Some of the writing is very faint, and so rather than the usual technique of using 3D capture (close range laser scanning or photogrammetry), I opted to use Reflectance Transformation Imaging (RTI) to obtain our best chance of reading all of the letters.
The results were fantastic, and I was lucky enough to be perhaps the first to read some personal names from 1,300 years ago, as I processed the data from my home office.
There have been many calls via Twitter to see more of the enhanced images, so here they are (released under a Creative Commons attribution non-commercial license) with thanks to Professor Michelle Brown and Dr Oliver Padel who conducted the interpretation and final transcription.
Last year (2016) I was asked by Dr Andy Jones from Cornwall Archaeological Unit to record and study the surface of Hendraburnick Quoit on Bodmin Moor in Cornwall. The work, funded by Cornwall Archaeological Society, involved detailed 3D recording of the surface of the two stones that comprise the monument (which isn’t actually a quoit, more a ‘propped stone’). It was a complex task for which I used photogrammetry as the 3D scanning method. It involved careful planning and then taking many hundreds of high resolution photographs to guarantee coverage at an even resolution across all sides of both stones to enable careful study with millimetric levels of accuracy.
After processing the 3D data I began the lengthy task of analysing the surfaces in great detail. Each potential feature was subjected to four checks using alternative methods, including virtual RTI and cross sections. I burnt a lot of midnight oil ensuring finding ‘cup marks’ and tracing the grooved lines that connect many of them together. I found 105 cup marks and 47 possible grooved lines connected or radiating from them, following the slope of the stone. This suggests that the lines were made in-situ rather than before the stone was moved in prehistory to its current location. This makes Hendraburnick Quoit the most known decorated or deliberately marked stone in southern Britain – possibly topping even Stonehenge in number of human-made features (152) on its surface.
The result was this plan:
A very reduced resolution model of Hendraburnick can be viewed on Sketchfab, to aid understanding of the monument.
As the Colossus is in relatively shallow waters, it is a popular and accessible location for diving. A physical dive trail already exists; concrete stations around the site on the seabed with tethered and numbered buoys and an accompanying guide. To help divers plan a visit, and to provide non-divers with a sense of what the site is like, we have created an “interactive plan”. Utilising Sketchfab, visitors can explore a simplified model of the wreck via computers, tablets and smartphones, tapping the dive stations to explore different locations in more detail, as well as viewing “diver-eye” videos of the site itself.
We opted to create a simplified 3D plan based on the archaeological survey, clearly indicating in bold colours the different components of the wreck, rather than adopt a photorealistic approach. The wreck itself is a mass of seaweed and sand-covered dark wood which can be difficult for the layperson to interpret.
In August 2015 I was commissioned by National Maritime Museum Cornwall to produce a 3D printed replica of the late medieval bench end in the church of St Senara in Zennor. It depicts a mermaid, which is famously known as the Mermaid of Zennor. This has, along with another medieval bench end, been incorporated into a seat called the “Mermaid Chair”. The replica was for an exhibition entitled “Mermaids: Women at Sea” highlighting the achievements of female sailors. Naturally, people would expect to also see a mermaid – a legendary woman of the sea – as part of the display.
The concept was to produce a half-size replica, which would be mounted upon a large photograph of the church interior. The 3D print would be positioned exactly over the portion of the photo occupied by the original – a collage of 2D and 3D. Visitors would be encouraged to touch the exhibit – something that should not be encouraged with the original medieval frieze.
Before any work was undertaken, we sought permission from the churchwardens. They were keen to hear about the capture methods and learn more about the exhibition, and kindly gave us the go-ahead.
I used photogrammetry – a form of 3D capture that uses photographs – to capture the entire bench-end. Using natural light only a tripod was necessary to compensate for the long exposure times. Before commencing the capture I planned my visit with advice from the churchwarden, when there would be no direct sunlight pouring in through the stained glass. This is essential for correct colour capture, pretty as that may have looked.
There were many challenges with the capture in-situ in the church. Aside from technical challenges, St Senara is a well-visited church, and everyone wanted to see the mermaid. And to find out what I was doing. So it took a little longer than I had hoped, as it is always good to explain what’s going on to curious visitors, but the results turned out very well indeed.
The resulting model consisted of tens of millions of measurement points which were used to create a highly accurate and ‘watertight’ 3D model. Given that the rear of the mermaid bench-end would be fixed to a display panel, the model was altered to create a smooth plane on the reverse to aid mounting. After preparing the model for 3D printing (checking the model for errors that could confuse a 3D printer) the data was sent to ThinkSee3D for printing. The resulting replica would be half actual size (182mm x 380mm x 20mm) – full size (370mm x 755mm x 40mm) would have been prohibitively expensive and would need to be printed in two parts then glued together. It would be printed in full colour using a solid gypsum-based resin.
