Community tools and Hobby projects

Hi everyone 👋

I’d like to share a hobby project I’ve been building: stargazer.earth.

It actually started as a simple tool to notify me about good observing weather - because, as many of you probably know, clear nights are rare… especially when you’re a busy dad trying to plan ahead 😉

Over time, it evolved into a more complete platform for planning and tracking astronomical observations. With 120+ users across all continents, it’s both exciting and slightly overwhelming. I’m doing my best to make sure it’s truly useful in practice.

What it currently offers:

• 🔭 Smart weather notifications (fully customizable - cloud cover, precipitation, visibility, even azimuth limits if you observe from a balcony)

• 🗂 Equipment management (scopes, eyepieces, filters, cameras)

• 📊 Observation session tracking with duration logs & statistics

• 🌌 Built-in Messier, NGC & Solar System catalogs

• 🤖 AI assistant (Nova) for planning help and potential session summaries

The app is completely free and will remain so as long as I can cover the hosting costs. I may consider optional donations or small support tiers in the future, but for now it’s just a side project built by one astronomy enthusiast for fellow enthusiasts.

I’d love to hear your thoughts - especially from astrophotographers here on AstroBin.

Clear skies to everyone (and greetings to all busy dads out there!)

Łukasz

Hi everyone!

A few months ago I launched CelestView, an Android app that lets you open and view FITS, XISF, and TIFF files right on your phone or tablet.

Sometimes someone shares a FITS file with you and you can’t check it until you get back home to your PC. That used to happen to me all the time, and whenever I searched Google Play I couldn’t find an app that did what I needed—so I decided to build one.

Key features

• Astronomical image viewer: Open and inspect your captures with high fidelity directly on your mobile device

• FITS support: Easily open FITS (Flexible Image Transport System), the standard format in astrophotography

• TIFF support: Open stretched or unstretched TIFF files

• New: Open PixInsight XISF files (including files with multiple attachments)

• Quick export: Convert to common formats like PNG or JPG without using desktop software—great for sharing results instantly

Upcoming updates

I’m committed to continuously improving CelestView. Future updates (besides quality and performance improvements) will include:

• Support for viewing DSLR RAW images

If you’d like to try it out, I’d love to hear your feedback (features you miss, devices it works/doesn’t work on, etc.).

Download link

Hi all. I’d like to share ASCOM Alpaca ESP32 Dome (Roll-off roof) & Safety Controller for Remote Observatory

This project upgrades the classic Rolling Roof Computer Interface (RRCI) to the modern ASCOM Alpaca standard using a single ESP32 dev board. It replaces the old USB-tethered Arduino solution with a completely wireless, OS-independent system for roof control. We have relay that’s controlling the roof motor, open, close and safety sensors. Safety sensor is attached to the mount while in the park positions so we have physical sensor on park.

📷 esp32_setup.jpg

The controller presents itself to your astronomy software (NINA, Voyager, SGP) as two distinct Alpaca devices:

• Dome Device: Handles roof movement (Open/Close).

• Safety Monitor: Prevents the roof from moving unless the telescope is safely parked.

****Dome Alpaca endpoints (Open/Close) do not enforce the park sensor and behave like a classic RRCI controller.****

1. ****The Safety Alpaca device exposes the park sensor state; users who want interlocks should configure their automation (e.g. NINA “Safety Monitor”) to require this device to be safe before issuing dome commands.****

2. ****If you want safety sensor for parking you have to use Safety Monitor driver in your astronomy software (like NINA). Then roof movement is blocked for unsafe condition (mount not parked)****

It eliminates the need for ASCOM Platform drivers on the client side—just connect over IP!

Features

1. Alpaca Native: No Windows drivers required. Works on Linux (Indi), Mac, and Windows purely over WiFi.

2. Integrated Safety: Physical "Mount Parked" sensor input acts as a hardware interlock. The driver rejects commands if the mount is not parked.

Dual Interface

1. ASCOM/Alpaca API: Automated control for imaging sessions.

2. Web Dashboard: A mobile-friendly manual control panel hosted on the device (Port 11111).

3. Manual Override: The Web UI includes "Override" buttons to force movement during testing or emergencies, bypassing safety sensors.

