On Technical Imaging
Part conversation, part handbook, our In The Studio series visits conservators, framers and other craftsmen to explore what it takes to look after an Old Master picture. For the fourth installment, we met with Dianne Modestini at the Institute of Fine Arts (IFA) conservation center in New York to investigate the powers and limits of technical imaging. This includes the use of techniques including infrared Radiography (IRR) and X-Radiography (X-Ray), which have become an integral and unusual aspect of her work as a private conservator.
Dianne’s success as a restorer is evident in her work for the Kress Collection. Formed by the retail mogul Samuel Kress, his collection of over one thousand old master paintings, primarily from the Italian Renaissance, were donated to museums across the United States. Along with her late husband, Mario Modestini (1907-2006), she co-founded the Kress Foundation Program for Conservation in 1989, in which young conservators took care of the Kress pictures as part of their training. In this conversation, Dianne discusses how advanced imaging technology has transformed our understanding of how artists paint, and can, when done right, help with thorny questions around attribution and conservation.
In The Studio with Dianne Modestini
For a collector who is contemplating a purchase – what would typically be the first technical image to consult?
UV is always the first thing people look at because it’s very easy to use. All the auction houses have UV lamps. People even have portable ones they carry in their pocket. A UV light essentially excites the surface coatings on the very top and exhibits different fluorescences. Recent retouches don’t fluoresce at all: they reflect back the ultraviolet light—that’s why they look purple. If they’re older retouches and are under several layers of natural resin varnish, the fluorescence of the resin tends to block them. Although, if you look at them long enough, you can see some indication in the case of older retouches under the fluorescent material.
But there can also be a masking varnish, which is, in theory, a varnish that will hide the retouching. In other words, it will mask the retouching by fluorescing.
It is rare for auction houses and private galleries to provide IRR and X-ray images for the works on offer. So if a client wants to know more about a particular work, when would it be necessary to commission other technical studies?
You would want to do some further investigation with infrared or X-ray to see what you’re dealing with. But I think that the beginning collector should really form a relationship with a conservator. There are many good conservators working privately.
How are X-rays used in a conservation context?
The use of X-ray on paintings goes back well into the early 20th century. It can show the distribution of heavy-metal-containing pigments, which is essentially white lead. Anything that has white lead mixed with it blocks the X-rays and will show up on the film as white.
This is the Met’s Memling Portinari portrait, which has been studied extensively over the years, and this is an X-ray of it. There is a lot of white lead in the background, probably because the panel was primed with white lead primer, at least underneath the background. You can see the changes in it even in visible light.
What should a collector look for in an X-ray? What does it show and what does it not?
In terms of the condition of a painting, it can show tears, for example, that wouldn’t necessarily be visible to the naked eye if they have been very well mended. And it can show paint losses, which appear as black holes in the middle of another ground.
In terms of attribution, it’s actually quite difficult to depend on an X-ray entirely. It can only give valuable information if an artist’s brushwork is very characteristic, and can therefore show that the brushwork is similar.
It can also provide further information about the artist’s process. There might, for example, be two or even three versions of a painting. Sometimes infrared and X-ray can show which is the first version, because there’ll be a very clear underdrawing for the first, while the second and third versions will be much more mechanical.
But it requires a whole lot of different factors taken together to use all this technical information to inform an attribution. Sometimes it can, and sometimes it cannot.
What are some common but misguided uses of technical studies?
The most common one I see is when a private collector has their painting analyzed by one of these private companies who do such analysis. The report might indicate, for example, that the pigments are exactly the same as the ones used in other Caravaggios at the time, leading to the conclusion that, therefore, it’s Caravaggio. There was actually a very limited palette of pigments from 1200 until 1730, so it means absolutely nothing if they found the same pigments!
So are imaging techniques considered diagnostic tools?
Yes, in the sense that they can record the actual condition of a painting or can help you understand some pigment alterations.
Can technical imaging prove or disprove authenticity? For example, certain artists, like Rubens, are not known to use underdrawings for their paintings, so could the existence of it be used against its attribution?
To look at the underdrawing, we use infrared wavelengths that vary, according to the instrument, in how deep they penetrate. It’s very hard to see underdrawing in Baroque paintings because there’s so much dark paint in it. But just because you don’t see underdrawing doesn’t mean that it’s not there. What would tip you off to a painter like Rubens or Caravaggio, who use very impastoed paint, is if you see something that looks rather thin and doesn’t have any real texture. You can see this, for example, in that Caravaggio copy sold at Sotheby’s several years ago, which was then contested by the owner. The X-ray told the whole story, because if you looked at the Kimbell painting, and the original painting of the Cardsharps next to this copy, the copy was just empty—there was no brushwork, it was just purely what you saw on the surface. And an X-ray will give you information about that.
How common is it for conservators to rely on technical imaging?
It’s not very common. Unless they’re working in a museum, they won’t have the equipment.
I’m curious to know, of all the paintings you receive for imaging here at the IFA, what is the rough percentage coming from private hands?
Very few, because we have a small staff and don’t offer this as a service. However, there are private companies that can do it too.
I imagine that all young conservators are trained nowadays to read different types of scientific images.
