Of Ancient Design
An undiscovered island. An ancient wreck.
It is July 1810. The brig Perseverance is upon the high seas, the wild Southern Ocean. Ship’s Captain Frederick Hasselborough is searching for new sealing grounds, with a crew of 25 men. On this gloomy, squally day, Hasselborough is certain he is the first to set eyes upon the uncharted emerald hills of Macquarie Island. Rowing to shore, the Captain and 6 men explore the mysterious uninhabited island, revealing massive numbers of seals and penguins around the rugged coasts, awaiting slaughter. The elemental landscape, wracked by constant winds, are colonised by strange plants unknown to science. All is new and unknown.
But what is that? Something hewn from from wood is poking through the tussock grasses near the beach. Approaching the object over a headland, it emerges, massive timbers sunk fast in sand, a wreck of ancient design, carved with strange motifs. Someone has been here before. Someone unknown.
Production Background – see my blog here
Macquarie Island is a small, remote island about half-way between Tasmania and Antarctica, set amidst the stormy Southern Ocean. Officially, it’s a part of Tasmania, and managed by the Tasmanian Parks and Wildlife Service. It’s a remarkable place – notable for its geology, wildlife, scientific significance and fascinating history. The definitive social history is covered in exhaustive detail in Cumpston’s Macquarie Island (1968) – to my delight I managed to pick up a first edition hardback through Abebooks. Other sources are more scientific: Selkirk’s Subantarctic Macquarie Island: Environment and Biology (1990), Terauds and Steward’s Subantarctic Wilderness: Macquarie Island (2014). In addition, there are hundreds of scientific and historical publications accessible online or held in various archives: a formidable treasure trove of information. They record and reflect the devastating impact of more than a century of sealing, slaughter and exploitation – from its discovery in 1810, until its cessation in the mid-1920s. This history is truly a picture of different times, where the natural world was seen as a source of endless abundance and dominion. In many respects it is hard to imagine the eighteenth- and nineteenth-century economic incentives and social conditions that drove people onto tiny ships, that entered into the deep unknown of the fiercest oceans on Earth, in the hope of finding their fortunes through the mass slaughter of millions of wild animals. In others, it is as relevant today as it ever was: a parable of human greed and ignorance. Fortunately the course of the twentieth-and twenty-first centuries traces the end of this industry and the gradual recognition of the island as a site of significant scientific and cultural importance. Indeed, today, it can be seen as a kind of sentinel that traces human impacts upon the natural world, and a symbol of hope, in that depredations can be undone, and life restored.
Captain Frederick Hasselborough, credited with the discovery of the island, is a fascinating character, about which little is known. Judging from the few notes he left behind, generally cited at one remove, he was clearly a literate man living in largely illiterate times. Whether this was common amongst sealer captains is unclear to me, but it is indicative of his socio-economic background and social class. It’s a pity not to have more of his direct observations. Although he’s referred to as ‘an Australian Sealer’, his name is variously spelt Hasselburgh, Hasselburg, Hazelburgh and Hasselbourg throughout various texts. Archive legal documents at the Australian National Library reveal his signature:
The autograph is clearly ‘Friedrich Hasselburg’ – suggestive of a Germanic origin. Given that the First Fleet arrived at Botany Bay in late January 1788, it is difficult to imagine that he was born in Australia. If he was a Captain in his 30’s or 40’s, I’d surmise he was born sometime between 1760 to 1780. Perhaps earlier in life, he worked with the Dutch East Indias company, and this connection might account for his being in Australia – but it’s guesswork. An interesting internet or archive rabbit-hole awaiting further exploration.
We can discern a few details of his life from McNab’s Murihiku and The Southern Islands (1907) and Entwisle’s Taka: A Vignette Life of William Tucker 1784-1817 (2005). Apparently, Hasselborough was separated from his un-named wife and involved in some court proceedings over property back in Sydney, suggestive of an unhappy marriage. During this first voyage he may have travelled with Elizabeth Farr – perhaps the first woman to set foot on Macquarie Island. Little is revealed about her. Tragically, upon the second return voyage en route to the island their deaths occurred in Perseverance Harbour, at Campbell Island – as reported in The Sydney Gazette 12 January 1811. Clearly there is a lot more to this story than meets the eye.
