This lecture was presented at the 3D Digital Documentation Summit held July 10-12, 2012 at the Presidio, San Francisco, CA
Applications of Reflectance Transformation Imaging (RTI) in a Fine Arts Museum: Examination, Documentation, and Beyond
In 2006, the Worcester Art Museum collaborated with Cultural Heritage Imaging (CHI) to develop reflectance transformation imaging (RTI) as a viable examination and documentation method for fine art collections. This included designing an RTI dome and establishing a digital workflow protocol for acquiring, processing, and storing RTI files. As the first American art museum to integrate RTI into conservation practice, the Worcester/CHI collaborators presented their achievement at the 2009 Annual Meeting of the American Institute for Conservation held in Los Angeles. Since then, the Museum has documented the surface topologies of a wide variety of art materials, with a particular interest in finding new applications for RTI. This paper presents a brief sampling of examples of RTI use, with an emphasis on two unique examples in which the application of RTI went beyond the sole purpose of documentation to enhance significantly our understanding of the artistic process.
The first example involves a research project that investigated a group of Greek red-figure vases and vase fragments in the collection of the Harvard Art Museums and the Worcester Art Museum. The project, which is a continuation of research presented by Worcester at the 2010 MRS conference in Boston, utilized RTI in conjunction with laser scanning confocal microscopy to help resolve a long-standing debate among scholars regarding the technique used to create the characteristic ‘relief lines’ of the surface decoration. This research also involved mock-ups for comparison to better understand the ancient technology and ultimately led to the conclusion that the relief lines were not produced by an extruded method as previously postulated, but with a brush made of two or very few hairs, termed linierhaar as first proposed by Gérard Seiterle in 1976. Two distinct types of relief lines were observed: the laid line (proposed by Seiterle) which has a characteristic ridged profile, and the pulled line (proposed in a previous paper by the authors) which has a furrowed profile. Additionally, it was determined that the relief line used to outline figures was applied prior to the contour line. These findings are primarily based on surface topography evidence visible using RTI.
The second example involves the discovery of an original inscription on an early-career portrait by the Flemish Baroque master Anthony Van Dyck (1599-1641). The portrait, which belongs to the Royal Museum of Fine Arts in Antwerp, is of an unknown man and is on long-term loan to the Worcester Art Museum. Obvious compositional changes to the sitter’s collar and background prompted x-radiography and infrared reflectography examinations to evaluate the extent of changes throughout. Neither of these methods revealed the underlying inscription; however, during visual examination at Worcester’s conservation department, a faint feature was noticed in the paint surface topography that suggested the presence of the underlying inscription. The inscription likely was painted over at the same time modifications were made to the sitter’s costume. Subsequent application of RTI enabled conservators to image the entirety of the previously unknown inscription, which now provides scholars with a firm date for the artist’s original composition and an example of how RTI can yield results where more established imaging techniques such as x-radiography and infrared reflectography do not.
Church: Alright, next talk is “Applications of Reflectance Transformation Imaging, RTI, in a Fine Art Museum: Examination, Documentation, and Beyond” with Philip Klausmeyer. Philip earned an art history and a studio art degree from the University of Massachusetts Amherst, an M.S. from the Winterthur University of Delaware Program in Art Conservation, specializing in paintings. In addition to 15 years at the Worchester Art Museum, his conservation work includes time at the Strauss Center for Conservation, Harvard Museum of American Art, and Pennsylvania Academy of Fine Arts. He received his PhD from Clark University and is currently a Conservation Scientist and an Associate Paintings Conservator.
Klausmeyer: Thanks Jason, OK, so today, I am speaking not in the capacity of a conservation scientist but to add my voice to this growing chorus of end users and conservators about the value and use of RTI in our day to day practices and its importance in answering defined questions, thus the title today; “Applications of RTI in a Fine Art Museum: Examination, Documentation, and Beyond.”
A quick overview of my presentation includes an introduction and then it is divided into two parts, the first part is using RTI to study Greek attic pottery, and the second part is an application of RTI in an ongoing conservation treatment of an early work by Anthony van Dyck. The last three talks today are from three conservators from three different institutions, one of which is the Smithsonian, one of which is the Metropolitan Museum of Art, and then there is the Worcester Art Museum. So I expect of these three, you might need a little bit of an introduction to one of them.
So the Worcester Art Museum is about an hour drive west of Boston and it is a classic American art museum founded right around the turn of the last century. The museum collection is really encyclopedic in nature and although it doesn’t have the numbers of the Smithsonian by any means, it’s got well over 35,000 objects in its collection representing everything from painting, sculptures, decorative arts, photography, prints, drawings, and more, and ranging from ancient to modern. On the left you see the facade of the 1930’s building and on the right you see a view of our beloved Renaissance Court with its centerpiece, the famous Worcester Hunt Mosaic, which is the largest ancient mosaic in North America.
