Technology

Facebook Twitter

Terms like high-resolution documentation and giga-pixel resolution are often used without a clear understanding of what is meant by this terminology in practice. In 3D scanning, high-resolution can mean long-range area-scanning at a resolution of 100 measured points per square meter or can be used to describe close-range surface scanning with 100 million measured points per square meter. In colour recording, confusion still exists between images that appear high resolution on screen and images that are in fact high resolution when printed at full size.

Lucida Laser Scanner developed with the support of Factum Foundation.
Lucida 3D Scanner has been created and developed by the artist Manuel Franquelo, cofounder of Factum Arte and it has been built and tested in Factum Arte under his supervision.

The association of digital with virtual is still at the root of many of these misunderstandings. When physical and digital are used in the same sentence the confusion is increased. The Q&A section on these pages is an attempt to open-up some of these issues and to explain why Factum Foundation is so committed to the work it is doing.

The biography of any cultural object will reveal a vast history, a hidden curriculum vitae that involves not just age, pollution, climatic changes, damage from movement and cleaning but also various levels of human intervention at different times in its life. Many of these interventions have been done to preserve the object. They are done with the best of intentions but the results often have unpredictable consequences. What the Foundation, through technology and skill, is able to do is record and archive the things we have inherited. The data can then be used for monitoring, study, dissemination and other valuable purposes such as replication and experimental re-creation. The concept of digital technology in conservation is the key on these pages - the results can be seen under Projects and at other locations on this site.

The Foundation turns to Factum Arte and other professionals in this rapidly growing field to execute the detailed and skilled work relating to the development and application of recording technologies. Factum Arte is an example of the new kind of working structures that are needed; it consists of a team of artists, technicians, engineers, software-writers, architects, conservators and digital/physical artisans dedicated to digital recording and mediation as part of a coherent approach to non-contact conservation.

Bespoke equipment, software and technologies are being designed to obtain optimum recording results. These non-contact methodologies are starting to have a huge impact on the world of conservation and are altering the way facsimiles are perceived by museums and experienced by the public. They are also helping to establish guidelines for conservation and documentation procedures. Factum Foundation and Factum Arte have worked and continue to work with many of the world's leading museums including; The Musée du Louvre, The British Museum, The Pergamon, Museo del Prado, Harvard Art Museum and Biblioteca Nacional in Madrid. Over the last decade they have also worked closely with the Fondazione Giorgio Cini, The Museo della Citta Bologna, the Supreme Council of Antiquities as well as many other international institutions.

 

Piers Wardle manipulating the scanner specifically designed to record Veronese's great painting The wedding at Cana at Musee du Louvre, Paris.

Input: Recording

One of Factum Arte’s core strengths is their interest in the qualities that make artefacts specifically what they are. This has led to the development in conjunction with the Foundation of a number of processes and the equipment necessary for carrying them out. The technology is constantly advancing and this allows for greater and greater quality and detail in the recording process - both of colour and of surface.

Q. Why does the Foundation work with technology?
A. The idea behind the Foundation has grown out of many years of working in high-resolution documentation and the realization that a lack of communication and understanding is hampering the application of technology in conservation.

The technology is now available to record the surface of objects in high-resolution. This recording can then be used to understand the object in intimate ways, to understand why it looks as it does, to monitor its condition and/or to create a facsimile when required. Facsimile is a misunderstood word that until recently often evoked an emotional and negative response – usually as a result of the poor quality of the object. The Foundation views an exact facsimile as evidence of the accuracy of the various types of recording that were used in its fabrication. The many hours spent recording and studying the object reveal unexpected insights - in every facsimile that the Foundation has commissioned from Factum Arte the object re-appears as a fascinating and complex subject. The creation of facsimiles has a number of applications, such as the possibility of being viewed by tourists in the correct environment whilst conserving the original that may suffer damage otherwise; it also allows artefacts to return to their right location or to be shown when the original cannot be moved - there are myriad uses that technology allows us to achieve (see below). However, the possibilities do not end here: it is the relationship between the digital archive and the physical object that is most important. Whatever happens to the object the archive is a record for future generations whose technology may allow even greater use of the data if it is stored in raw form that can be reprocessed.

