Volume 17, Number 2 .... May 1995

Conference Reviews

Three conferences are reviewed in this article:

1. Digital Imaging Applications in Conservation, October 17, 1994, Marina Del Rey, CA. Review by Chris Stavroudis.
2. Varnishes: Authenticity and Permanence, September 19-20, 1994, Ottawa. Review by Mary Piper Hough, Virginia Rasmussen, and Chris Stavroudis
3. The Structural Conservation of Panel Paintings, April 24-28, 1995, Malibu, CA. Review by Montserrat LeMense

Digital Imaging Applications in Conservation

October 17, 1994, Marina Del Rey, CA

On October 17, 1994, the Documentation and Scientific Programs of the Getty Conservation Institute hosted a seminar on Digital Imaging Applications in Conservation. Eric Lange, a Research Fellow/Consultant at the Institute, spoke on "The Development of an On-Site Digital Imaging System for Site Documentation".

The documentation program used the David Alfaro Siqueiros mural in downtown Los Angeles as a difficult, real-world field test. The mural, painted in 1932, is 18 feet tall and 77 feet long and is located on the exterior second-story wall of the Italian Hall on Olvera Street. The mural is presently covered by a wooden and fiberglass lean-to which allows access only in a very narrow corridor. Because of the access problem, an overall photograph of the mural has not been possible.

The advantages of using digital image capture over conventional color photography include: broader dynamic range (the Leaf Systems digital camera has a range of 10 f-stops as opposed to 7 f-stops with film); linear response; balanced panchromatic response; faster processing; and the ability to preview the final image at the time of exposure.

The goal of the project was to use commercially available image acquisition and manipulation hardware and software, so that the system would be accessible to, and replicable by other individuals and institutions. A Leaf Systems digital camera back was attached to a Hasselblad camera. The camera back is based on a 2048 x 2048 CCD array which is thermoelectrically cooled to improve the signal to noise level. The camera captures images in 14 bit depth. Color images are acquired by taking three images sequentially through red, blue, and green filters. A Macintosh Quadra 950 with 128MB of RAM, 1Gbyte hard disk and a magneto-optical recorder with 1.3 Gbyte optical cartridges was used to download and process the images from the camera.

Many of the details of the project were dictated by the final resolution desired for the image. Four pixels per square millimeter was chosen as a compromise between ultimate detail and manipulatable digital file sizes. Based on this target resolution, the number of pixels in the CCD camera, and the limited distance the camera could be moved back from the surface of the mural, the ideal focal length lens was calculated to be 42 mm. A 40.9 mm (nominally 40 mm) Zeiss lens was chosen. This allowed each exposure to cover 39 square inches. A 2" overlap was placed on each side to facilitate alignment of the images into a final mosaic. The mural was imaged with 156 (6 vertical by 26 along the length) individual high resolution pictures.

The necessary even and constant illumination was supplied by two electronic flash heads with soft box diffusers. These were mounted on a custom designed camera/light stand. The lighting was balanced by imaging a grey card and, using software, enhancing color and tonal differences as a guide in positioning the lights.

Because the floor in front of the mural was very uneven, a track and rolling scaffolding system was constructed to allow accurate positioning of the camera and lights. Because of the accuracy of the positioning system and the evenness of the lighting, images could be assembled into a mosaic without distortion or additional edge smoothing.

Each RGB image was acquired in 42 bit depth (14 bits per color channel) resulting in a 22MB file for each image. Because Adobe Photoshop 2.5 software can only manipulate 24 bit color, the images were sampled down to 8 bits per color channel by Leaf's camera control software. The high resolution Photoshop files were 12.6MB per image. To assemble the individual pictures into a final composite, each exposure was resampled to 335 x 335 pixel low resolution images, and these were combined into a single 50MB image. This represents the first contemporary overall view of the mural. [A brilliant trick for final registration of adjacent images: Invert the colors of the image being added to the composite and set opacity to 50%. When the images are aligned, the overlapping regions of the images will be an even grey. If the images are out of alignment, the grey will have an embossed appearance.] As Photoshop 2.5 can include 16 layers in each image, color coded overlays were drawn into separate layers to indicate: border of the individual images, plaster joins, cracks, water stains, previous restoration, and a 1932 black and white image of the mural. Each layer or channel can be turned on or off interactively.