After some small test prints and experiments with colour, and a coat of varnish to replicate the finish of the real medieval wood, the finished 3D replica of the Mermaid of Zennor was ready. Museum staff mounted the print carefully over one of my photos – the result was rather effective!
Today sees the release of the January/February 2016 edition of British Archaeology. It features an article by Prof Michelle Brown and I on the early medieval cross-base at Gulval Church and its unusual depictions of the Four Evangelists.
Whilst my copy of the magazine has yet to be delivered the story has been picked up by Discovery News in a piece entitled “Medieval Monument Depicts Animal-Headed Saints“. Use the arrow buttons or square page buttons to navigate through the story.
Findings also suggest that people from the U.K. at this time were traveling to the Middle East, and vice versa, negating a prior belief that contact between this region and the U.K. was interrupted from about the end of the Roman Empire in the West in 476 A.D. to the start of the Medieval Crusades in 1095 A.D.
Below is the text from the newsletter which I wrote after an intensive couple of days research in the Society of Antiquaries library. I have many more thoughts and notes on the matter and will write a follow-up post.
A Reinterpretation of the Rock Art at Boscawen-ûn Stone Circle
In 1986 Ian Cooke first recorded, on the north-east side of the central stone at Boscawen-ûn stone circle, the presence of a pair of carvings interpreted as representations of neolithic stone axes. This remarkable discovery is relatively obscure and beyond the initial plans to record their location, and beyond Peter Herring’s excellent report in 2000, little research has been undertaken.
In early July 2015 the author was kindly taken to Boscawen-ûn by Adrian Rodda and shown the carvings. During this visit the central stone was recorded using photogrammetry, a technique which uses photographs to create a highly detailed 3D surface model.
Close analysis of the resulting 3D model of the central stone revealed the carvings clearly. The model was straightened so that they could be inspected more closely. Surface colour was removed and digital techniques were employed to accentuate any surface features. The results were surprising.
The carvings appear to represent not two stone axes but a pair of feet, soles outwards, carved in low relief. A row of ‘toes’ can be discerned, especially on the right-hand foot.
They bear a striking resemblance, albiet weathered and on coarse Lands End granite, to those recorded at Dolmen du Petit-Mont at Arzon in Brittany. Barbara Bender noted in 1986 that the stone bearing the feet motif had disappeared. Presumably it was removed or destroyed in WWII. However the author has discovered a photograph in Péquart & Le Rouzic’s 1927 Corpus des Signes Gravés des Monuments Mégalithiques du Morbihan, which is reproduced here. The feet are quite similar in appearance, albeit slightly smaller at 230mm (ours are about 450mm long).
This is not the only surprise at Boscawen-ûn. The data has revealed about 500mm above the feet a pair of circular features, also in low relief, which appear very similar to carvings interpreted as breasts on some allée-couvertes in Brittany (Tressé, Prajou-Menhir, etc).
Along with the presence of the possible cromlêh noted by Dr Borlase, perhaps we could consider that the stone circle was, for reasons unknown, constructed with reused stone from a much larger chambered tomb which incorporated decorated stones in the Breton style. Those symbols may have had potent significance, enough to position them in the centre of a new monument. Reuse of decorated and standing stones is known in Brittany (see Scarre 2011, p147). Many of the inner faces of the stones at Boscawen-ûn are flat – perhaps once lining a small passage or chamber.
More research is in progress and a more detailed article is in preparation.
Barbara Bender with Robert Caillaud. The Archaeology of Brittany, Normandy and the Channel Islands: An introduction and Guide. London: Faber and Faber, 1986.
Peter Herring. Boscawen, St Buryan, Cornwall: archaeological assessment. Truro: Cornwall County Council, 2000.
Martha et Saint-Just Péquart & Zacharie Le Rouzic. Corpus des Signes Gravés des Monuments Mégalithiques du Morbihan. Paris A. Picard & Berger-Levrault, 1927.
Chris Scarre. Landscapes of Neolithic Brittany. OUP. 2011.
I haven’t blogged since March, and since I have some data processing in the background and can do little else while I wait, I thought I’d post a little update.
