The Method
Permission to access the death mask was granted by both the National Portrait Gallery and the Tate Britain (which houses the Turner Bequest and owns the rights to the death mask). Once access was granted, a subsequent date and time were arranged to photograph and laser scan the death mask. The death mask, created in 1851 by Thomas Woolner, was fully intact and in good condition. The death mask was positioned upright and photographed, with a special attempt at recreating the angle of the sitter in the Varley portrait. Following this, the 3D laser scanning was accomplished using the Polhemus FastSCAN Scorpion Handheld 3D laser scanner. This scanner operates by passing a fan of laser light over an object in ‘sweeps’ (in the same manner as using spray paint). The image then instantly appears on the computer as each sweep is registered. The generated sweeps can overlap to ensure minimal gaps in the virtual death mask, of which 4 were created and the one with the least background ‘noise’ (usually caused by metal artefacts) was selected.
The superimposition
Photographic 2D comparison
Of the photographs taken of the death mask at the gallery, one was selected which portrayed the death mask the most similar angle and orientation to the Varley portrait. The death mask was selected from it's background using Photoshop. This enabled a direct visual comparison between mask and portrait, utilising Photoshop’s functions for layering and altering the transparency of images (see gallery for stills of this process). Certain facial features were used as points of reference in the scaling and angling of one image over the other. These were as follows: * The nose was utilised in the establishment of the correct camera angle and scale, examining the length and width of the nasal aperture * The nostrils (size and position) were used to determine the correct degree of lateral tilt in the death mask * The position of the external ear in relation to the nose was important for determining the correct angles * The distance between the nasion and subnasale aids in scale * The outer canthus of the eye provides the proper degree of rotation to the death mask and enables the precise alignment of features. For an accurate photographic superimposition, the angle of the camera shot, the distance of the object from the camera and the degree of facial rotation need to be as similar as possible. I generated feature-based comparison tables as an aid in the visual comparison process. This feature-based comparison generated results that were not supportive of the Varley portrait’s subject being JMW Turner. I also undertook a computer montage comparison, looking for correspondences between the shapes and proportions of the soft tissue by vertically dividing portions of the face and examining the details along the join line of the two juxtaposed halves. This showed a better correlation between the two images, but the obvious difference in jaw line was troubling. Finally, a grid/line comparison was undertaken using Adobe Illustrator software. Rectangles of varying colours were created around facial features. The related facial features from the Varley portrait to the death mask were connect by lines drawn down the vertical and horizontal, forming a grid of spatial and proportional comparison. The results of this comparison were inconclusive, as the grid analysis made obvious the differences in facial angel and rotation. Computer aided 3D comparison FreeForm is a 3D modelling system which operates by the use of ‘haptic feedback’; the virtual clay in the 3D interface generates resistance (forced-feedback) which provides the operator with a sense of touch, enabling the modelling of virtual clay with the use of the Phantom Desktop or Phantom Omni devices. The scanned 3D death mask image was opened in FreeForm and the Varley portrait was scanned and fixed onto a flat 2D ‘plane’. The plane can be moved with the Phantom devices so that it is orientated and aligned precisely in front of the death mask. A morphological comparison is then made. Observing historical records, it becomes apparent that Turner had been suffering from a long and severe illness. The disease process itself would have altered the soft tissues of the face, shrinking the fatty tissues and changing their natural depositions. This is especially obvious in the optical region where the eyes appear deep-set, with deep furrows present in the orbital margin and at the base of the eyelid. The shape of the external ear on the death mask was remarkably similar to the ear on the Varley, especially with regards to the shape of the choncal fossa. Although the orientation and angle of the ear showed minor discrepancies, as well as the definition of the outer helix on the Varley portrait being more pronounced, the general morphology is rather similar. The death mask presented myself with pathology of the soft tissue in the form of small lumps (two of these lumps were visible, anterior to the left ear and side burns). There was a suspicious lack of shading in the same region on the Varley portrait, which could have caused the sideburn hair not to grow at this location (a common occurrence with moles and other such marks). To be certain of this would present a considerable contribution towards a positive identification. The orbits aligned precisely. It was also apparent that the distances between the outer canthi of the eye and the tragus were in exact accordance. The levels, positioning, size and shape (both of the eye itself and of the superior and inferior rim) correlated well. The eyebrows (although not clearly delineated on the death mask) had a similar shape. The mouth and jaw on the Varley portrait exhibited obvious differences with the death mask. Large differences in age, health and dentition would account for this, as well as the positioning of the mouth when the death mask was created. Notwithstanding, one image is of a healthy young man while the other is of a cadaver. The jaw on the Varley is much fleshier and robust, unlike that on the death mask, which is thin and lacking dentition. This lack of dentition would result in bone resorbtion, reducing the height of the mandible and maxilla. These facial features are therefore incomparable. The only comparable morphology was the position of the mouth in relation to the subnasale (the point at which the nasal septum and upper lip meet). This distance appears similar in both images. The dimensions of the forehead showed no comparable discrepancies. The only noticeable difference between the two foreheads was found to be the liens of facial expression: the lines were more pronounced on the Varley. An explanation for their apparent absence on the death mask could be that, during its creation, the weight of plaster pushed back the skin and smoothed away many of the lines and wrinkles on Turner’s face (a pseudo facelift). The nose was the only feature in the upper half of the face that showed discrepancies. The features of the nose that matched well included the base of the nose when examining the left nostril and the curve of the alar. Those that didn’t included: * The nasal septum, itself more pronounced and superiorly projected on the Varley * A more pronounced undulation of the bridge present on the death mask * The nasal tip on the Varley was shorter and up-turned * The nose on the death mask showed a bifid tip while no such trait was present on the Varley Some of these differences can be accounted for: the nasal tip continues growing with age; the weight of the plaster during the death masks creation may have compressed the nasal tip. However, the lack of a pronounced bifid nasal tip is of concern; such a facial feature does not disappear with age. Yet there does appear to be some shading in this region on the Varley. It is possible that the sitter was placed in direct sunlight, obscuring this trait, or that the artist didn’t emphasise this trait... as it was considered aesthetically displeasing to either the artist of the sitter. Following the completion of a morphological assessment (and having taken into account the affects of age, ill health, hairloss, tooth loss and eventual post-mortem effects on the face of Turner), I conclude that many of the morphological features were found to be in agreement and that discrepancies could be logically accounted for (with the exception of the conspicuously absent bifid nose). On the basis of these findings, I conclude that it is probable that the Varley portrait and Turner’s death mask represent one and the same person.