In a mere ten years after its discovery, this paradise for southern wildlife was utterly transformed. Captain Douglass, Master of the Mariner, provides a description of life there in 1822:
Distilling the Story
This is a complex story – a human and non-human drama with many historical threads that could be explored – and many fascinating themes to tease out. I had to distill it down to a few elements that could be covered in, at most, 5 minutes, without too much exposition: a ship heads South on a voyage of discovery; new uncharted land is discovered; seals and penguins are seen and heard – ripe for exploitation, slaughter and greed; a mystery reveals itself. A classic adventure mystery, yet a true story, with an element of hubris.
Of course, the main mystery proposed is the ship ‘of ancient design’ that Hasselborough refers to in letters written to the Scottish merchant Robert Campbell, the owner of the Perseverance. No more details are provided, and the ship was never seen again – presumably being burned by sealing gangs left upon the island, as there are no sources of wood available – no trees. It vanishes like a ghost from the pages of history, a mere footnote. Was it of Polynesian origin? A Maori vessel? A Dutch galleon? A Chinese treasure ship? What happened to the people aboard her? Were they marooned to die upon the island, perhaps decades, perhaps a century before the next humans to stumble upon this remote, forbidding crag? What sort of people were they? It remains an enduring maritime mystery that will never be resolved.
Even cutting it down to this skeletal level brings a lot of challenges – how to model and animate a convincing seascape, replete with spume and storms, an animatable cloudscape and sky capable of accurate geolocation and simulation; a 160-ton brig built in the style of the late-18th, early-19th century; a crew of 25 men (and one woman), dressed in period costume, that can walk and talk and interact with the world; an accurate model of Macquarie Island; representative models of Elephant seals, penguins and other fauna; representative models of the unique flora of the island and finally, a way of visualising the mystery ship, without being too specific.
Looking at that list now, it was a bit mad to imagine I could do all this in such a short time, entirely on my own – but as I show – it is do-able using tools available with Unreal Engine (UE), especially its intuitive Blueprints visual programming interface, various plug-ins that extend its capabilities, a range of original, free and paid assets from around the web, and appropriate accompanying software such as the Reallusion toolset and UE pipeline for ‘digital human’ character creation and AR-driven mocap. And of course – thanks to the great UE documentation, the YouTube training we received and the myriad online tutes and tips posted by UE users from all around the world.
This is work-in-progress – below I’m aiming to provide a series of technical notes and visualisation considerations that informed each shot or sequence – including some of the many problems encountered along the way. More soon!
Robert Campbell’s introductory monologue is based upon letters of commission published in Cumpston (1968), setting the narrative in motion. I dithered between leaving the screen entirely black and having the voice over some kind of visualisation of the ship’s voyage from Port Jackson to New Zealand, the Auckland Islands, the discovery of Campell Island and finally Macquarie Island. Serendipitously, I came across the marvellous David Rumsey Historical Map Collection and did a search for the island. Lo and behold – the 1828 J.W. Norie & Co (London) map of the Tasman Sea and New Zealand. It couldn’t have been more perfect – this map was published a mere 18 years after the island’s discovery – which in those times would constitute perhaps the most current and up-to-date data available. It is a beautiful thing – and generously released by the collection under a Creative Commons 3.0 CC BY-NC-SA license. It’s actually a pity to rush across it in such a short sequence, as the detail is extraordinary, the quality of printing and the sense of old paper and discovery so alluring – and the omissions fascinating – what is not shown also reflects the state of contemporaneous geographic knowledge. But, of course, this story is about the time before the map – when Campbell and Macquarie Islands were not even known to exist.
The voice part was read by my brother, Alex. We both share some ability to put on Scottish accents thanks to the Scots side of the family. In a competition between my accent and Alex’s, he is the expert, so he kindly donated his voice for the opening monologue. He’s not an actor, he’s a GP. Of course, thanks to COVID and travel restrictions this had to be done remotely – so it was recorded on his iPhone at home in Fremantle and emailed to me here in Tasmania.