I have been working for most of the week with Azook, rebuilding cornishmemory.com, the web front-end to their Re:collect digital archive of historic photographs, films and audio from across Cornwall. The updated site will launch in a few weeks with much larger images, better search, an API, and lots of great features. I have also been continuing to work on the MemoryFish project digitising historic photographs if Cornwall’s fishing industry. If that isn’t enough, I have helped Azook to tweak some of their workflows and effect some useful changes internally. I’m really enjoying being in at the deep end of digitisation and hope to continue to work with them on future projects.
When not working with Azook, I’m continuing to work with the Morrab Library. Having designed their digitisation facilities, which I believe are among the best in Cornwall, implemented their database and workflows, I’m also using the system myself to digitise some wonderful 19th century lantern slides. The Morrab Library have applied for some funding to allow us to digitise a unique collection of over a thousand glass plate negatives of archaeological subjects and Cornish cultural events such as the first Gorsedd in 1928. If it works out, these will all be made available online.
Heading back to more ‘direct’ archaeology, I have been busy here too. In the middle of July I visited Professor Charles Thomas FSA with Jacky Nowakowski FSA where we had a chance to visit the church of St Clement near Truro. We recorded the medieval cross in the churchyard using photogrammetry to further investigate the inscriptions. Interpretation is ongoing.
I have also been up to London to use the Society of Antiquaries library to research in interesting find I have made at a stone circle here in west Cornwall. More on that when I have finished the research, but it’s quite an exciting find.
As I type, I am processing both photogrammetry and Reflectance Transformation Imagine (RTI) data for a privately commissioned analysis of a stone thought to contain fine carvings. I am not convinced, but it important to assess the stone carefully. At 50mm x 40mm I have 13 million measurement points and a very nice mesh.
I have also undertaken a commission for the National Maritime Museum Cornwall (our largest and busiest museum) to 3D capture the famous medieval mermaid bench-end at the church of St Senara in Zennor. This was a tricky task because of low light levels, high specularity from the church windows on its highly polished wood surface, and many eager visitors. Despite the challenges, the results are fantastic. I provided the model in a format ready for 3D printing, which will be reproduced half size full colour out of solid resin for a forthcoming exhibition.
I’m also preparing an article for publication with Professor Michelle Brown FSA on the Gulval Evangelists to explore the significance of the find more fully.
And then there’s the Tywardreath Priory project, where we hope to launch a public archaeology project to locate and fully investigate a Benedictine priory in the village. I helped out with the open day during the Festival of Archaeology 2015, explaining 3D technologies, LiDAR, and the wider landscape archaeology.
Publishing 3D models online used to be a pain. It always relied upon plugins, and the results could never be very detailed. Sketchfab has changed all of that, and made publishing beautifully detailed 3D models a breeze. You could use the analogy that Sketchfab is a kind of YouTube for 3D models. It is also a social network in that favourites, comments and forum discussions are all part of the system.
Here’s a taster of Penzance Market Cross with a detail enhancement filter applied to it (think inverted chalk rubbing). You may need a fast broadband connection and fairly recent computer to have a smooth viewing experience of this model.
Since November I have been working on 3D capture data from St Piran’s Oratory. It has taken a long time to process the 3D photogrammetry data as it is a hugely complex task to record every stone. This process caused the untimely demise of my 5 year old MacBook Pro, which was struggling anyway. It’s 8GB RAM just couldn’t cope, with the sheer quantities of data, and then the screen began to flash purple, before randomly restarting. Or just turning off. Recording the remains of this old building finished it off. Hence the tabloid-esque title of this blog post…
Happily, colleagues at Archaeovision stepped in to help (thanks James), and I have been able to deliver the (300 million points) raw data to the client, along with a textured mesh of the Oratory and surrounding Scheduled concrete, and a series of orthographic renders of each elevation.
I have since been able to purchase a new computer – an iMac Retina 5K with i7 4Ghz, 32GB RAM, SSD, and upgraded graphics. The screen is gorgeous – so crisp – it makes going back to using a ‘normal’ screen quite difficult. It’s the perfect balance for me in terms of display quality and performance.
The header image of this post shows the last stage of this project. I am rendering a short animation of the chapel structure to help people visualise it without its concrete shelter.
Since the concrete walls surrounding St Piran’s Oratory do not allow enough clearance for conventional photography, this has proven to be a very useful exercise to record and understand the structure that we see today – a confection of medieval to 20th century rebuilding and repair.
When the animation is available publicly, I will post a link here. I also hope to upload a model of the unencumbered Oratory to Sketchfab so that it can be inspected at will. I will post that on this site in a new blog post when it is ready.