The art generated by this project has nothing original to it. The aim is to be faithful to the source, the original. The aim is also to eliminate guesswork and maintain accuracy and authenticity. A combination of strategies is used to realise these. To reconstruct, we rely on a number of visual, textual and narrative sources. To reconstruct the lost parts of the painting below, for example, we used visual references from two different sources. Source one: a sculpted replica of this painting from the Kailasanatha temple in Kanchipuram (shown beneath the reconstructed picture), which actually antedates the painting making it more appropriate to refer to the painting as the replica. Source two: a painting of Maya from Ajanta, which is, though similar, quite different also, particularly in respect of the position of the left arm and the angle at which the subject is painted.[1] [2]

To give accuracy and authenticity the best chance, the project involves hereditary artists who know the mythological themes and the grammar of Indian painting and iconography. Art historians assess the reconstructed art. We will also be presenting a direct, interactive comparison of the original and the reconstruction, as shown at the bottom of this page, for everyone to see and compare with the original the output that we generate.

It is also worth mentioning that the older photographs of the Devi mural at Panamalai, photographed in the 1950s by C. Nachiappan, now Koviloor Swamy, and published by Kalakshetra[3] show the mural to be just marginally, very marginally, better-preserved compared to its present state. No clues emerge from this picture as to how the left arm and the torso might have been.

It is also proposed that we involve more than one artist, independent of each other, to reconstruct the painting, the logic for this being that if different versions converge, then we most likely have the best reconstruction possible.

Above left: Devi, from Talagirisvarar Temple, Panamalai, Viluppuram, TN. Right: a digital tracing of the mural. In the tracing, the lost middle portion and the left hand are recovered through references to other sources, e.g., a sculpture of Devi (see below) in the same pose in the Kailasanatha temple at Kanchipuram; also from a reference to a painting of Mahamaya in a similar pose in Ajanta.

Devi sculpture

Above: A sculpture of Devi in the same pose as and in a mirror image of the Panamalai mural, in the Kailasanatha temple at Kanchipuram.
Picture courtesy of KT Gandhirajan.

Above left: Musicians in the procession that leads the palanquin of sītā on her journey to reunite with rāma, the scene as it is now seen in the Chengam temple. Right: The reconstructed look. Please note that the character on the extreme right who is half-seen in the composition has been omitted from the reconstruction. Consequently, the hidden arm of the drummer is imagined. The grey background is interpolated to mitigate the whiteburn when viewed on screen.

Ultimately when the drawings are replicated on other media, the colours will adjust and approximate to the nature and demands of the destination medium: on kalamkārī cloth, the colours will tend to those that are allowed by the resist-dye and direct painting processes of kalamkārī; in digital animation, a moving canvas, there will be constant interplay of light which will significantly alter the ‘original’ colours as seen on the mural, with base shades that approximate to the original.

One yardstick used for colouring, decolouring actually, is to pick the amount of ‘browning’ or dirt layering that is seen on what should originally have been a white canvas and assume that to be the overall extent of dirt accumulation and minus out that much of ‘brown’ from the canvas on a uniform basis in the process of filling colour.[*]

The methods of tracing and reconstruction have one great utility. It can serve as a great mock-up before physical conservation / restoration is undertaken. In fact, there is enough reason to argue for this to be a mandatory practice for anyone attempting to undertake any work with heritage property.  It must also be emphasised that "the reconstruction to the original colour palette as attempted in this project should not be seen as a finality, that it will always be an approximation and somewhat conjectural in principle. It should be understood as reconstruction to ‘as near the original colour palette as possible, based on physical reference of paint from the mural’." (IFA).


[*]It seems that one can be surer of the reconstruction of the lines than of the colours. The approximate results of the colour reconstruction is because of the absence of proper pigment analysis, or at least the public availability of the results of pigment analysis. For the foreseeable future, accuracy can be expected of the line work in all our efforts and a certain creative license may have to be provided for in analysing the colourisation of the drawings.