Team DIU (well, half of it!) have been visiting the Rijksmuseum again for the biennial conference on 2 and 3D photography. 2 full days of speakers followed by another workshop day left us with lots to think about. This year’s conference built on the last, Robert Erdmann released the open source code for his amazing curtain viewer which can be tried out in the Bosh Project here http://boschproject.org/#/ . Malcolm is going to delve deeper in to Erdmann’s latest developments below. Otherwise 3D technology seems to be taking root, with debate over the level of quality and detail needed, and advances such as ‘videogrammetry’ and ‘unstructured light field renderings’ (see below) entering the fray.
As with the last conference, emphasis was put on the need for heritage institutions to adopt consistent standards (Metamorfoze and FADGI are now being implemented more widely, and the over-arching ISO/TS 19264-1:2017 has now been published) and equipment suppliers were encouraged to adapt to our requirements.
Additionally, the Rijksmuseum has published a thorough new object photography guide here https://www.rijksmuseum.nl/en/2d3d/rijksmuseum-manual-for-the-photography-of-3d-objects . This outlines the practice and standards at the Rijksmuseum with in-depth information on approaches taken for different collection types. Furthermore, Froukje van der Meulen gave a fascinating insight into the team co-ordination required for the Rijksmuseum’s mass digitisation project, encouraging us all to tackle the ‘Cabinet of doom’ (that we all have…) early!
One of my favourite presentations was from Micah Walter who discussed the Cooper Hewitt Pen. Visitors to the Design museum are issued with a digital pen which has a number of features: it acts as a ticket; allows visitors to digitally ‘collect’ museum objects; use in-Gallery apps and interactive elements; and to take home a permanent record- a link custom built with the visitor’s points of interest allowing later close study of the design collection. This innovative use of technology also provides data to the museum about statistics, peak traffic areas and times, favourite items in the collections and post visit link hits. Developed using APIs to allow later bolt on’s, and the project was assumed to have a 5 year lifespan. More information can be found here https://www.cooperhewitt.org/new-experience/designing-pen/
I found the workshops particularly useful. In ‘Unstructured Light Field Rendering: Enhanced Photogrammetry Using On-Camera Flash’, Dan Dennehy and Gary Meyer discussed software under development which allows the 3D capture of shiny objects using a simple on camera flash set up- this flies in the face of current photogrammetry which struggles with shiny or glossy surfaces and usually requires very diffuse lighting. What is more the results demonstrated were astounding- a bronze sculpture really had the surface texture and reflectance properties of a bronze. Unfortunately the software is not yet available commercially- it may become part of other programmes such as Agisoft.
My final workshop ‘Photographic Techniques for Transparent Glass Objects with Andrew Fortune and Frans Pegt was excellent. Andrew and Frans showed several different studio set-ups and how very subtle changes to the lighting could have a dramatic impact on the quality of the image, bringing up important details and defining edges. I came away with a new set of tools for the next time the Anatomy department deliver a tricky specimen in a jar!
I’ll hand over now to Malcolm, who will discuss his conference highlights.
Susan Pettigrew, Photographer
The first presentation of note for me was “A CRITICAL REFLECTION ON THE USE OF 3D TECHNOLOGIES FOR THE DOCUMENTATION OF ENDANGERED HERITAGE” by Hendrik Hameeuw (Research fellow, University of Leuven, Belgium)
Hendrik Hameeuw argued that the 3D models we are currently capturing are not detailed enough to be considered as a preservation standard. Suggesting “3D is Dangerous” he went on to demonstrate how inaccurate these models are in comparison to the high end 2D cameras currently available. The captures are missing a lot of surface detail and low level relief can become obscured and inaccurate. The processing power required to capture high end 3D makes it unviable for all but a few organisations. He makes an important point if we are to embrace 3D technologies as a preservation archive tool. It can be very easy to get swept away with the ‘wow! factor’ of new technologies but just how useful are the models we are currently capturing? In contrast to Hendrik’s presentation Robert Erdmann’s “PUSHING THE BOUNDARIES OF IMAGE PROCESSING AND VISUALIZATION FOR CULTURAL HERITAGE” (Senior Scientist, Rijksmuseum, Amsterdam, The Netherlands) displayed the pinnacle of what is currently achievable using medium format photogrammetry. Erdmann says “Objects from the Rijksmuseum and the Bosch Project provide strong motivation for the new techniques, including the recently acquired Bacchant sculpture by Adriaen de Vries, the Hartogh Plate, and many works by Bosch, Rembrandt, Hercules Segers, and Vermeer”. Erdmann tasked the photographers at the Rijksmuseum to use their medium format cameras to capture high end photogrammetry models of new acquisition Bacchant sculpture by Adrian de Vries. In this video below he demonstrates the photogrammetry model rotating.
Erdman then went on to demonstrate “The Mother of all Contact Sheets”. He demonstrated a live online interaction over wifi with a file that was 171 Gigapixels in size of every image digitised so far by the Rijksmuseum. The images were sorted by neural network organised by semantic content and “no metadata was harmed in making the image”, indeed no metadata was even used. Erdman’s mantra in reference to the code he used was “follow the source Luke”.
The last presentation of note was “THE MAKING OF AN ONLINE EXHIBITION: THE LEGACY OF ANCIENT PALMYRA” by Liz McDermott and Frances Terpak, Getty Research Institute, Los Angeles, USA”
A fantastic insight in building an online exhibition from Getty Research Institute. The level of pre-planning that went in to this was astonishing. Notably everything was drafted using paper and boards in the first instance because it was quicker than getting a developer to work on it from scratch. The project ran to 24,000 words and 108 images. For the launch they developed a digital strategy which included Facebook and online review policy. I would urge our own developers to look at what they have achieved with online delivery, it’s quite remarkably well thought out and flows like any good story. I think this work and the other websites listed below highlight the need for developers, designers and story tellers to work in conjunction to deliver these complex sites. Its almost akin to making a movie with distinct skill sets needed for each part.
These three websites below were described at conference as the current bench mark for image driven cultural heritage web sites.
Malcolm Brown, Deputy Photographer.