Q. Why does the Foundation promote digital conservation?
A. The Foundation wants to record forensically and accurately objects as they are. It is dedicated to finding innovative ways to study and analyse the information. It is essential that we understand any physical interventions that may have been imposed on the object - the aim is to generate more accurate information so that we can ‘see’ the object better. Technology does not replace interpretive skill, it informs it and enriches our understanding. This is essential to the preservation of our cultural heritage, given that future generations will have a more advanced understanding of digital analysis and material conservation. But they will not have, unless we safeguard the evidence, access to what we had - our inheritance as we received it.

Viewing the physically layered colour and overlaid relief data on a rendering of Rubens' Triumph of the Eucharist over Idolatry.

Q. What are the most effective technologies involved in recording cultural objects?
A. Computer vision and sensor technology provide enormous benefits and uses for the recording of objects. White light and laser scanning, video, still and time-lapse photography, photogrammetry X-ray and infra-red recording are techniques that are used or commissioned by the Foundation.

Below is a brief description of the technologies frequently used in recording cultural objects.

Close Range scanning – Laser scanners emit controlled laser radiation and detect the reflection of the emitted signal to create an image of the recorded object. The public perception is that these scanners are very expensive, a factor that depends largely on levels of accuracy; Branded scanners, which are designed for industrial processes and not for high resolution conservation, can be expensive, as their makers need to sell them and create demand. That means that not just development but also marketing and future risk are aspects that need to be costed in, as well as profit. Along with the recording device, software is needed to interpret and manage the recorded data - this can often soar the price by an additional fifty percent. In addition, software is not statically priced - it needs upgrading to be compatible - so yes, it is generally true that scanners can be expensive.

However the Manuel Franquelo's Lucida 3D Scanner reverses this trend. It is a high resolution low cost laser scanner with its own integrated software that records relief information as raw black and white video. Factum Arte and the Factum Foundation have been working under the direction of Manuel Franquelo for 2 years to develop the Lucida 3D laser scanner (see video and further detail below) using readily available components.

Manuel Franquelo's Lucida 3D Scanner, developed with the support of Factum Foundation

The Foundation will be releasing very soon the construction and operational details as well as the component specifications for the Lucida Scanner Version 3 in order that the widest public has access to this body of knowledge.

Long-range and middle distance scanning.
For large installations and objects the use of Time of Flight (ToF) middle-distance scanners is most appropriate. These scanners measure the time taken for a signal to bounce back from the target. With ToF scanners the user chooses a series of points defined as important on his target and records these; it is a selective system that creates an impression of the target. Middle-distance time of flight scanning is used when it comes to obtaining an impression of sites and buildings. The specifications of ToF scanners allow them to scan approximately 10,000 to one million points per second up to a distance of 200 meters, with an accuracy of 5 millimeters; this operation is then followed by a significant amount of post-processing, required as the data needs to be manipulated with complex and expensive software.

Photography and Photogrammetry (also see a full description of parallel and panoramic photographic processes in later questions). The critical development that is taking place in the photography of paintings and objects is the ability to stitch high resolution images together to form composite images of extraordinary detail. When this is merged with 3D and multi-spectral images the interpretive potential is vast.

Q. What is the difference between white light and laser scanning?
A. The development of White Light scanning has focused on the capture of form (the recording of volumetric form). It also records the surface of these forms but the noise/information ration generally means that significant processing is required to extract the data from the noise - therefore reducing the correspondence between the data and the object. The field is developing rapidly and advances and innovations are constantly being made. Hand held laser systems mounted onto 7 axis arms can also be used to record volumetric objects but suffer from a similar loss of specific detail. To optimize the correspondence between the information and the surface the movement of the recording device must be as constant as possible. While a laser scanner mounted onto linear guides will record more accurate data it is limited to the recording of flat or almost flat surfaces. There is usually a direct relationship between speed of scanning and the correspondence between the data and the surface. There is a great deal of misunderstanding about resolution and accuracy.

Process of white light scanning and how the contours of an object are captured.

