Robert's Home Page | Mona Lisa Home Page | Mona List
 Previous Mona Image Page | Next Mona Image Page

Mona Lisa Images for a Modern World

Digital Mona Lisas
A survey of the use of the Mona Lisa as a metaphor for digitizing images, including related efforts.
by

Perhaps due to its universal capacity to be instantly recognized, the Mona Lisa has come to be adopted by computer hobbyists and scientists as the prime motif upon which their technical experiments rely. Even so, the popularity and ubiquity of  the Mona Lisa icon is not the only cause for its use. The sophistication of Leonardo's color and the range and subtlety of his tonalities, coupled with the work's relatively poor state of conservation, set a usefully high mark for experimenters to set goals. Moreover, Mona's face has come to stand for the universality of the creative process, and, in particular, the process of image making. For these reasons the Mona Lisa has come to serve as the epitome of all that belongs to the public, to civilization and to every culture that claims her. The examples chosen for this page fairly exhibit the range of experiments in this arena. Examples show how a variety of picture elements have been used to serve these ends. They may range from the use of ASCII (alphabetic) characters, building images from Lego blocks, and even use of images, themselves, as picture elements.

Not chosen for this survey is the way computers have been used to manipulate the image of Mona Lisa, such as morphing and intentional distortion. (See Lillian Schwartz's experiment and the Mona Lisa flip book discussed above.) Also not treated, are Mona Lisa works based on the kind of distortions and patterns suggested by typographic technology.
fig. 43A An ASCII Mona Lisa from http://www.geocities.com/h2lee/ascii/monalisa.html. Note how the ASCII characters form picture elements. I suspect that each character in the ASCII set was given a density value, and then the image was scanned and associated bit-by-bit with the correspondingly dense ASCII symbol. The resulting image, while recognizable, is rather crude. One wonders how many variations in grey tones the ASCII characters permit. In a more sophisticated fashion, this idea was picked up by Robert Silvers (See his "photomosaic" Mona Lisa, below.) who uses unaltered images as picture elements.

On related experiments, see: http://www.digibarn.com/collections/books/computer-lib/dream-machines/dm10-large.jpg

fig. 43B A different kind, perhaps even more advanced ASCII interpretation may be found at this location: http://onickz.ifrance.com/onickz/deck91gfx/onickzakasc_c.htm (12/02). In this case ASCII characters are given tonality and are placed against a tinted background to create a kind of chiaroscuro effect whereby the characters can be either lighter or darker than the ground. At left is a small detail of this rendition -- here converted into a graphic file. The color of the background field seems to be optional, but should be a middle tone to show off highlights and shadow. In contrast to the example above, it does not seem as if the characters are associated with density values -- all the characters are uppercase san-serif alphabetic.

Another example, from 1965, with expanding close-ups: http://www.digitalmonalisa.com/

fig. 43C The two images on the left depict the work of Eric Harshbarger whose medium in this case is Lego brand interlocking blocks. The work, executed in November 2000, is entitled "Mona Lego." It uses the standard Lego "primary" colors: red, green, blue, yellow, white and black and is composed of over 30,000 blocks, measuring six by eight feet, and weighing over 45 pounds.

The artist takes Lego commissions. Here is his webpage that describes his projects and offerings: http://www.ericharshbarger.com/lego/

fig. 43D The technology employed ultimately depends upon the pointillist experiments of Seurat, who used primary colors that depended upon optical resolution to produce the impression of mixed colors. On the left is a detail of Eric Harshbarger's "Lego Mona." The two photographs shown derive from the artist's website, which may be consulted for additional details.

The Lego Monas are closer to the digital ideal in that they use repeating shapes of equivalent values. Only the color changes. Many of the works classified as "digital" here belong to the virtuoso mode of Mona Lisa creations and can be thought of as (unconscious?) attempts to rival the Leonardo's inventiveness.
fig 43E (below, left),
fig 43H (below, left bottom		  

fig. 43F (below, right, top),
fig 43G (below, right bottom)
From Byte Magazine of March 1987 comes a report on the status of image processing. The cover painting (left) and the work on right are by graphic artist and illustrator Robert Tinney. In them we find an image in the process of being painted with a brush, but as the brush lays down color, the image begins to show the mark of digitization. As such, these images serve as metaphors for the process of transforming analog (painted) data into digital data. Among the articles in this section is one on Digital Image Processing in Art Conservation by John Asmus (p. 151). The author highlights the "non-destructive" benefits of conserving and recreating art via digital methods. He was entrusted with the search for one of the most elusive of lost works, the Battle of Anghiari, a mural Leonardo painted in the Hall of the Five Hundred within the Palazzo Vecchio in Florence. This famous work, a model for many artists, may have been lost forever or merely just covered up by the painter Vasari. Asmus was also commissioned to investigate the possibility of restoring the Mona Lisa by using a surrogate image (a transparency) on which digital tools could be used to see what the painting might look like were the varnish removed. (See p. 102 ff.) By digitally removing the varnish "the sky changed from brown to blue, the skin from yellow to alabaster, and the gown gained a deep greenish tint. ... The main limitation to emerge from that demonstration was the 512 by 512 resolution of the system." [That was in 1981!]