The afternoon speakers, Adrian Heritage and Alison Sawdy, are graduate interns at GCI; Adrian with the Documentation Program and Alison with Scientific. Each presented their 3rd year diploma research projects conducted while in the Wall Painting Conservation Course at the Courtauld. In their internships, Alison and Adrian will be collaborating on a project investigating combinations of soluble salts in various substrates. The action of salt growth and its deleterious effects will be documented with video microscopy.

Alison spoke on "The Deterioration of Wall Paintings in Hardham Church, West Sussex, England: The Role of Salts and Organic Coatings". Adrian spoke on "The Potential for the Use of Video Microscopy In-Situ". A video microscope consists of a video chip mounted behind a special lens. The system is light weight and appropriate for use on site. Magnification ranges from 5x - 1000x (as measured on a 14 inch video screen) are possible with different lenses. At magnifications above 40x, the depth of field is very shallow and the system is prone to vibration.

Varnishes: Authenticity and Permanence

September 19-20, 1994, Ottawa

Part II

(for Part I see WAAC Newsletter, Jan. 1995, Vol. 17, No. 1)

[Ed. Note: We have just learned that the proceedings for this conference will be published by Archetype Books, U.K.]

Because of its significance for conservators, the WAAC Newsletter is providing a longer than normal review of the colloquium "Varnishes: Authenticity and Permanence". The event was sponsored by the Canadian Conservation Institute and Canadian Heritage, and was organized by Dr. Leslie Carlyle and James Bourdeau.

The first part of this review covered the talks by René de la Rie, Jill Whitten, Alan Phenix, Stephen Hackney, David Rainford, Raymond White/Robert Feller, Stefan Michalski, and James Bourdeau, as well as a separate review of the workshop, written by Neil Cockerline. A final installment will appear in the next issue and will include summaries of the talks by Cathleen Hoeniger, Helen Glanville, and Vojtech Jirat-Wasiutynski.

Anne Ruggles, Paintings Conservator, National Gallery of Canada, Ottawa, spoke on "When and With What to Varnish -- Some Practical Considerations".

Anne Ruggles discussed recent treatments of paintings by artists of the Canadian Group of Seven in preparation for an upcoming exhibition. Ruggles investigated unusual characteristics found in the condition of some of these paintings. A Lawren S. Harris which had been pastelined and varnished with PVA displayed sensitivity in the orange when recently spit cleaned. Cross sections revealed red-orange pigment on top of the PVA varnish layer. Had original pigment migrated to the top of the varnish? Records indicated that the unvarnished painting had been cleaned in 1951 with Borax before the PVA was applied. Ruggles' conclusion was that the paint had been rendered sensitive either by the Borax or by the varnish's solvent carrier.

Several colors in an unvarnished 1913 painting by J.E.H. MacDonald also showed sensitivity to water and mild solvents. The yellow (identified as zinc), the orange, green and red all displayed the same sensitivity. Ruggles questioned whether the paint itself may have been inferior, or if some prior cleaning of the painting may have affected the stability of the paints. To help protect the sensitive colors, Ruggles applied wax locally to some areas.

A 1929 landscape by Yvonne McKaque-Housser also showed unusual characteristics. In a recent cleaning, varnish which was removed looked red-brown and was observed in cross-section to contain pigment particles. In addition, the yellow pigment in the paint film again showed sensitivity.

The Group of Seven paintings, which are now about 75 years old, appear to have the following problems in common: The yellow-orange and red-brown paints are sensitive. The zinc yellow and red-browns are resistant to drying. Pigment migration into the varnish layer is common and appears to be induced by the varnish delivery solvent.

Ruggles noted that it is important to record the materials used in treatments. She cited a paper by Leslie Carlyle in 1990 which quotes period literature recommending 6 months to several years before revarnishing. She recommended that the stronger the solvents used in cleaning, the longer we should wait to revarnish. In recommending varnishes, Ruggles indicated that, historically, dammar in turpentine seems to be generally safe. She closed with the observation that in fact these paintings may not have been intended to be varnished.