White light (structured light) scanning systems record 3D data by analyzing the deformation of parallel lines of white light projected onto the target. As they pass over the surface the deviation to a theoretically straight line reveals the shape. The patterned light is seen from one or two cameras positioned just off the axis of projection and at a fixed angle from it. The principles of this system were originally developed in the 1920’s in France by Claudius Givaudan.

Laser scanning projects a thin laser strip on the surface of the target. As the line moves over the surface it is recorded by one - or two - cameras. The theoretical centre of the line is calculated mathematically eliminating the speckle noise (see video). The movement of the scanner can be controlled by linear guides or by hand.

All optical scanning systems are affected by the glossiness of the surface and by contrast. A great deal of research is being devoted to systems that can scan black and white, matt and gloss surfaces at the same time.

Q. What is the Lucida scanner?
A. The Lucida 3D Laser Scanner has been created and developed by the artist Manuel Franquelo, cofounder of Factum Arte. It has been developed whilst at the same time used in many locations to create contact-free high-resolution images of the surface of paintings and low relief. It is a two camera one laser scanner with bi-directional recording. It has a software application that under and over exposes each frame to record accurate data from both the dark and light areas of the surface. The scanning head is small and robust and mounted onto x and y axis linear guides fitted to a light-weight rigid aluminium structure. All the scanning processes are controlled from a portable computer and the bespoke software application allows the user to select a specific section of the target should re-scanning be necessary. The Z axis is manually controlled. All stitching is done as a depth-map greatly reducing the size of the files and increasing the accuracy of the operation. The graphic user interface is intuitive and easy to operate.

Development and importance of the Manuel Franquelo's Lucida scanner in the Foundation's conservation work.

As with the previous version, the system is lightweight, powered by battery or mains electricity and stores the 3D information as raw video data. The new (version 3) Lucida Scanner is being used to record the surface of works of art in a number of different projects, all of which have significant implications for the protection and management of cultural heritage.

 

The Lucida scanner head in progress.

Lucida is the Foundation's preferred scanner for documenting highly detailed work that requires a close correspondence between the surface of the object and the data that is recorded from the surface. At present the Lucida uses a red-light laser but this will soon be replaced with a super-luminescent diode that will significantly reduce the speckle noise and increase the depth of field. The Lucida records tonal data, at present this is being processed at 100 million measured points per square meter.

3d Scanning Test, Manuel Franquelo's Lucida Scanner, March 2013.

This approach to the management of 3D information has significant implications for the management of the data and for its use in 3D printing and routing. The surface of the Hereford Mappa Mundi, (detailed in its corresponding Project page) can now be routed in one piece - using existing technology this was impossible due to the size of the files.

Scanning Hereford Mappa Mundi.

Q. Who uses or potentially will use Lucida scanners?
A. The first and second versions of the Lucida 3D Scanner, created and developed by Manuel Franquelo, were used by Factum Arte to carry out the work to date - these scanners were prototypes that have meant that the final Lucida is a fully effective and tested piece of technology. Its operational procedures are simple and intuitive which means it is easy to use. With one day training most people with basic computer skills will be able to obtain good data. The Foundations aim is to deliver the system free of charge to museums and other custodians of cultural heritage. It is the Foundation's intention that up to 100 Lucida Scanners will be deployed in various places around the world.

Q. What is the cost of the scanners - I see in the press that they are very expensive and need endless software and upgrades too?
A. These Lucida scanners will generally be provided to the institutions or custodians free of charge, and training will be given to local operators so that scanning can become a normal and continuous procedure. In the case of this system there will be no software or upgrade costs. The software has been developed in-house at Factum Arte so that cost is minimized and efficacy optimized. When upgrades are required they will be available free of charge.

The initiative of the Foundation is to provide museums and institutions with a free scanner, complete with the software updates that often make most market scanners impractical for medium and long-term use. The tension between cost and utility has slowed the application of technology for recording and monitoring purposes. The aim of the Foundation is to reverse this trend.

Q. Who will own the data? Where will it be stored?
A. One of the central ideas behind the Foundation’s approach is to ensure all the data that is recorded is safely stored. The simple answer to the question is that the custodian of the object should own and have 100% control of the data and 100% of any revenue that the data generates. One way of ensuring safe storage and optimum data quality is to process it centrally, store it safely and supply it to the owner in a form they can use.