The "digitized" image at the right (which I have dubbed "Next") by illustrator Christoph Niemann was prepared to accompany a letter by Philip F. Palmedo to the New York Times of  February 10, 2002. In his letter, Mr. Palmedo responds to the ongoing debate as to whether the Mona Lisa should be restored. Some believe it should be returned to its original state, while others believe that any such restoration will produce only speculative results and may in the process permanently destroy some of what Leonardo intended. He suggests that restoration should not be executed on the original, but should be carried out by a computer, which shall produce a surrogate. He says: "I suspect that Leonardo, for whom art and science were one (the seeking of essential meaning through observation and representation), would have applauded this notion. ... He would have been the first to try it out at the computer -- which he would have invented if he had had a little more time."

note: 9/16/03: Mr. Palmedo writes: "I was fascinated by your Mona Lisa web site where you cite my letter to the editor of the New York Times. Although the illustration was, of course, of the Mona Lisa, the subject of the article to which I was responding was Leonardo's "Adoration of the Magi." It is in far worse shape than the Mona Lisa, alas." How telling, therefore, is it that the editors of the New York Times, in all likelihood despairing of the ability of the "Adoration" to communicate the writer's message, chose to use the Mona Lisa instead. ed.

"Next" demonstrates the extent of the evolution in the use of the digitization metaphor from the Byte images to now. Whereas the earlier images focus on demonstrating how painting is turned into pixels, this latter image looks at the Mona Lisa as just another representation in an album where there will always be another image that comes next.

Letter-writer Kevin Pease (4/21/03) finds fault with the above "picture-album" metaphor. Rather, he suggests, correctly, it seems to me, that the imagery derives from the popular video-game "Tetris." He writes:

The falling tetromino in the upper right indicates that the context is in fact meant to be the game Tetris, seen as played on a low-resolution handheld device. As such, the icons adorning the frame are not indicative of an digital picture album; rather, the arrows are game controls and the word "next" indicates at its right the next shape waiting to fall into the playing field.

At first, the point of this seems vague, and it would seem that the illustrator is making the common crude generalization that video games represent the sum of everything computers do. But, upon reflection, it seems that to see the image being put together like a jigsaw puzzle is appropriate to the article, because that's what the machine is essentially meant to be doing: assembling a replica without adding any creative influence. [See further.]

Another variant of the digitization metaphor may be found in an advertisement for Compaq Computers that began to appear around 1999. In this case, the "ascii style" substructure uses only "1s" and "0s," the abstract symbols that serve as everyman's metaphor for bits of electronic information. Unlike other digital depictions, the Compaq advertisement attempts to show that it is a surrogate of the original work. Rather (taking a cue from Andy Warhol), the Mona Lisa is itself a page -- self-consciously a reproduction. Illustrated is a single page version of the advertisement that also appeared as a two-page spread.

 

 

 

fig. 43i, 43j and 43k
While the image on the far left looks as if it has the hallmarks of digitization, in fact the squaring "cubist" effect is the result of an optical technique that was employed in 1973 "during an experiment directed at finding out how people remember faces." The Smithsonian Museum once showed a similar portrait of Lincoln's profile, presumably the one (right) developed by two ATT Bell Labs scientists, Leon Harmon and Bela Julesz. Their experiment attempted to determine the minimum information needed for an individual to identify a well-known facial image. Lincoln and Mona, in this regard lend themselves well to these endeavors because their profiles and other features are so well known that they have become iconic. Not only was the experiment  useful for defining the minimum information needed for identification, but it also served to demonstrate the ability of the human mind to fill in missing information based upon expectation. Interestingly, as soon as the abstract arrangement of squares resolves itself in the viewer's mind, it is no longer possible to see the image without content. In this respect they are parallel to the work of painter Chuck Close and make possible the kind of mental reconstructions necessary to see the "photomosaics" of Robert Silvers (below).

These quasi-digital experiments have had significant influence on the fine arts. Without doubt, as already suggested, Chuck Close paid close attention to their system of communication. We also find Salvador Dali (below) using the Harmon/Julesz image of Lincoln, and an unknown creator of a sequence of Mona Lisas that replicate well-known cultural and individual styles (above), seemingly transforming the Mona image into a pseudo-mosaic.

