingly rare event may be granted to us—will show whether the chosen sequence in this test setup truly works as intended. For now, the dry runs have been promising.