Marion Barclay, Senior Conservator of Paintings and Contemporary Art at the National Gallery of Canada was the next speaker who discussed "Aspects of Varnishing Historical and Contemporary Paintings".

In this presentation Marion Barclay discussed the importance of the finish of a painting, whether a painting does in fact need a varnish, attempting to understand both intent and context, and how certain varnish options age both physically and aesthetically.

In her abstract she observed that, in a retrospective, paintings of one artist's work from different institutions appeared dissimilar. In contrast, specific institutions can tend to homogenize a broad range of paintings with treatment by one specific approach. She stated that inflexible individual varnishing approaches can do a great disservice to individual pictures and to their authors.

She asked the following questions: should we be more sensitive to the amount of time allowed between cleaning and varnishing a painting? Should we be more sensitive to the solvent delivery or to the resin/solvent ratio?

In 1980, she prepared varnish test samples on a study painting. Fifteen different coatings were brushed in strips on the painting: AYAA, AYAC, AYAF, AYAT, B72, B67, B66, AW2, Dammar, n-butylmethacrylate (Lucite 44), and polycyclohexanone (Ketone Resin N), all 20% w/v in toluene; and the commercial varnishes: Grumbacher Mastic, Winsor & Newton Copal, Soluvar Matte and Gloss. A control strip was left uncoated. The painting hung for nine years on the west wall of the conservation lab, with no UV protection, and some temperature and RH fluctuation.

She observed that the accumulated dust was more dense on the strips of Lucite 44 and Soluvar Gloss. Under UV light, the copal appeared orange and opaque, the mastic had a yellow tinge, and the dammar did not show much fluorescence. In terms of gloss, the synthetics remained the same while the natural resins, which were glossy when applied, had calmed down with time.

The physical surface can be a very subtle thing. She illustrated this point with a number of examples. In an 1890 painting by Leduc, dammar varnish was applied following the brushstrokes of the paint. She termed this final expressive feature a "painterly varnish". In a 1924 Chagall, the artist added a local glaze to the red roof. Carl Schaefer applied a partial coating over the trees and house in his 1931 "House in the Woods".

She presented statistics on the surfaces of paintings which were included in the "Crisis of Abstraction in Canada" traveling exhibition. Out of 120 paintings, 24 had highly glossy varnishes, 14 of those applied at the request of a dealer and 10 applied by the artists; 46 paintings possessed a matte and glossy combination; 51 had matte finishes.

With the introduction of artist's acrylic paints in the 1950's, the interplay between matte and glossy areas tended to become less prevalent, resulting in more uniformly matte surfaces. A choice of varnish for these paintings continues to be problematic. As artists began to use a wider range of paints the interplay of matte and glossy areas on the surface of the painting has become important.

With the introduction of synthetic varnishes in the 1960's, the hope was to have an easily reversible, non-yellowing coating. In fact, these varnishes have become quite gray and opaque, especially Lucite 44 and Ketone Resin N when beeswax was used as a matting agent. With aging, synthetic coatings can produce surfaces which become flat and dead in appearance within 20 to 30 years time.

Of the synthetic resins available at the moment, B72, introduced to the conservation department in the 1970's, seems to offer the most possibilities. It appears to have stood up well on paintings in the National Gallery.

She suggested that before applying a varnish that one should consider whether a synthetic varnish will age more appropriately on a cool colored painting, and whether a natural resin varnish will age more appropriately on a dark, warm colored painting. Will varnishes with additives, e.g., wax, age appropriately?

Barclay does not believe in a standard studio varnish. The knowledge of how varnishes age in a museum environment is of critical interest to her. She believes each painting and its context should dictate which, if any, varnish should be used.

Historically, the National Gallery used mastic from 1913 to 1930 and dammar from 1940 through the late 1950's. Paintings varnished in these periods are generally exhibitable in their present state. Visual problems arise when the varnish was applied too thickly or unevenly, when wax was used as a matting agent, or when two different resins were applied on top of each other.