Squared Gioconda. From Roy McMullen, Mona Lisa: The Picture and the Myth. Boston, Houghton Mifflin, 1975, p. 231, fig. 141.
See further an article in ABC Science News, which illustrates a work by Chuck Close and an article by Denis Pelli in Science Magazine ("Close Encounters--An Artist Shows that Size Affects Shape") upon which the ABC story is based..

  fig. 43l (left)

fig. 43m (right)   

A cover of the Scientific American issue of April 1995 depicts what the editors call a "virtual museum," which they define as "an infinitely expandable gallery of great art that can exist only within a computer." This 1995 demonstration piece for three-dimensional imaging required the efforts of four Macintosh computers laboring for 86 hours. In its virtual space we see light effects, including shadows, reflections and transparency.

The result may be heralded as a triumph of technology to represent subjectively determined images, but it can also be stigmatized as a tragedy for the cause of authenticity -- all depending. On one hand, the technology can just as easily be used to model future exhibits of physical works of art (which today in 2002, it is) as it can be employed to give shape to an individual's fantasy. In this case, the graphic artist seems to have taken no heed of the physical specifics of the works chosen. The resulting images ignore as much the original scale as it disregards the media of the original works. For instance, the so-called self-portrait of Leonardo (far left) is not a painting, but a drawing. The Mona Lisa seems larger than it does in real life, and the juxtaposition of one image against another seems not to make any sense -- no art-historical sense anyway. But perhaps that's not the point.

Rather, after having claimed success at digitizing works of art (if "success" is the operative word for the effort), what is interesting here, is that the technology has moved on to show that it can reproduce works in selected environments, subject to the subjectivity of vision, location and (perish the thought) curatorial disposition.

In this same issue of Scientific American appears an article by Lillian Schwartz entitled "The Art Historian's Computer" (p. 106) in which she demonstrates how computers employed by art historians can produce imagery to help test theories, suggest theories,  reconstruct damaged or altered sites, and create associations that might otherwise escape the human eye. In this article Ms. Schwartz puts forward the idea that Leonardo's Turin "self-portrait" (far left) proves that the Mona Lisa is based on the artist's own features. For many professional art historians, this popular theory is not proven, but the use of the technique at this time was groundbreaking, and even today continues to draw attention.

note: See references to Ms. Schwartz, above as follows: MONA15.htm, MONA16.htm, and MONASV12.htm, as well as her own website: http://www.lillian.com/.

fig. 43n

figs 43o, 43p

The image on the left is a demonstration image published by Adam Finkelstein and Marisa Range in their 1998 paper entitled "Image Mosaics" (Technical Report: TR-574-98, Computer Science Department, Princeton University, Princeton, NJ 08540, USA). It demonstrates, in principle, a method for constructing mosaic-based images somewhat like those created by Robert Silvers (below). The technical explanation that appears in this paper is outside the scope of this discussion. It is cited here primarily because the authors acknowledge that the urge to build image mosaics is derived from the work of modern painters. Referring to the work of Leon Harmon and Bela Julesz. (above), they first cite a technique employed by the Impressionists in which paint strokes are laid down that, in themselves, seem to have no representational value. When the observer, however, recedes from the canvas, the individual strokes merge and coalesce visually, causing the representational image to manifest. In addition, the authors cite Salvador Dali, who has produced works that use pictures as elements with which to produce larger pictures, as seen in his popular lithograph Lincoln in Dalivision (above right) where designs suddenly reveal themselves as elements belonging to a bust of Lincoln, the design of which doubtlessly was derived from the Lincoln pattern created by Harmon and Julesz (above). They also cite works by Chuck Close where large images are produced by combining myriads of abstract figures and designs. (See, for example an article on Chuck Close appearing on Artsnet of Minnesota or the articles above that discuss the work by Denis Pelli on the effects of size and scale.) Another precedent, not cited by Finkelstein and Range is the work of the Renaissance painter Arcimboldo, as typified by his allegory of Autumn (above right). Arcimboldo takes natural elements and recombines them as to suggest human and other shapes. In the examples provided by Finkelstein and Range, a single image is adjusted in tonality and then positioned as a function of the tonality in the larger image. See the Mona Lisa composed of Mona Lisas, above left. Readers may wish to identify this image's mask-like effect -- looking as if the face has been painted with layors of miniature faces.