No rights change hands. The data should be made available free of charge for monitoring, research and study but the commercial rights will remain with the custodian. The Foundation wants to work hand in hand with museums and institutions to make sure that the data is recorded, stored, preserved and disseminated. It is also dedicated to finding new uses for the data that can generate more revenue for the appreciation and preservation of our heritage.

 

Manuel Franquelo, the artist and engineer who is behind the conception of the Lucida 3D Scanner.

Q. Why not record in a studio?

Photography

A. Photographic recording (or scanning) of works of art at high-resolution is not as easy as it might seem. Sometimes it can be studio based - sometimes it has to be done in-situ.

In a photographic studio heavy immobile equipment can be used to create stable, high-quality recordings. The subject is brought to the equipment in order to be photographed or scanned; Most of the ‘top-end’ equipment has been manufactured in Germany and Japan where there has been significant investment in CCD technology (charge-coupled device). CCD technology is the digital equivalent of film.

A tri-linear CCD has 3 lines of luminance photosites or cavity arrays where photons collect upon exposure to light, rather like a solitaire board with indents that can take more than one ball - digital cameras have millions of them. Each of them is filtered for primary red, green and blue (RGB) colours (normally a Bayer Array), which move across the target recording the colour of the surface, which must be illuminated with the correct level and type of light. The length of the linear CCD determines the maximum width of the target to be scanned, the length being determined by the length of the rails onto which the CCD is mounted. These scanners can achieve very high quality results but are ultimately limited in their resolution - none can really record any large surface with a resolution above 300 dots per (lineal) inch or dpi, a standard measure of spatial density. This is due to the limitations of the top-end tri-linear CCDs that have a limited length and number of photosites. The major issue is issue is that these machines are immense and so cannot be moved to a site.

Q. What is parallel recording?
A. One solution is parallel recording: Factum Arte worked from its early days developing machines that were lightweight, modular and easily transportable. Systems that could scan and record color in difficult locations and environments - most of the scanning targets were either fixed or too large to move. Most were also highly fragile.

Parallel recording machines were first developed using modified flatbed scanners (similar to desktop scanning machines), digital single lens reflex (DSLR) cameras with macro lenses and medium format cameras with digital backs (a back is the removable part of a sophisticated camera that can be changed to allow for a variety of operations but using the same expensive lens housing), attached to linear guides, either manual or motorized and automated. The principle is simple - set up a railed platform on the ground, parallel to the surface of the painting (or 2D target). A vertical mast or frame is then moved along the rails, parallel to the surface of the painting. The recording device and lighting are fixed to the mast or frame. As the platform moves parallel to the surface of the target a number of image captures are made at fixed positions, each overlapping with the previous and the next, to cover the whole of the surface. All these image captures are then assembled and digitally stitched together to produce a single large image file, at a resolution of up to 1200 dpi, depending on the recording system used.

This kind of system can only record a flat surface with minimum relief and deformations over the surface, and in these conditions is not limited by the size of the surface to scan. While portable, a system like this takes time to install and operate and is limited by the time needed to record the surface, in small increments.

It was this kind of system that was used in several highly delicate and important recordings: in the Musée du Louvre for Veronese's Wedding at Cana (for which the scanner became known as the Cana Scanner) and other museums, churches and in the tomb of Tuthankamun.

Q. Are there problems with recording with parallel capture systems?
A. The most obvious weakness of parallel recording appears when the target has a non-flat surface. The other is the time taken for installation and operation while keeping the very high recording quality and resolution. Time is always the determining factor - Museums and sites are normally open to the public and keep works on public display is an important part of their mission. Parallel recording systems tend to be slower than panoramic recording systems.