The establishing scenes and title overlays introduce us to the journey ahead. They break down into three broad sections – Leaving Port Jackson, Heading South and Southern Ocean. The main challenge here was to portray an appropriately architected Tall Ship (a Brig) on a voyage from temperate climes to the wild Southern Ocean, presenting months of voyaging across thousands of kilometres, in a few minutes.
I wanted to capture some of the visual style of 18th and 19th-century maritime seascapes – there are lots of good Romantic era examples in the Wikimedia commons. Painters I particularly love are J.M.W. Turner, Ivan Aivazovsky, Caspar David Friedrich, Andreas Achenbach – there are dozens more sources of inspiration, an entire genre of Flemish maritime imagery for instance – that capture the ever-changing atmospherics, grandeur and sublimity of the deep ocean, the tempests and wondrous effects of light, shadow, wave and water. These are challenging subjects for real-time computer graphics, requiring not only some kind of water simulation, but particle and texture effects for sea-wrack, spume and spray, and volumetric effects and lighting for skyscapes, clouds, sun, moon and other astronomical bodies. I also re-watched Peter Weir’s epic Master and Commander: The Far Side of the World, especially the ‘making of’ documentaries that accompany it, which provide a wealth of detail about the technology, life and times.
I was aware that these were do-able as a roll-your-own in UE4.25, and there are lots of examples of, for example, Gerstner-wave Blueprints development online, but as I wasn’t confident in getting the entire environmental set-up done in time, I was looking for a more off-the-shelf solution. It was a bit of a dilemma, as UE4.26 was on the horizon but had not yet been released – and it promised many of these features. I compiled the GitHub 4.26 preview releases as they became available – but it looked like too much of a risk to rely on these as there was no guarantee of compatibility with other assets rated for 4.25. Given this, I evaluated some UE marketplace water/environment systems such as Oceanology (3.0) and Waterline PRO, and finally came across trueSKY™ – which provides a 30 day free demo. They’re all terrific systems with a range of areas in which they each excel, but as a computer researcher myself, trueSKY really appealed because it presents itself as a turnkey simulation environment that I could simply plug into UE4.25 – and it just worked. But it has some limitations – as I was to discover further down the track – but nothing really that would stop me from getting the job done.
TrueSky provides a lot of parametric control through its native UI and BP support, and made convincing water and atmospheric simulations accessible in UE4.25, based upon geophysically accurate simulation parameters. Like all these complex systems, some aspects are work-in-progress and not everything is documented terribly well (understandably, as it is a huge job to thoroughly document large parametric systems), though they provide some very useful online tutorials and I found support friendly and responsive. A key attraction was that they provide both an infinite ocean and sophisticated sky simulation, that can accurately simulate atmospheric/environmental effects for known lat/long positions on the Earth sphere – this meant it was relatively easy to set up a series of scenes matching the ostensible voyage track of the Perseverance from Port Jackson to Macquarie Island. Another crucial aspect was that it performed well on my machines of various specs – the home machine (Dell Aurora R7, 16GB RAM, 8GB GTX1070) and a studio machine (Dell Precision 5820, 64GB RAM, 8GB GTX1080 – later upgraded to an 8GB RTX2070 Super).
Whilst it is possible to manually or procedurally animate a ship on the ocean, in varying sea states, TrueSky also provides a tractable BP buoyancy component system, which simplified this objective no end, once I’d got everything set up properly. This required setting up buoyancy probes on the ship model I’d elected to use, in such a way that the physical simulation looked reasonably convincing, and attaching some some BP physical force vector components to drive it forward through the ocean. It involved a fair bit of experimentation to get the physics simulation working (a balance between the ship sinking to the bottom of the sea, tipping over or flying off into the air), centre-of-gravity offsets and understanding how UE collision detection works. Once that was all done, it became a stable and predictable system, freeing me up to concentrate on filming the ship at sea.