Refer to "A Short History of Photo-tiled Pictures," by William Leigh Hunt.
fig. 43q and fig. 43r.
The technique devised by Robert Silvers is considerably more sophisticated than the one described by Finkelstein and Range, above. Silvers', Mona Lisa "Photomosaic" (his trademarked word) represents one of the most intriguing variants of image digitization. The technique that Mr. Silvers invented has earned him a broad commercial success in advertising and portraiture. Each larger image is composed of a set of photographs. When combined, the photographs form a mosaic-like matrix that builds a composite image of his subject -- as if each contributory image were a large pixel. The technique is actually more sophisticated than just combining images that have been adjusted or typed by density and tonal value. Instead, each image seems to have been classified by its own unique internal arrangement of color and density values. When they are combined, they produce an effect that seems to have a higher resolution than the texture of the wall of separate images would suggest. The illusion thus created appears to pop out of the matrix in such a way that the larger image appears to be denying the structure that created it. In addition, Silvers selects his component images so that they are iconographically related to the larger image. For instance the Mona Lisa is made up of myriads of pictures of paintings, his Babe Ruth is fashioned out of pictures of baseball cards, and so on.

(At right, detail of Mona Lisa's smile from a "Photomosaic by Robert Silvers."

Silvers' technique and its aesthetic and stylistic sources employed in his picture book Photomosaic Portraits, is discussed in a review by this author. In the review is an example of Silvers' version of  Dorothea Lang's Migrant Mother, from which a detail has been extracted that demonstrates the effectiveness of the technique.

Earlier uses of the "mosaic" technique employed picture elements that had no special pictorial or aesthetic interest -- they were merely subservient to their pictorial function. The figures selected were often individually uninteresting, such as the ascii characters, Lego blocks, or the miniatures of the main image itself. Silvers' component figures differ because they tend to be interesting images in and of themselves -- especially when juxtaposed within the main image. The consequence of these multiple centers of interest is to create a tug between the main image and its components. Knowing that the observer can't have both at the same time, he finds himself perpetually attracted back and forth between the whole and its parts, unwittingly drawn into a trap formed by the dual magnetic poles of Silvers' macro- and microcosmic inventions. Ironically, oftentimes in Silvers' images (but not with the Mona Lisa) one finds that the bevy of smaller images are more interesting in themselves than the single larger one.

As in the portraits by Chuck Close, cited above, when the observer approaches one of Silvers' images, getting ever closer to it, at some point the large image unresolves and only the small images can be perceived. Silvers' "photomosaic" technique represents an interesting stage in the evolution that digitization forced upon images. The use of ASCII characters to simulate a given image, primitive though it was, and while understood as an attempt at a virtuoso performance, nonetheless aimed (no matter how imperfectly realized) at verisimilitude. At this point, successful quality reproductions now possible, naturalistic image-making has taken off in new directions, attempting to explore the aesthetic potentials of the technology, which in this case also investigates the contradictions inherent in competing modes of representation.

Additions:
=>Reader Chris McKinstry (May 8, 2000) submitted an image of the Mona Lisa that includes a cross-section of the brain rendered tomographically -- the use of non-visual techniques, such as ultra-sound, magnetic resonance, x-rays, etc. to produce highly focused images at specified depths. These images can be combined by computer, to create apparent three-dimensional surveys of the subject -- generally limited to animate nature . Mr. McKinstry created this image as an illustration for a book under preparation -- "Hacking Consciousness, which has to do with the digitization of human consciousness." The picture relies on the Mona Lisa as a metaphor of the universal receptacle in which to report digital applications for imagery. More than a diagram (which could have served as an abstract vehicle for the author's purpose), the Mona Lisa adds a dimension of humanity, consciousness and recognition unavailable elsewhere. (Added 7/23/04 -- Image used with permission of its author.)

=>Readers of this page might be interested in Robert Silver's website or the gallery that represents him, through which a variety of posters, artworks and jigsaw puzzles may be purchased, including designs based on the Mona Lisa image shown here.

=>April 3, 2006. Discovery News, in an article by Rossella Lorenzi, reports that at a recent meeting at the Uffizi Gallery in Florence (Associated with the exhibit "The Mind of Leonardo -- The Universal Genius at Work,") a theory was put forward that claims that Leonardo technique in "consists of countless of dot layers applied with a technique of micro-divided brushstrokes." It continues by adding that "Jacques Franck, a consultant at the Armand Hammer Centre for Da Vinci Studies at the University of California, reported that the technique is somewhat similar to pointillism used by the French Neo-Impressionists in the late 19th century."
See further: http://dsc.discovery.com/news/briefs/20060403/monalisa_his.html  (added 4/8/06.)

Robert's Home Page | Mona Lisa Home Page | Mona List
 Previous Mona Image Page | Next Mona Image Page