Q. What is a panoramic recording System?
A. If the camera is fixed in a single position panoramic technology can be used to calculate the position and angel of a camera mounted onto a 2 axis cradle. The Foundation and Factum Arte work with a Dr. Clauss system that has several modifications including the projection of a laser pattern to assist the focusing. The recording device (a professional digital single lens reflex camera (DSLR) with long telephoto lens of 600mm or even more) is mounted onto a motorised panoramic head which rotates around the perspective centre or nodal point. The surface is then scanned and recorded by rotating the head and capturing still images each overlapping with the adjoining images. The lighting is by flash (or fixed constant source on the whole surface in the case of recording a relatively small surface) which is mounted on a second motorised slave head to the recording unit so that it follows and illuminates only the section being captured for each position. If the flash is installed on the same head as the recording system it results in unusual shadows and hotspots - the separation of the light-source from the recording system is essential to achieve the desired level of control.

Panoramic photography - Giovanni da Modena in San Petronio.

Systems similar to this have been in use for some time but the developments in the system, in CCD technology have made it more useful than ever. Developments in stitching software are also playing an increasingly important role. As with parallel recording, the final image is produced from assembling all these image captures and the result can be in similar levels of resolution, depending on the size of the surface to scan and the camera/lens combination used.

This kind of system has been used to record large areas at high-resolution when access is limited. In the Sala Bologna (Vatican) the map of Bologna by Lorenzo Sabatini, the largest city-map of the renaissance measuring approximately 50 square meters, was recorded at a resolution of 400 dpi in under 7 hours, with 3,000 single captures.

Q. What are the differences that characterise parallel and panoramic systems?
A. Apart from the speed of set-up and working the main advantage is that the Panoramic system can record uneven, complex and inaccessible places without extensive scaffolding. A good example of this might be a vaulted ceiling.

The main limitation is that with panoramic recording the final stitched image derives from a single point of view - the position of the panoramic head. Anything that cannot be seen from that point of view will not be recorded. Parallel scanning also has an advantage in cramped conditions and requited less space. For example panoramic recording with a 600mm lens requires a minimum focusing distance is just over 5 metres.

The process of "stitching": merging colour and 3D data for perfect registration.

Stitching is the name by which the post-processing of the images is commonly known; stitching all single image captures in a panoramic system is much faster, easier and more precise than with parallel recording as overlaps show identical data between adjoining captures as all are shot from a single point of view; the warping and re-projecting to make the final image are a result of mathematical calculations. Each of the images, being captured by the same device (camera and lens) from the same position, share common parameters (only yaw and tilt parameters are individual for each capture). Lens distortion and projector shift can also be accounted for mathematically. This means linkage can be optimised and a sub-pixel accuracy is possible even when working with thousands of single captures as the calculation is the same for each.

Q. Can you accurately record colour?
A. During the color recording extensive colour notes are made using a series of colour sticks made on-site and matched to specific points on the surface of the painting. These are fixed into a book containing a 1:1 scale line-drawing of the original. A bit of the colour stick is cut off and fixed into the book at the corresponding point on the painting. The aim of the colour adjustments is to match the colours of an image printed using Factum Arte's flatbed printer to the colour sticks recorded.

 

Colour is one of the least understood and most complex subjects - Adam Lowe talks about his approach.

Colour is one of the least understood and most complex subjects. In the production of a facsimile you are seldom dealing with a standard flat colour. Most coloured materials age in complex ways - among the most important traits of ageing are the changes in transparency revealing or obscuring the layered nature of the paint. Complex changes in texture result in an irregular surface complete with shadows and highlights and an uneven surface reflectivity.

Digital Storage

Q. Why store digital data?
A. The speed at which new developments are appearing show that imaging and recording technology are still far from fully developed. The Foundation is focused on storing the digital data as securely as possible in a format that can generate revenue for the custodian but be accessible for monitoring and study. Future generations will continue to develop new technologies – these will hopefully be able to extract more information from the raw data we are gathering now. Un-abstracted raw data is the key to preserving the past. Future generations will find better ways of processing, analyzing, viewing or applying the data. The aim is to make it available - a record of the past and present for the future.

Q. Can you accurately record colour?
A. During the color recording extensive colour notes are made using a series of colour sticks made on-site and matched to specific points on the surface of the painting. These are fixed into a book containing a 1:1 scale line-drawing of the original. A bit of the colour stick is cut off and fixed into the book at the corresponding point on the painting. The aim of the colour adjustments is to match the colours of an image printed using Factum Arte's flatbed printer to the colour sticks recorded.