4. Low Cost: Built for ~20€ using standard components.

Hardware Required (~20€ on Aliexpress)

• ESP32 Development Board: (Recommended: ESP32-WROOM-32U with external antenna connector for better range inside metal domes).

• ESP32 Expansion Board (38-Pin): For easy screw-terminal connections.

• 5V Relay Module (1-Channel): To trigger your existing roof motor controller.

• 3x Magnetic Reed Switches (Normally Open): For detecting Roof Open, Roof Closed, and Mount Parked status.

Pinout & Wiring

Refer to User Settings in the attached sketch.

1. Relay Trigger GPIO 26 Connect to Relay Input (IN). Triggers the roof motor.

2. Roof Closed Sensor GPIO 32 Connect sensor between Pin 32 and GND.

3. Roof Open Sensor GPIO 33 Connect sensor between Pin 33 and GND.

4. Mount Safe Sensor GPIO 27 Connect sensor between Pin 27 and GND.

Note: The sensors use internal INPUT_PULLUP. Wire them to switch to Ground when active.

Libraries Needed (ArduinoIDE):

1. ESPAsyncWebServer

2. AsyncTCP

3. (Standard ESP32 WiFi libraries)

Configuration:

1. Open the sketch and locate the // ========= USER SETTINGS ========= section.

2. CRITICAL: Change ssid and password to match your observatory's WiFi.

3. The default port is set to 11111.

How to Connect (NINA / ASCOM)

Because "Alpaca Discovery" (UDP) is disabled to save resources for the web server, you must add the device manually.

1. Open ASCOM Diagnostics.

2. Choose Device > Choose and Connect to Device

3. Select Dome as Device Type > Choose > Alpaca > Create Alpaca Driver

4. Add Dome name eg. "ESP32 Dome" and click OK

5. Enter the IP Address of your ESP32 when it connects to your WIFI

6. Enter Port: 11111.

7. Click OK.

8. In the main Device Connection Tester Window (from step 2) Select Choose and pick your newly created Dome (ESP32 Dome), click Ok and then Connect to test.

9. Do the same for Safety just pick Safety in "Select Device Type" dropdown in step 2

Web Dashboard

1. Access http://[YOUR_ESP32_IP]:11111 in any browser.

2. Standard Web Controls: Respect the "Mount Safe" sensor.

3. Override Controls: Ignore sensors (Use with caution!).

4. Status: Shows real-time feedback for "Roof Open", "Roof Closed", and "Park Sensor".

📷 esp32_UI.jpg📷 esp32_alpaca.jpg📷 esp32_alpaca_safety.jpg

Safety Warning

!!! USE AT YOUR OWN RISK !!!

Remote observatory control carries risks of equipment collision.

The "Mount Safe" sensor implementation relies on your physical sensor placement.

Always have a backup method to cut power to the roof motor in case of software failure.

TODO: My router allows me to make a static IP based on MAC from my ESP32 board (which is what I did), but some people could benefit from static IP setting in sketch file so it's on my to do list.

Link to esp32 ino file https://pastebin.com/QkQr1W9e

Hi! I made something I think is awesome :)

I have a 360 camera, and wanted to get PERFECT horizon files into nina and stellarium. So I made a simple webbased tool (that does ZERO server-calls, and ZERO logging - save the page to your disk for offline usage when travelling to star parties!)

https://neuronburner.com/hrz-creator/

There’s a Load Example button that uses an image as an example. It displays your 360 image in two projections at the same time. Have fun!

And then it exports .hrz files etc :)

📷 image.png

Hi!

I am pleased to share a fabulous experiment: I have implemented and exposed an MCP Server for PixInsight. And also a set of skills / instructions. As a result , I work out my processing flow in Claude (chatgpt style), and Claude drives PixInsight to run everything. Processing an image starting from the raw integrated images takes really just 10-15 minutes (just the processing time) even for more than 20 processing steps. It is really great and revolutionary. I had posted it a few days ago but I guess I did an error as I don’t find my post anymore, I hope I was not considered as spam because it is a real project, so I start again.

You can find the sources of my project on GitHub: https://github.com/aescaffre/pixinsight-mcp

I posted videos in a comment down the thread.

Happy to discuss more if it raises interest. Here is a NGC 891 processed with it (it is a crop, C9.25 with 0.63 reducer, the frame is much larger)


📷 Capture d’écran 2026-02-20 à 00.08.28.png

Thank you for your nice comment, yes, it is not uncommon to see thousands of frames getting accumulated by lack of time. That’s definitely one use case.