The finish of several paintings at the National Gallery was discussed. Those that have been varnished with mastic or dammar have held up well, and the varnish remains easily removable. B72 has also held up well although at times tends to make the painting look like a reproduction of itself. A Hans Ewer painting had undergone several cleanings and revarnishings in the past. After a recent cleaning, it was left for several months prior to being revarnished with several thin coats of 3-5% dammar in distilled turpentine. Six to seven weeks were allowed between brush applications.

Barclay concluded her presentation with the following observations. She noted that display, location, and lighting are critical. Large immaculate institutions can tend to homogenize works of art. Secondly, there is a great range of materials available to the conservator which raises the specter of a "perfect finish" and inflexible perfection. Do we use state of the art materials and techniques just because they are there? Thirdly, she suggested that artists have not changed much over time. Some care about craft and follow available instruction manuals to the letter, while others ignore the manual completely. Without some clear documentation, it is difficult to say what type of finish the artist would have preferred. Fourthly, communications between the artists, historians, scientists and conservators have a bearing on how paintings ought to be finished. Her final point was that fashion can play a major role in how paintings are finished. Fashions are a standard of beauty projected by the different generations within cultures. This being true, there is no final truth about the surface characteristics of paintings and it may be said that each age will restore its own view.

Leslie Carlyle, Paintings Conservator from CCI, was a co-organizer of the colloquium. She spoke on "Reproducing Traditional Varnishes: Problems in Representing Authentic Surfaces for Oil Paintings".

The premise of her paper was best stated in the abstract for her presentation: "In recent years, many paintings conservators familiar with synthetic resins have been returning to the use of natural resins or have been turning to newly introduced synthetics which attempt to reproduce the effect of natural resins. Some are making this choice on the basis that the natural resin or its synthetic imitation can achieve a surface more like the appearance of the original, or traditional, varnish."

"Many assume that dammar is a good example of an original varnish. In fact, dammar was apparently introduced only in the nineteenth century. Although it may have been popular in Germany, it was not widely adopted in Britain, where mastic was the only spirit varnish recommended for paintings throughout the nineteenth and well into the twentieth century. As well, spirit varnishes have not been used exclusively; oil-resin varnishes have had an even longer history of use. Although oil-resin varnishes are certainly not a desirable alternative for a paintings conservator today, our view of what could have constituted an original varnish must include these materials as well."

"If we are going to adopt natural resins on the basis of an "authentic" appearance and search for alternative synthetic materials which better replicate the appearance of natural resins, then we should consider more closely what was actually in use at a given time and place. Are mastic and dammar actually as interchangeable in appearance as has been assumed by some twentieth-century researchers, and how do these spirit varnishes compare with oil-based varnishes?"

Leslie examined published recipes for historical varnishes and found a wealth of information. Reproducing the historic varnishes yielded some surprises. The 1833 recipe for mastic varnish recommended sorting the resin (selecting only the cleanest beads leaving out the dirty and soft ones). Bruising the resin by rolling it in a glass jar with glass chips. This removed a waxy component onto the sides of the jar. After the resin was dissolved in turpentine and filtered, it was aged open to the air for nine months. It remained clear and slightly yellow (dammar remained cloudy). Older turpentine that had been allowed to settle ensured a cleaner product.
Other recipes of the 1840's include the addition of 1/16 part rosin to act as a clarifying agent. It also produced a fatter surface.

The biggest surprises were in the appearances of the historical varnishes. The mastic was found to have yellowed noticeably on a painting after one month. The copal was a gentle yellow when applied thinly and produced a soft surface to the varnish. These observations contradict some of our most cherished beliefs. The golden glow did not take years to develop; the artist was aware of and probably considering the yellow tone of the final varnish; and the tone of the painting was modified when the varnish was applied.

She suggested that perhaps we should reexamine our premise that a varnish should be clear. And provocatively asked, should we be using slightly yellow varnishes on older paintings?; and should we think in terms of historical tone and gloss of painting surfaces?