So the Worcester Art Museum has a history in conservation that is one that we can be proud of, and it is thanks in large part to these two gentlemen that you see here. The conservators in the room will probably recognize George Stout, who’s pictured on the left; they probably don’t know that he was the director of the Worcester Art Museum starting in 1947. He was very instrumental in bringing up the gentleman you see on the right. His name is Edmond de Beaumont and you see him working with some state of the art
X-ray equipment from the 1940’s. Edmund, I really want to give him props in this talk today because so much of what we are talking about is documentation. For all the opportunities I might have to differ in opinion in terms of the conservation treatments that took place during his tenure, Edmund was an incredible documenter. He left an incredible record of all the things that he did and the value to us today is really hard to underestimate.
Today, the conservation lab remains a vibrant part of the museum’s mission to preserve, collect, and exhibit works. There are two labs to the museum. Pictured here is the main lab that houses our objects, conservation center, and the paintings conservation. On the far left, you see the head of the department, Chief Conservator Rita Albertson, myself, and on the far right is Birgit Strähle, our assistant paintings conservator. Winifred Murray is shown at the microscopes standing. She is one of two Andrew Mellon fellows that we have. We have these rotating positions of fellowships and they are typically a three year position and they rotate between specialties of paintings, objects, and paper. In the distance you will see Paula Artal-Isbrand who is our objects conservator.
The museum has tried to remain true to its proud heritage of conservation and in 2004, we got the first of three grants from the Andrew W. Mellon Foundation and that enabled us to purchase some scientific instrumentation. That included an FTIR spectrometer, an XRF spectrometer, improvements to our existing polarized light microscope, infrared reflectography and what was the original line item was budgeted for a more high-powered X-ray tube. That was really the interest of Lawrence Becker, who some of you might know as the head of the objects conservation department at the Metropolitan. He was, at that time, head of our department at Worcester.
When Larry left for the Met, so did the interest for the high-powered X-ray tube because no one was going to be carrying out the project of looking at the bronze sculptures. So we deemed our current Lorad X-ray tube, which is a Lorad 200, as sufficient for our purposes. I lobbied heavily to re-appropriate that money into the budget for equipment to go towards starting up the museum in what was then kind of a fledgling technology of RTI and PTM. That took a lot of letters and reports to Angelica Rudenstine. Some of you may know her as a previous head of the Mellon’s Museum Department. Correspondence could be a very trying thing sometimes with Angelica, but I am glad to say she became a believer in this technology, and then I went out, and it was a bit of serendipity, but I went to a talk where Tom Malzbender was and just happened to find myself next to him at the lunch table. He put me in touch with the team at Chi, Carla and Mark and it wasn’t long before a fantastic collaboration started up. We designed, in 2007, the RTI fixed light array dome that you see here. By 2008 it was installed and we were up and running with it. And then by 2009, we gave the presentation listed below:
Reflectance Transformation Imaging: A New Conservation Tool for Examination and Documentation.
That was at the 37th Annual Meeting of AIC that was held in L.A. that year. So, since then, we’ve had time to integrate it into our own workflow and what I am thankful I don’t have to talk about today are the principles behind the method. But what I want to talk about is two case studies really where it became a really powerful tool in our investigations. The picture here is of Paula Artol-Isbrand, who is our objects conservator, and as the photograph suggests, she has a very keen interest in Greek attic pottery. Well, she had known about a long standing debate about the production of Greek attic pottery, and it launched our own investigations that ultimately led to a presentation at the 2010 MRS fall meeting, and the title of that you see here.
Thankfully, we’ve had the opportunity to expand on this research since 2010. Originally it was based on works from the Worcester Art Museum, but because of this research, the Harvard Art Museums had made accessible to us about 18 other Greek vase shards or vases to stack up against our findings of this original research. Now, some of you probably could give this talk a lot better than I can, but Greek attic pottery just a general overview, represents a period from 600 B.C.E to 330 B.C.E, and it is generally divided into two major areas; the black figure period and the red figure period. Both of these works are in the Worcester Art Museum’s collection, and we were particularly interested in looking at some features on the surface of the red figure vases so the Stamnos that you see on the right, from the Tyszkiewicz painter, which had recently been conserved and asked for loan by the Getty became really, the centerpiece of this original research.