But the other is that , if you are picky, and looking for perfection, you can achieve like 20 various tests , even asking Claude to try and learn, and it works in no time. And this is still kind of the genuine way, we can build a lot on this, and I guess it has the potential for raising the level end results. I’ll share a video by the end of the week end.

Hi Alain, what a great idea! I definitely want to try this myself. I’m really looking forward to the video, as I’ll need some guidance to set it up on my own. Thanks for recording it!

Hello Alain
… and wowww, great.

As I am using Google-Gemini, I instantly started a thread with Gemini, whether this can also work with Gemini.
The answer was yes, plus step-by-step installation guide for my MacBook.

Is it okay for you that I download your githup code as zip … and what are my duties to use it?

Best regards.

Michael

Michael · Feb 22, 2026 at 05:58 PM

Hello Alain
… and wowww, great.

As I am using Google-Gemini, I instantly started a thread with Gemini, whether this can also work with Gemini.
The answer was yes, plus step-by-step installation guide for my MacBook.

Is it okay for you that I download your githup code as zip … and what are my duties to use it?

Best regards.

Michael

No problem with using it, it is open source, MIT License. Also it works with any agentic framework as. principle, but it needs to be on desktop, not the ChatGPT or gemini online services nor from your phone obviously.
Also I really think Claude is the best (clode code or Claude desktop)

Interesting. I was thinking about attempting this with an agent through Openclaw. I already have a dedicated and isolated OpenClaw box running where I can just drop files for a (multi-session) target and it manages my library and automatically stacks it using Siril. The obvious next step is processing but that is a bigger deal. Good job!

Kevin Guion · Feb 21, 2026, 09:47 AM

Backfocus Calculator — Visual Optical Train Builder

- Visual builder — telescope on the left, camera on the right, drag parts in order, real-time backfocus calculation with color-coded gap

- 12,100+ product database — 125 brands, 22 part types, 29 connection types (M42, M48, M54, EOS, Nikon Z, Sony E, ZWO bolt mounts…)

- Connection check — verifies thread diameter and male/female pairing, warns on mismatches

- Auto-complete — finds 1–3 part combinations from your inventory to fill the remaining gap

- FITS/XISF Backfocus Analyzer — star detection, PSF fitting, radial/tangential elongation diagnosis, 3x3 mosaic view inspired by Ekos Aberration Inspector

- Save/export/import — individual configs or full data as JSON

No external dependencies for the core app. https://github.com/ARP273-ROSE/backfocus

---

Both run on Windows/Linux/macOS, bilingual EN/FR, dark theme.

Feedback and suggestions welcome!

Hi Kevin!

Very interesting and helpful tool to visualize and help find the right back focus for those that are having trouble with this subject.

One question though, how would you take filter thickness into account, when applicable?

One typically needs to add 1/3rd of a filter’s thickness to the total back focus (so 55mm back focus becomes 56mm when using 3mm thick filters).

Filters aren’t part of the catalogue, but perhaps it would be something to point out when selecting a filter wheel or drawer for the optical train?

I’ve been incorporating AI tools more deeply into my professional work over the past few years - using them to structure thinking and build small utilities around real problems. Recently, I started extending that exploration into astrophotography as well.

The current project I’m working on is a simple deep-sky target planning tool. The goal is straightforward: help answer the question, what should I image tonight based on the equipment I have and will it actually frame well with my setup? The tool looks at what’s in the sky for a given night and location, and then evaluates suitability based on practical constraints like focal length, camera sensor size, and general visibility window.

Instead of just browsing very long object lists, the idea is to see targets that suit your equipment. Will this nebula meaningfully fill the frame? Is this galaxy going to be too small at my focal length? Does the target stay high enough in the sky for a reasonable imaging session? It’s less about complex modeling and more about bringing these considerations together into one planning view.

This isn’t a commercial effort - it’s primarily a learning and experimentation space where I’m exploring how AI can support structured planning in our hobby. That said, I’m very open to thoughtful feedback or serious discussions around workflow, framing logic, or how others approach target selection. explore it here.

Link: AI Playgound - Dr Gopal Krishna Kunwar Observatory

📷 Target Selector-Banner.jpg