From Buffering to Black Holes: How Streaming and Amateur Astronomy Are Following Parallel Paths

In this week's Friday Musings, I want to share something a bit different – a few astrophotography shots I've captured over recent months with my Dwarf2 smart telescope. These images of the Horsehead Nebula, Orion Nebula, Whirlpool Galaxy, and the Andromeda Galaxy represent more than just pretty pictures; they're the culmination of a teenage dream to photograph all the Messier objects. It's fitting that my early passion for astronomy led me to a career in streaming, and now I'm watching streaming technology transform amateur astronomy through devices like the Dwarf2.

From 1990 to 1992, while volunteering at the local University in holidays before my astrophysics degree, I encountered what I now recognize as one of the earliest examples of streaming over IP networks – though we didn't call it that back then. I worked with the Starlink network (not 'that' Starlink!), an interconnected IP Network of VAX Mini computers linking international observatories that had emerged as a response to the 1987 supernova. My job? Processing data from million-dollar CCDs – 1000x1000 pixel arrays that captured photon counts from binary star systems on some of the worlds largest telescopes.

The parallels to modern streaming are striking. We were essentially running a primitive content delivery network, albeit one where latency was measured in light-years rather than milliseconds. Each "frame" from our telescopes was painstakingly collected onto magnetic tape drives, processed pixel by pixel, and transformed into intensity maps. These would eventually become stop-frame animations showing binary stars in orbit. The workflow mirrors today's streaming pipeline: content origination, transfer, computational processing, rendering, and finally, delivery to viewers who would assess what we'd now call QoS (Quality of Service) and QoE (Quality of Experience).

The Software Revolution

Just as software tool like OBS and VLC have transformed live streaming from a specialized professional task into an accessible creative medium, tools like Stellarium and Siril have revolutionized amateur astronomy. Stellarium serves as a virtual planetarium and telescope control interface – think of it as the astronomy equivalent of a streaming production switcher. It helps plan observations and can even control smart telescopes directly, much like how streaming software manages multiple video sources and transitions.

Siril, on the other hand, is the astronomy world's answer to media encoding and processing software. Where streaming engineers use FFmpeg and hardware encoders to process video streams, astrophotographers use Siril to stack, align, and process multiple frames of astronomical data. Both domains are essentially solving the same problem: how to process large amounts of visual data efficiently while maintaining quality.

The Democratization of Complex Systems

Just as streaming evolved from those early days of academic networks to today's consumer-grade 4K experiences, amateur astronomy is undergoing its own renaissance. Where we once needed million-dollar CCDs and access to academic networks, smart telescopes like the Dwarf2 now put sophisticated astronomical imaging in anyone's backyard.

In the 90s, creating those binary star animations required months of data collection, manual processing, and specialized knowledge. Today's smart telescopes handle everything automatically – from star alignment to image stacking – much like how modern streaming platforms have abstracted away the complexity of video delivery.

Edge Computing in the Stars

Back on Starlink, each observatory was essentially an edge node in our primitive network, collecting and processing data before sharing it across the academic network. Today's streaming architectures push processing to the edge for lower latency and better performance. Smart telescopes like the Dwarf2 follow a similar principle, processing image stacking and enhancement directly in the telescope, much like how edge nodes handle video transcoding.

The Role of AI and Machine Learning

The manual pixel-by-pixel processing I did in my gap year has been replaced by sophisticated AI systems. In streaming, we use AI for content recommendations, quality optimization, and predictive caching. Similarly, smart telescopes use neural networks for object recognition and image enhancement. What once took months of manual work can now happen in real-time.

AI poses questions for the Astronomy sector too: As a quick example I passed the cover image for this article (Andromeda Galaxy) through Siril as part of the processing. There is a tool that removes all the stars, to allows you to grade the background object (in this case the galaxy) and then re-introduce the stars to create a controlled balance of brightness. This effect allows the editor to produce a much more enhanced image - much like background replacement can enable fake news to be produced! It raises questions of artistic purity, and even, potentially, ethical ones!

What i do like about the image is it shows what the Andromeda Galaxy would look like to a traveller who has left the milky-way (our own galaxy) behind them and is seeing Andromeda as they approach from an interstellar perspective...

Real-time Processing and Feedback Loops

Back on Starlink, those binary star animations required patience – lots of it. The feedback loop between data collection and viewing could span months or even years. Modern streaming and amateur astronomy both operate on much tighter cycles. Smart telescopes continuously adjust their positioning and imaging parameters based on environmental conditions, much like how adaptive bitrate streaming adjusts video quality based on network conditions.

The Social Layer

The Starlink network was, in many ways, an early astronomical social network, allowing observatories to collectively respond to celestial events. Today's streaming platforms have evolved into social spaces where content creators and viewers interact. Amateur astronomy has followed suit, with smart telescopes integrating into mobile apps and cloud platforms, enabling real-time sharing and collaboration.

Looking Ahead

As I reflect on my journey from those early days of astronomical data processing to today's streaming landscape – while ticking off Messier objects from my teenage bucket list – the technological parallels are clear. Both fields have evolved from specialized, high-latency systems requiring deep expertise into accessible, real-time platforms that anyone can use.

The future promises even more convergence. As streaming moves toward immersive experiences with VR and light field video, amateur astronomy is pushing boundaries with multi-telescope arrays and citizen science initiatives. The Dwarf2 and similar devices are just the beginning, much like how those early academic networks laid the groundwork for today's sophisticated streaming systems.

For streaming engineers, these parallel evolutionary paths offer valuable insights. Whether we're delivering the latest Netflix series or capturing images of distant galaxies, we're all working to make the impossible accessible, one frame at a time – though thankfully, we're no longer measuring latency in light-years.

---

The Commercial Bit

If you want help building premium, scaled up live streaming workflows and platforms do give me a shout at [email protected] - we have been doing it a loooooong time and have some great tools and software we licence out to developers and systems integrators :)

(Think of norsk.video as infrastructure-as-code for live video, but smarter. Instead of stitching together ten different vendors and hoping it all works, you just tell Norsk what you want to achieve. Clean REST APIs on top, deep technical control underneath when you need it. Use what you need, build what you want....)