The Structural Conservation of Panel Paintings

J. Paul Getty Museum
April 24-28, 1995, Malibu, CA

The Symposium was attended by over two hundred participants from 20 countries. The paper topics included: panel painting techniques, panel construction; history of panel conservation; current conservation techniques; wood properties, analysis, deterioration, and treatment for insect infestation; framing; microclimates; and problems of transit. The Proceedings will be published by the Getty in the next year and will undoubtedly become a basic reference for this subject. The following two talks are briefly reviewed here as they were of more general interest.

Microclimate Boxes for Panel Paintings. Jorgen Wadum, Mauritshuis, The Hague,The Netherlands

Mr. Wadum discussed the increased use of microclimate boxes over the past three decades and described three basic types of microclimate boxes and some of their variations.

A candidate for a microclimate box was described as any panel exhibiting chronic problems. An important factor in a panel's reactivity is the exposed surface to volume ratio. A high proportion of exposed surface to volume of wood means a more reactive, more at risk panel, or more simply put, thinner, bigger panels are more sensitive to environmental fluctuations.

Some of the considerations in building a microclimate box are: the panel's past problems and conditions; the travel or display situation; materials to be used in box construction; type of buffer to be used -- wood, paper, textile, Artsorb, silica gel, etc..

Mr. Wadum characterizes three general types of microclimate boxes: Type A - a box with a panel and a buffering material; Type B - a box with a panel; and the less common Type C - a box with a panel and an altered gaseous content.

The typical Type A box holds a hygrometer, and utilizes either conditioned silica gel or Artsorb, to maintain RH. A famous example of a Type A box is that containing the Mona Lisa, an example of a painting and its frame within a microclimate box. An aesthetic advance is a Type A box that conforms closely to the panel and is able to fit within a frame, making the microclimate box no more intrusive to a viewer than glazing alone.

Type B boxes don't rely on a separate buffer, they maintain their own equilibrium at changing temperatures. Research on these boxes, using data loggers placed inside to monitor the environment, determined that the changes in the internal climate were negligible, particularly if the interior air volume is as small as possible. An example of a type B box: a panel painting on honeycombed spacers to allow for air movement, a layered safety glass front, and a plexi backing.

Data loggers are valuable tools for monitoring the microclimate environment, particularly useful for traveling exhibitions, as they can be set to take readings at different intervals: frequent readings while in transit; at longer intervals while in the more stable exhibition environment. The inexpensive and easily read RH strips are also excellent in microclimate boxes, and are fairly accurate.

Technical Considerations for the Transport of Panel Paintings. Mervin Richard, National Gallery of Art, Washington, D.C. Charles S. Tumosa, and Marion Mecklenberg, Conservation Analytical Laboratory, Smithsonian Institution, Washington, D.C. The talk was presented in two sections: the structure and mechanical behavior of panel painting materials as they relate to transport, and specifications for transporting panels safely.

There are four elements to consider and guard against in the packing and transportation of panel paintings: temperature, relative humidity, shock and vibration.

The first step in transporting panels safely is to have a policy governing the loan of panel paintings. There are some panel paintings that should not be moved. Structures differ greatly, and they must be evaluated individually. Size can be an important factor in this determination. The larger the panel, especially if thinned, the more complex the structure, the more dangerous it is for it to travel. Condition must be considered as well. Any insecurity in the paint layer or the panel, any recurrent, non-controllable condition indicates it is not a good candidate for travel. The panels that can travel are the stable ones, and those with minor and controllable problems.

The information on the reaction of panels to temperature and RH fluctuations involved a number of graphs and is best obtained from the Proceedings or other publications of the CAL group. Worth noting, however, is that at low temperatures, paint will crack and shrink. The glass transition temperature, at which paint is at increased risk of cracking, is higher than might be assumed: oil paint -10 C, alkyd -5 C, and acrylic 5 C.