Yes. And there are some projects, like the Bruegel and the van Eyck that have been put online by The Metropolitan Museum. In the online entries for some of the Flemish pictures, they have put up the whole set of technical images, which is very useful. The Rijksmuseum has started to put the Night Watch material online as well, but to do so is still relatively rare.
The technical imaging that’s done mostly by museums and institutions is usually not available to a wider audience. It’s held very closely, which limits the utility of these scans. Museums and people working with this material will give lectures and publish it. But the published image is not good enough to really study—you would need a high-resolution digital image for that. There have been efforts, however. I think the EU was hoping to have a project containing many scanned images and archives but there was a lot of pushback from the stakeholders and custodians of this material. They just didn’t want to give it up.
So, in the public domain, not only are paintings rarely studied this way, even if they were, the result is rarely made available to others.
Yes, people are very jealous of this information! It gives them more power—it creates a sort of cast—because there are very few institutions that possess these tools. Especially the latest innovations, including XRF (X-ray fluorescence) scanning and XRD (X-ray powder diffraction) scanning, provide so much absolutely essential information about a painting. It’s very unfair to anybody outside these institutions, as it puts people at a great disadvantage.
Yes indeed, the Met’s recent technical study on Jacques-Louis David’s Lavoisier portrait was all about putting these powerful new tools to work. Is cost an important factor?
No, all of these things are subsidized. Most museums have government subsidies that allow them to do this. So you could, in theory, submit a Freedom of Information Act.
Is there a bias, in terms of artists or schools, in the paintings with published technical information at the moment?
Some artists are studied extensively because there’s a concentration of paintings by this artist in national collections, like Dutch paintings at the Rijksmuseum. Other painters are not studied at all. I’ve never seen a technical scan of a painting by Chardin, for example, and yet, it would be very interesting and fruitful to study. Then there is the Bruegel project, for which conservators took an extensive look at every Bruegel they could find. The technical images are online, so if a collector wants to get an idea of what infrared shows, they could go and see. Sometimes infrared doesn’t show the underdrawing, but it can show the pentimenti, losses , and some simple pigment identification . All of these techniques require a great deal of scrutiny.
We think of conservation as a science but it appears that Alain Goldrach, who I spoke to earlier in this series, almost entirely relies on a ‘feeling’ when it comes to cleaning pictures. The parallel approach of science and artistry is fascinating: it appears that even when using scientific tools, one must exercise subjective judgement.
Yes, the approach to all of these technical studies is very interesting for understanding the artist’s process. They don’t have much influence on the actual cleaning or restoration of a painting, which remains fairly subjective except in rare instances.
Could you think of a rare example?
Yes. Last year, one of my students worked on a painting by Pompeo Batoni for us. As she was cleaning it, some color began to come off the lapel, so we stopped immediately. It was not clear whether it was the retouching or the original, but it seemed to be damaged. After we stopped, we looked at it again with ultraviolet light, and I saw immediately that the buttonholes were on top of a fluorescing varnish. I knew that in the 18th century, many artists, including François Boucher, used a layer of resin in between their dead coloring and the final coloring. Jean-Baptiste Oudry writes about this in his second Discourse to the Academy. And so we started looking at it again and took a cross section, in which you could clearly see this interlayer of yellow varnish. Because this layer was easily soluble, there was a great danger of removing the original paint. This also happened once with a Gainsborough, where the final work was on top of the varnish. If you just go by the principle of ‘if it doesn’t fluoresce, it’s not original paint’, you can get in trouble.
So in this case, the technical study helped you ascertain that it was not retouching but original, damaged paint. Can you tell us more about the Rembrandt Research project, which you worked on with John Brealey during your time at the Met? How has the attitude towards technical studies changed over the course of your career?
I worked at the Metropolitan Museum from 1974 until 1987. The Rembrandt Research Committee was formed in the 1970s. Their remit was to look at every painting by Rembrandt, and to decide, as a committee, whether it was by produced by the master, the studio or a follower. They made pronouncements all around the world, which were quite different from the traditional attributions, and some of the institutions that owned these pictures weren’t entirely in agreement. In those cases, X-rays were somewhat helpful in looking at the brushwork or pentimenti but, of course, a Rembrandt follower imitating the master could also have pentimenti, so it didn’t really prove anything.
At the Met at that time, the Dutch scientist Peter Meyers had a relationship with the nuclear accelerator facility at Brookhaven . They had come up with a technique for irradiating a painting so that various atoms were excited and increased to a higher energy level. By exposing the painting to these [X-ray] films , you could see the [pigment] distribution—this separated out bone black, for example, because it identified the element phosphorus, one of the elements that got very excited.
John became very interested in this technique, called autoradiography, for looking at works by Rembrandt. So they would get images of, for example, just Rembrandt’s hat, which was painted with bone black, and another image just of the flesh tones, and another of something painted with green or blue. This created a whole set of images, which said much more about the genesis of the painting than a simple X-ray did, and in a very conclusive manner. It established a lot about the way Rembrandt worked, because they would see the same thing in picture after picture. This was very helpful in attributing Rembrandts and to distinguish between studio work and work by the master. It was time consuming, however, because someone had to go out with the painting to Brookhaven to have it irradiated and then go out every time these isotopes came down to remove the film. I don’t think anyone has ever done it since; other things have come along, which have superseded it as a useful technique. ❖