Refreshing Mona Lisa: part one

My colour adjustment of the Mona Lisa (reducing yellowing)
click to enlarge

I often repair or adjust old pictures that I use on this blog, doing such things as reducing the age-yellowing of paper, taking out stains, repairing tears and so on. Sometimes, I adjust the colours so that they look better with other images on the page. A few days ago I thought I might attempt to restore the Mona Lisa to something close to what da Vinci might have seen while working on it. I also thought that this must have been attempted before but did not look for examples as I did not want them to influence me in any way. It was obvious to me that severe yellowing has occurred on the painting as skies are  rarely green, but the criterion I picked was to lessen the yellow just until I saw no trace of it on the whitest part of the skin. I thought that da Vinci would have wanted no trace of yellow in that area. In doing so, I first noticed that the skin tones became very natural looking and she started to look like a real person and not someone close to death from jaundice.

This is the source image I used which had been adjusted to be close to the colours of the painting in its current state. You can also click on it to see it at about the same size as my adjustment (mine slightly cropped in selecting it).  da Vinci, himself, was  worried about yellowing: he used walnut oil instead of linseed oil and would tint his glazes to counteract the effect of age-yellowing. Charles Lock Eastlake, in Materials for a History of Oil Painting, Vol. I, 1847, explains his techniques and concerns:

"With all his sense of the advantages of the new process, Leonardo participated in the dread of oil which was so common among the Florentines. He preferred nut oil, as less coloured than that of linseed, and took infinite pains to extract it in the purest state. He appears at one time to have believed that not only this but all the fixed oils could be rendered perfectly colourless; but he must have found that, after all such precautions, time ultimately deepened their hue. He distilled these oils in the hope of obtaining a less changeable vehicle, with no greater success. With better promise of attaining his object, he confined himself to certain colours in the earlier stages of his pictures, with a view to couteract the subsequent yellowing of the oil. He prepared, and even completed them (their final glazings excepted), in a purplish tone, and thus provided by anticipation a remedy for the evil which he dreaded. With the exception of the Adoration of the Magi, in which the fainter shadows are greenish, there is scarcely a picture by Leonardo, whatever stage of completion it may have reached, which does not exhibit this more or less solid purplish preparation, varying from an ink-colour scarcely removed from grey — as in the Mona Lisa, as in the " Vierge aux Rochers " at Charlton, and as in an unfinished head in the gallery at Parma — to the almost violet hue of the Holy Family in the gallery of the Hermitage at Petersburg. De Piles remarks that the carnations of Leonardo incline for the most part to the colour of wine-lees, and that a violet colour predominates in his pictures; Rumohr notices the same tints in the head of Ludovico Sforza in the Ambrosian Library at Milan, and in other examples. Leonardo himself, describing a mode of painting with gum-water, recommends the use of lake and black among the colours for painting the shadows of flesh, the darker shades being strengthened with lake and ink. It appears, both from the unfinished Adoration of the Magi, and from Leonardo's writings, that he preferred a yellowish ground or priming. This was the opposite hue to his dead colour, as his dead colour was again the opposite to the mellow tone which glazing and time would give. On the same principle the tempera painters dead-coloured their flesh green, that the carnations might look fresher, and the Venetians prepared a sky with cream colour as a ground for blue."

When I had finished, I was surprised to see the grey sky, but I noticed that the blue of the water appeared very natural and the redness in the earth was a dead ringer for the Italian ochres of the siennas and umbers. More importantly, I noticed all of the subtle colours in the grey of the sky. The glazing technique, which  uses multiple layers of tinted oil often only a few molecules thick allows for such subtle variations and the the Venetian painters, especially, favoured the  juxtaposition of grey and colours: the former adding great luminosity to the latter. I think that da Vinci had borrowed that effect quite deliberately. I can demonstrate the effect of the glazing technique with the following graphic:

The colour chip on the left is RGB: 255,0,255; next to it is the same colour with its transparency adjusted to 50% and it overlays a chip that is 0,255,0. I have selected the area of the overlay and you can see that it is 127, 128, 127 which is as close as you can get to the mid-range in an RGB system of 255 for each colour. Creating well over 16 million different colours, this is a mathematical system which has nothing at all to do with human colour perception which is far more limited. The colour wheel, a relatively modern (18th. century), and thoroughly unscientific construction, was unknown in da Vinci's day. In the modern system, opposite colours are considered "complementary" so red:green, violet:yellow, etc. At the very least, it is a poor choice of words because if you paint a thin line of red against a background of green, the vibratory effect can almost set your teeth on edge. What you are actually seeing is discordance, something like the wrong note being played in a familiar piece of music. My late wife, Carin Perron, was working on a new colour system that was based on normal human perception: we see different ranges of colours with greater or lesser strengths. We evolved that way. Other creatures can see colours we do not, so that, in itself, reveals the subjective nature of the colour wheel: wavelengths move down into infra red or upward into ultra violet but they never circle back on themselves in nature (coincidentally, in an invented colour system). What Carrie ended up with, geometrically, was a "colour egg" so the actual complement of green was not red but had moved toward the violet. This is how da Vinci started his paintings, using green and violet over the ground (which in his case was not white, but a creamish colour). da Vinci worked on the basis of perception and not theory.

The subtleties in the thicknesses of  each of his glazes, something unavoidable, anyway, when dealing with only a few molecules, gave da Vinci the effect of  very lively greys and this, also, is enhanced through simultaneous contrast in human perception. If you are very colour perceptive, you might see variations in the warmness and coolness of the grey overlay in the graphic above. Yet, wherever you place the colour picker, the results will be the same. Colours are not 100% reproducible on a monitor. If the overlay is not neutral grey, you should adjust your monitor's colour. When I used to design new colour scheme in RGB for eventual printing, I had to actually work blind, with many trial printings to get them exactly right and colours cannot be designed mathematically, either, because of the human perception factor. When a large corporation tried to reverse engineer the colours of the maps I was doing for them rather than paying for mine, they gave up after a year because it cannot be done.

I will conclude tomorrow with the nature of the lakes Eastlake speaks of; what has faded in the painting; another effort to adjust the colours of the Mona Lisa; and a copy of the Mona Lisa that is thought to represent da Vinci's original colours. Of course, I will be somewhat critical of these!


John's Coydog Community page