So the production of this red figure ware and the black figure ware really does represent a mile stone in the history of ceramic production. It wasn’t until 1970’s or 1980’s when a scholar really rediscovered this technique of how these things were made, and it’s really a series of different steps in the firing process under three different conditions. The first condition is an oxidation condition, and the decorative paintwork that is applied to this surface, I should say from the beginning, has more of a silica content than the rest of the vessel itself. So in the oxidation stage, both the vessel and the surface decoration turn red, then the firing conditions are changed to a reduction condition and both of those surfaces turn black. But at that point, the temperature reached such a high point that the silica in the glaze material becomes vitrified and is no longer really reacting with oxygen. So then it goes under another condition of re-oxidation and the vessel turns back to red again and thus we get what we typically know as the red figure vessels. So between the red figure and the black figure vessels the technique I just described can be used for both of them, and the surface decoration itself, although varied and certainly beautiful in so many different ways, the materials and the techniques that they used were actually limited in number.
One technique in particular that we were interested in investigating is how did they achieve what’s called the relief lines? The relief lines are those thin black lines in the garments and to describe the collar and the features of the face; the eyes, the nose, and eyebrows, those are called relief lines. It has been a long standing debate about how were those are achieved. What method is used to achieve them? Scholars differed on what tools were used. So we wanted to go back and reopen the debate and see if we could add any information to what was known based on an investigation into the morphology of the lines.
The two techniques that we used, we figured were pretty complimentary. I already told you we were up and running with RTI within our lab, but we also had access via some great colleagues over at WPI to the use of a laser scanner confocal microscope. So, you’re getting pretty familiar with all the pros and cons of RTI today. Laser scanning confocal microscopy, I think, although more of a quantitative technique, whereas RTI is more of an image-based technique. Laser scanning, one of the big limitations of this project is just the set up. You can see we were limited to just shards. We are not going to put vessels on that stage and there is not really any opportunity with that particular instrument that we had access to, to modify the instrument to accommodate for larger objects. So working with shards was actually a pretty convenient use of the confocal microscope. Whereas RTI allowed us to look at whole vessels.
So we continued with that and you’ve seen some live RTI’s today. These are still captures. The one on the right is a still capture, where it’s really specular enhanced, it’s really a map of the surface normals and these images were so rich and informative, we were picking up on things right away. And again, it’s the way it facilitated comparison, easy comparison, you capture these things, and then you had them on file, you could look at them, go back and forth, no need to go to the objects, and it was all right there on the computer screen. You can from the surface normal image that the specular enhancement really improved our view of the surface and even with this lousy image, I think you can see that already the incised lines are jumping out, you know those initial lines that were made to do the sketch of what the image was going to be, and then you also see the relief lines live up to their name. There is a three dimensional property to them.
So, what were some of the techniques proposed? Well, it was all over the board, from reeds to brushes to single haired or very few haired brushes, and even syringes where it was more of an extrusion technique. And this image on the right of a bladder and some sort of delivery nozzle was taken from one of the major scholarly publications on how were the relief lines produced.
So, the other component of our research, other than looking at things with RTI and laser scanning confocal microscopy, was to do mock ups for ourselves, investigate the research, can we reproduce what were are seeing, what we have a better standing of through RTI? Can we reproduce it right in the lab with our own instruments or tools? One of the things that RTI was showing right from the beginning was that the relief lines were starting to separate out pretty consistently as two distinct types of lines. There was a ridged line, so either a rounded topography to the line, or there was a furrowed line and some of the straight lines, the straight diagonal lines that you see, those are the furrowed lines. If you look closely you’ll see a little trough that runs down the center of that line. Whereas the lines on the snake head and around the eye or the curved lines to describe the ends of the garment, those are the ridged or domed lines. So, we’re thinking there’s got to be a clue in the fact that there is these two different profiles that we’re seeing.
So again, ridged line versus furrowed lines and went back to our RTI’s and did further investigations. The little dots here, that’s a separate discussion I’m not going to get into right now, but I want to call your attention to the loops at the top and the way the specular enhancement with these RTI images really brought out the physical properties that characterize these types of lines. How the way the ridge itself folded over around the tight curve at the loop. Well we launched into a series of different mockups. Here’s one of our setups. The image on the left shows our extrusion instrument is basically a tube but for the nozzle we had used the shaft of a feather and using basically like a Gesso mixture in this case, tried to start producing the types of lines that we were seeing on the surface of the Greek vases. The image on the right, those are two different stitched images from our confocal work where we are looking for our mock-ups, and not finding out anything great, but I just wanted to point out that it’s pretty irregular and again and again, no matter how hard we tried, at the start of the line there would be a heavy deposit right in the beginning and it was almost impossible to eliminate that feature.
So brushes didn’t work, you couldn’t get the turnover properties of that and then we got into these single hair brushes that had been proposed by a Greek vase scholar, Zadeli decades before. Here these are all different brushes that Paula made up. They included horse hair, human hair, cat hair, pig bristles, and she really put it through the paces and here is a charged brush. Here she is laying the brush down in a manner proposed by Zadeli. He talked about laying the hair down and just lifting it up. So, it’s laid down and then lifted up, and sure enough we were getting that ridged feature, over and over again with that technique. But we couldn’t get that furrowed line using that same technique. Here’s a small bristle that we used to produce the loops and an analysis of those loops with confocal images really provided some pretty compelling evidence that the features produced with what is a modern reproduction on the left compared well with what we say in ancient loops on the right.
