Making 3D digital archives available – the Archaeoptics Archive

It’s completely fair to state that Archaeoptics, a Glasgow-based company established in 2000, were true pioneers of 3D data capture and processing in the world of heritage. Ahead of their time, they travelled around the UK and abroad visiting archaeologists, museums and artists, 3D scanning tiny objects up to large buildings. It really was ‘scanning’ in the true sense of the definition, using laser-based triangulating or time-of-flight 3D laser scanners. The company hung up its scanners in 2006 and “went into hibernation“, but their spirit – and data – lived on.

Archaeoptics didn’t just scan objects. They created their own data processing and archiving pipelines, as well as developing software called Demon3D for clients to do things with their 3D scans. Back in the early part of the millennium, there just wasn’t the software that museums and archaeologists could use. The software that was available, such as 3ds Max, was expensive and struggled to open the dense datasets that scanning created. Demon allowed big datasets to be opened on commodity computers, measurements and cross-sections to be made, and for surface features to be enhanced. Those of us in heritage were interested in what 3D scanning could do to help us understand and analyse the surfaces we were capturing – to help us read eroded inscriptions, study tool marks to help us understand how something was made, and understand worn decorations (to mention but a few). We wanted 3D scan data to be part of the archaeological or interpretative process, not just a presentational tool. The combination of high density meshed 3D models and Demon allowed us to do that – it was all genuinely way ahead of its time.

I was lucky to work with the small team at Archaeoptics on occasions from 2002, notably on some projects at Stonehenge, Knowlton Henge and some archaeological finds when I was at Wessex Archaeology. It’s fair to say that I learned the foundations of 3D scanning from them. Not just how to scan, but how to be critical of the data, how to enhance it, and how to archive it. Many of those concepts I still use today, and I’m deeply grateful to them.

So I was heartened when I heard that Alistair Carty from Archaeoptics was planning to make their digital archive available online. This doesn’t mean that all of the data is available (clients owned the data they paid to create), but you can see what Archaeoptics have scanned, who the client was, and submit a form to request the data. Requests are reviewed by the Archaeoptics and the IP owner, and if possible, the data will be supplied as a download.

Most organisations at the time had no capability for managing large volumes of 3D data and most of our shipped data has languished in boxes and cupboards or, quite often, been completely lost.

The knock-on effect is also that the information about what we actually scanned has been unavailable as well. The Archaeoptics Archive is designed to help rectify this.

Archaeoptics Archive / https://archive.archaeopticscloud.com/data

The archive, with its refreshingly clean design, allows 3D datasets to be explored in several ways. By ‘object’ (individual objects scanned) or by ‘assemblage’ (essentially a group of objects or an individual job). Objects are also displayed spatially by a map.

Each object and assemblage also has a ‘permalink’. This is a unique and human-readable link, such as https://archive.archaeopticscloud.com/permalink/assemblages/olympics that will not change and can be used in citations. Too many heritage websites don’t pay attention to permalinks (and making them permanent) and change their links with redesigns, breaking citations, or having dynamic links that can’t be cited at all. Top tip: if you find a broken link for any website, try using the Wayback Machine to see if it has been archived.

When their Minolta 3D scanner was used, objects had to be planned and scanned carefully from different angles to capture all details. A single 3D object could be made from many combined (or ‘registered’) scans. Archaeoptics have even made these individual raw scan frames available, alongside reference photos. These are the source files collected directly by the scanner and can be re-processed using new methods if desired.

For anyone interested in 3D scanning in the heritage sectors, this is the real deal; raw and processed data, data as-collected, available in their original formats and widely used ones.

A digital image of a large 18th century carving depicting a person holding a torch aloft. This is the surviving part of a carving attached to the rear (stern) of a late 18th century ship.
Reprocessed and painted data of the late 18th century stern carving from HMS Colossus 3D scanned by Archaeoptics in 2006, refreshed and reused in 2021 by us.

Some of this data, at the time of writing, is over 22 years old. It’s been preserved, documented, and now presented online. Some of the ‘old’ Archaeoptics data has been used recently for some new projects, such as in our 2021 “Painting the Past” project, added to new research projects, and even used to make jewellery, in 2022.

In this tweet, Alistair makes an essential point. He tweeted “Interestingly, if I hadn’t maintained the archive, at least 50% of the datasets would have been lost entirely‚Ķ”. Even today, in the 2020s, many organisations using 3D scanning don’t know what to do with the data. Many don’t even think about long term data storage at all. Not much has changed. This is why I also keep an archive of all our digital projects, maintaining a standard directory structure with multiple copies, in original and common file formats and have contributed guidance to GLAM3D.

As 3D capture (in general, including photogrammetry/SfM) becomes more commonplace the focus doesn’t seem to be on capturing great amounts of surface detail. To many organisations, it’s about presentation, and sharing a nice-looking object online. It’s all about high definition textures, with details preserved via ‘normal’ and ambient occlusion texture maps. Popular photogrammetry pipelines like Apple’s ObjectCapture API don’t even create a dense mesh (highly detailed surface geometry) that the user can access. It’s troubling to see tweets from people who use systems such as this think that they’re ‘archiving’ objects. They aren’t if the details needed to study the object in the future simply haven’t been captured. If using photogrammetry, it’s still better to use professional software such as Agisoft Metashape or RealityCapture and a good camera, and focus on dense surface geometry to capture details that can foster study and learning.

This is why seeing the Archaeoptics Archive is heartening. From the beginning they captured data at a resolution appropriate to the size of the object; sub-millimetre for smaller objects or ones with faint features (such as rock art), and centimetric for large buildings (such as a church or earthwork). Their quality control processes made sure that the data presented to the client was good, taking into account aspects such as point density, angle of incidence, noise, etc.

And guess what? It’s stood the test of time. Now we can all learn from it.

Head over the Archaeoptics Archive and explore it for yourself.