In discussing shock, the worst case scenario is the toppling of a panel, say a crate falling from a truck. In a case of shock, a panel that is not well supported will flex; the size and thickness of the panel affects the breaking point. The speaker differentiated between canvas and panel paintings when making recommendations for the design of a packing case to withstand shock: 50 G's protection was suggested for a canvas, 30-35 G's was recommended for a panel. The right amount of cushioning must be used to protect the panel against impact. If the cushioning layer doesn't give sufficiently, it won't mitigate the shock; if it is too soft, it won't insulate it from the shock. This refers to the compression of the foam on impact. The speaker suggested referring to the Dynamic Cushioning Curves in the Art in Transit Manual.

In choosing a method of transport, truck or airplane, shock isn't as much an issue as temperature change. A heated, air conditioned interior is a must. It's not really a problem for air travel as all cargo bays are above 5-10 degrees Celsius.

In handling and packing, the goal is to avoid moisture content change. Enclose the panel in moisture impermeable materials or a microclimate box. The storage of the packing cases is also important. If the case comes out of an uncontrolled storage environment, it can create a problem. If a painting wrapped in polyethylene is placed in a cold case, the environmental RH will drop and the materials will take up the moisture, raising the moisture content of the panel.

Tests indicate that condensation is not a significant problem in transit. It happens primarily when the panel package is cold in transit and is unpacked in a warm environment before it has adjusted to the new temperature. Allow time for it to equilibrate.

Insulation in a case is no substitute for climate control, as it only slows or buffers temperature change for a brief time. Polystyrene and polyurethane are better insulators than wood.The double packing case design is recommended, with 10 cm total of poly foam - 5 cm in the inner case and 5 cm in the outer case. Polyethylene is better for shock protection, polystyrene for insulation.

A final issue in transportation was raised. The speaker expressed concern over the popularity of the hand-carry for panel paintings. The appeal of a hand-carry is in having the package in responsible hands at all times, with uninterrupted supervision, in a climate controlled environment. However, the benefits can be off-set by several factors: if the panel is a little too large for a hand-carry, it is tempting to reduce the quantity and quality of packing in order to reduce the overall size; in hand carrying anything but a small panel, there is sometimes difficulty in storage or maneuvering in the small confines of the cabin.

In the Question and Answer session that followed several interesting points were discussed.

Q: Are there problems associated with the loss of microclimate environment due to loss of pressure in cargo areas?
A: Essentially a non-issue. Refer to the IIC conference paper that addresses this issue. (M. Richard) Strain gauges that monitor small dimensional changes have been used to monitor panels in this situation, they recorded no physical damage within boxes. A concern, however, is that plexiglass can bow and touch the surface of a panel painting if the microclimate box is too well sealed when subjected to low pressure. This should be considered when constructing boxes, perhaps the back surface should more flexible. (J. Wadum)

Q: In defense of hand carrying, panels that fly cargo often must be consigned to the "care" of airline handlers for as much as 8-10 hours prior to loading, a situation which allows strangers to handle the crate unsupervised. There is also the possibility of an uncontrolled temperature in the cargo holding hangar. Finally, due to last minute "re-packing" of the cargo space, the crate might be repositioned contrary to flight direction, which can damage a painting on landing and take-off.
What foam does M. Richards prefer?
A: Acknowledged validity of points brought up. Polyester Urethane foam gives a broad margin of error.

Q: There is a problem with newer paintings that are still off-gassing. Is an occasional flushing of a microclimate environment recommended?
A: Suggests an activated charcoal cloth to absorb various gases. Microclimate boxes in his institution were primarily used for short term loans. (J. Wadum)

Q: A microclimate box warning: a framed painting was hung for display within a vitrine; the vitrine was packed, not the painting, which came loose within the vitrine during transit.
A: Agreed with warning, emphasized boxes should be as small as possible.

Q: Recommended moisture barriers?
A: Referred to Oxford conference. A caution: never seal the back of a panel with a better moisture barrier than the front has. (M.Richard)

Q: Should a panel be packed sideways, if necessary, in order to have the end grain supporting it?
A: It is, in general, better to have the grain vertical. (M. Richard)

Comment: A participant suggested not wrapping the painting but lining the inside of the packing crate instead. He was concerned that a many venued show increases the amount of wrapping and unwrapping, often done by preparators, and that this increased amount of handling increased the risk of damage.