So, what about that furrowed line? Well, we proposed that there is actually a distinction between techniques, the laid line and a pulled line. A pulled line is when you put in the single hairs down or double hairs, and pulling it across that surface, you achieve a trough profile. Here’s two different confocal elevation maps and again the evidence for the similarities in the profiles are really compelling.
I’m going to skip through these parts, more metrology stuff. One of the other parts of the investigation was to look at the contour line. The contour line is that big thick line which you wouldn’t see without RTI, but right along here is a ridge that first inscribed or cut out a larger shape. So we wanted to know what came first, the relief lines of the figure, or these contour lines? Then RTI was pretty great at allowing us to zoom into these features and pick-up properties that had to do with sequence. You see these relief lines extended into the background and then you see this contour line overlapping and again and again we saw that. Enough so that we felt that there was a pretty standard chronology used here. Very gratifying to go down the Mass. pike and go to the Boston MFA and see a shard that has been the study from a lot of Greek vase experts, called the kylix fragment and there you see a painter using who knows what kind of tool to decorate the surface, you know scribes, brushes, everything has been proposed but with RTI it was really nice to see that the ridge there, between the instrument held in the hand and the vessel, really pretty much describes that of the Linierhaar single haired brush.
So, this is going to be lightning speed but if you’ll allow me. This is an ongoing treatment at the Worcester Art Museum and it’s on loan to us from the Royal Museum of Fine Arts in Antwerp. You’ll see right away that there is something odd about the portrait. It’s said to be by early Van Dyke and it is pretty much agreed by all scholars that it is, but you’ll see some pentiment around the collar and around the sleeves which merited our investigations using infrared reflectography. You start to see even more of the pentiment, the fact that the man originally had one of his millstone collars that extended all the way around and being a slave to fashion, he brought the artist back when the styles changed and changed the collar completely. So in x-ray, it really jumps out. The two vertical bands are strips of canvas that are on the panel joints, but you see that millstone collar and you can also see plenty of changes around the sleeves. But neither of those techniques brought our attention to something that was noticed in the lab by somebody who was less than five feet tall, standing in front of this painting on a sunny day, and noticed a little feature in the surface of this painting. We thought and we looked at it and yeah, there is something there. There might be an inscription underneath there that was covered over at the time of the changes to the portrait. So we went back to the x-ray and right into that detail area which was in the top left, and there was nothing there on x-ray. We went into IR, there was nothing there in IR. We even did a multi-spectral approach with false color IR, UV induced physical florescence, false color UV, none of this was showing up, but of course topography and having RTI in the lab, we were hungry to look at this with RTI. Here again, the detail in visible light and then in RTI, you can see, this might be an act of faith on your part, but you can see some features in here that suggest an inscription.
Meanwhile this is being carried out by Matt Cushman, and I suggested to Matt, let’s go back, let’s look at our RTI, we can never see the whole inscription in view all at one viewpoint. But, because it’s interactive, let’s move around the lighting and you can, with confidence, outline those areas you feel are definitely features. So we went ahead with that and this is the overlay that he came up with. This represents multiple views or multiple captures taken from the RTI and then compiled as overlays later. So it was great that we could go back to the museum, the lending institution, and say, you know, your painting that you’ve often wondered what year it was painted, we now have evidence to tell you that it was painted in exactly 1619.
Just a final thing is that we did map scanning with XRF, elemental map scanning, nothing in lead, nothing in mercury, nothing in copper, but this is all subsurface analysis. I wanted to do a full map of that area. It was pretty great to come to iron. This is the area, just focused on one of the letters which is shaped as a J. You’ll see that same letter comes up. Not only in iron, but in manganese, which suggested it was done with a iron oxide in umber pigment. So, it was pretty great that we could go that far in RTI, having facilitated it. So the conclusion, maybe it’s coming out like a mantra today. RTI can in some cases, provide insight where other established methods, such as X-rays and infrared reflectography cannot and therefore is a really valuable method in the larger field of technical art history.
Thank you very much.
Philip Klausmeyer earned art history and studio art degrees from the University of Massachusetts, Amherst, and an MS from the Winterthur/University of Delaware Program in Art Conservation, specializing in paintings. In addition to fifteen years at the Worcester Art Museum, his conservation work includes time at the Strauss Center for Conservation, Harvard University and the Museum of American Art at the Pennsylvania Academy of Fine Arts, Philadelphia. In 2009, he received a PhD from Clark University where he focused on applications of immunological methods in the analysis of art materials. He is currently Worcester’s conservation scientist and associate paintings conservator.