Jump to content

A post of RTI-HSH stills of lead-soap microprotrusions on Google+


Recommended Posts

Dear All -


Dale Kronkright (Head of Conservation at the Georgia O'keeffe Museum) and I recently had a convo / exchange about his forum post in re to 'Lead Soap'


Honestly, I had zero idea of what lead soap actually was, so below is our thread:


Check out the G+ Image gallery of O'Keeffe's "Pedernal", 1941, oil on canvas 19" x 30"


Question (Marlin): 

Dale, this is interesting. Can you tell us more? Is lead soap a cleaner or a 'primer' like layer that was on the canvasses?
Are u extracting that information sub surface? Or is the image a 'blank/empty' area of the canvas?


Response (Dale):

The precise mechanisms are elusive, although the chemistry is clear. In O'Keeffe's case, an in the case of many 20th century painters who used commercially primed canvases from Belgium, France and Germany, the manufacturers appeared to load the lead-in-oil priming with additional non-drying oils to prevent full hardening of the priming. Remembering that from manufacturer to wholesaler, exporters and shipping, retailer art supply and then to the artist, who my keep their 20 yards for a year or more before cutting, stretching and painting, the canvas would need to remain pliable and stretchable for two years. A pure linseed oil priming, even if lean, would harden in a couple of months.

The free fatty acids in the non-cross-linked oils are available for alkaline hydrolysis. Each time there is a change in surrounding RH, water molecules move through the unprimed verso of the canvas, one direction or the other. These water molecules come in contact with available PbCO3 to form PbOH and CO2 and the alkali consumes a little free oil to create a soap. THe works that appear to have travelled the most, historically, have seen thousands of humidity changes and thus the alkalis have drilled tiny worm-holes toward the recto surface.


Ultimately, all the oils and fats are consumed leaving a hard, thin-shelled "blister" that often erupts, leaving a tiny pit. Often, the free oils migrate around the microprotrusion creating islands of saturated shadows around the protrusion. THe Doerner Institut in Munich found the same phenomenon in the paintings of the same period by Max Beckmann.

Stabilizing the humidity in the verso of the painting using micro-environmental framing technologies appears to slow the development of the micro-protrusions.


We are using RTI to both identify, map and monitor the rate of change of these tiny blisters - 50 - 200 microns in diameter.

(The 1/2" bearing in my Google+ set, when imaged in false color normal reflection vector visualization, is awesome,as it shows what the vectors are doing)

Link to comment
Share on other sites

Very interesting and clear explanation of the chemistry, Dale!  You can even see that some of the blisters have erupted, with small pinholes in the center, while others are less developed.


The last two images showing the normal maps appear to be taken at a sample resolution of about 2,500 to 4,300 ppi (e.g., ~1,500 p / 15 mm x 25.4 mm/inch = ~2,500 ppi).  What kind of camera and lens were you using to get this magnification, and did you have any difficulty using the highlight method at this magnification?  I found it was difficult to avoid shadows and get an even flash distribution at such distances.  Working alone made this even more challenging, especially when trying to avoid bumping things.


The highlights in the sphere in image 3 are quite blurry, but this didn't seem to be a problem for processing your RTIs.  This isn't surprising for macro-photography, when your depth of field is paper-thin.  There isn't much of a workaround when you need to place the spheres in front of the canvas to frame the area of interest.  In similar situations, I found I had to compromise and allow the spheres to be a little out of focus, since I really wanted the surface of the canvas to be sharp.  You mentioned in the Google hangout doing studies of the reproducibility of normals--did you notice any increased variability when the spheres were out of focus?  Fortunately, the RTIBuilder software seems pretty tolerant of blurriness of the spheres and can still find the centers of the highlights. 


To map the locations of the blisters, do you have an x-y coordinate for each blister on the canvas?  How do you control the position of the camera to map the entire surface of a large painting at such high resolution, if indeed that's your goal?


Thanks for sharing these!

Link to comment
Share on other sites


This topic is now archived and is closed to further replies.

  • Create New...