Storing raw data. Adam Lowe explains why the data is stored in this state - pre-process and pre-manipulation.

Q. What is raw data?
A. Raw data is the product of the diverse recording processes (video, photography, X-ray, 3D scanning or other data capture) without any post-processing applied to it. The data produced is a mass of information- some if this is noise some of it useful; when the relationship between signal and noise has been ‘cleaned’ (tidied up) it corresponds to one aspect of the object (its surface, its colour, its structure). Mediation and transformation are already two words that are in regular usage in both Factum Arte and the Foundation. Abstraction and condensation are equally important when referring to raw data. Most recording systems for the commercial market automatically process the data to varying levels (a lot more will be written about this as the site develops) – removing noise, adding sharpness, cleaning the appearance of the image etc. These all impose levels of abstraction in which information is removed - lost forever. The image of condensation seems more relevant to the long-term preservation of information.

Multi-layer files - digital inputs from various source technologies.

Q. What happens to the data after it has processed and prepared?
A. The data collected is sent to Factum Foundation to be stored permanently in the Foundation's secure archive after it has been carefully codified for the future - future technologies may well develop some astonishing techniques that we can only roughly estimate and prepare for. The data belongs to and remains the property of the custodian of the object as subject - Factum Foundation’s role is to protect, preserve and make this data available for study, maintaining the digital information in a safe and secure archive for this and future generations. Factum Foundation does not want property rights - this is not the aim and the Foundation cannot understand a world where our heritage - not created by us but given to us - is willing to bargain over a common good.

 

Q. Who has access to the data?
A. The data recorded is made available to the public openly through the Creative Commons license on behalf of the custodian. This means anyone can see it and use it for research and analysis but anyone wishing to use the data for commercial purposes will have to go to the custodian who may license or allow use. Should a payment be necessary, this will go directly to the custodian, not Factum Foundation in any case. It is important to stress this point, as the data and rights of all data recorded in all sites globally will remain the property of the custodian at all times, who will benefit from all rights of use and any revenues created by the data. In this way our heritage is conserved, recorded and made available generally to the public ensuring that rights remain in the hands of their custodians.

Virtual Applications

Advances in digital technology are rapidly and radically changing our understanding of our shared cultural heritage. The conservation and preservation communities are realizing the importance of high-resolution digital recording and this data is starting to be integrated into professional protocols.

One field that is emerging is that of digital restoration- the in-depth study and application of restoration techniques in digital form before they are imposed on the original object. Any intervention is a reflection of its time and of the team responsible for the restoration. As recording technologies develop the subjectivity of this approach is revealed. The Foundation is encouraging an approach that will result in more and more emphasis on virtual restorations and a greater period of time for reflection. To some degree all restorations turn the original into a reproduction of itself – it is as essential that all changes are fully documented as that all actions are fully reversible.

Q. What is digital conservation?
A. Digital restoration refers to restoration projects carried out on the digital archives. An important advantage of digital technology is the potential to manipulate the data. It is possible to digitally restore the object both by working with specialists to modify the data and by scanning fragments that have been dispersed in museum and private collections worldwide. In this way the data can be merged to produce a digitally restored facsimile of a damaged or dispersed original. The precise protocol for this has yet to be established.

Q. What is a multi-layered archive and why is it important?
A. A layered archive is one that has been made up with a number of or many inputs, be they laser scanning, X ray, infra-red, photography and many more. The skill is to integrate these various inputs to create a single final file that contains the manipulated data from all the inputs such that the original object can be viewed in many ways but where the whole is there and visible. It is important to note that layered imaged depend greatly on skilled post-processing. For information on the techniques and processes for layered images, please visit the Layered Archives micro-site.


Lucida 3D Scanner
Created and developed by the artist Manuel Franquelo, cofounder of Factum Arte
Built and tested in Factum Arte under the supervision of Manuel Franquelo.

© Copyright 2018 Factum Foundation | Terms & Conditions

This website uses cookies to improve your experience online. By using our website, you are agreeing to our use of cookies. To find out more, read our cookies policy. Close