3 Responses »

With a little help from the computer…

Computers are a big aid, also for the conservator. Roughly five years ago, the Van Gogh Museum entered a new and somewhat unusual research partnership with experts in the field of  Electrical and Computer Engineering. Our collaborative goal was to develop new computer tools for the study of artists’ canvases. This initiative developed into the ‘Automated Thread Count Project’, co-directed by Professors  Richard C. R. Johnson, Jr. and  Don H. Johnson, from Cornell and Rice Universities in the USA. Application of this tool has led to fascinating new insights regarding Van Gogh’s choice of canvases in relation to other painters of his day, as this one example shows.

Counting the threads 

Simply put, the computer is able to precisely count the number of threads that run near-vertically and horizontally across each square centimetre of a painting canvas. Usually the automated thread counting is performed on a digital X-ray image, which reveals the canvas weave structure of a painting.

Detail of an x-ray of a painting revealing the canvas weave structure, with overlaid counting grid.

Detail of an x-ray of a painting revealing the canvas weave structure, with overlaid counting grid.

Weave maps

Slight fluctuations in the spacing of the threads across a piece of canvas emerge, a natural consequence of the weaving process. These variations appear as colour-coded stripes in ‘weave maps’, forming a distinctive pattern for each canvas. When the pattern of stripes in the weave map of one painting exactly aligns with the pattern of stripes in the weave map of another painting, we can conclude a ‘weave match’ i.e. that the two canvases were cut in alignment from the same piece of woven cloth.

A match

We were quite excited when the computer turned up a weave match between two pictures painted in Paris in 1887, one by Van Gogh and the other by Emile Bernard, his younger friend and colleague in the period.

Reconstruction of how the two painting canvases were cut from the roll. Emile Bernard’s Portrait of his Grandmother was turned 90˚ in alignment with Vincent van Gogh’s Plaster cast of a woman’s torso.

Reconstruction of how the two painting canvases were cut from the roll. Emile Bernard’s Portrait of his Grandmother was turned 90˚ in alignment with Vincent van Gogh’s Plaster cast of a woman’s torso.

Weave map illustrating a match in the pattern of warp threads running through both canvases down the length of the roll.

Weave map illustrating a match in the pattern of warp threads running through both canvases down the length of the roll.

An intimate link

It appeared that the canvases used by both painters had been cut from a single piece, not by the painters themselves, but by a commercial manufacturer. The reverse of both pictures is now covered up by an added support canvas. Yet a figure ‘6’ stamped onto the back of  the original canvas used for Plaster cast of a woman’s torso shines through the backing canvas. This confirms that the canvas was commercially prepared and sold ready-stretched on a standard-sized frame, in this case corresponding to a Landscape 6 size ( 41 x 27 cm).

So it is likely that Van Gogh and Bernard visited the same artist supplies shop in Montmartre to purchase their ready-made canvases. Surely one cannot imagine a more intimate link between the practice of two painters than a shared shopping venue.

 

3 Responses to “With a little help from the computer…”

  1. Philippe Renette says:

    How fascinating a life Vincent (and other artists in Paris and around) had in those days!

    Your piece was very informative – a lot has been going on recently around Van Gogh – the new version of his letters in a glorious series of six books, the idea of a portrait of his brother and so on.

    Am a fan of Vincent and would advise everyone loving him or interested in him to travel to Auvers-sur-Oise, 25 miles north of Paris where he lived the final 70 days of his life. Only there in the little room of the Auberge Ravoux do you get to know him, to feel him, to understand him and to realize who he was and is.

  2. terence kelly says:

    The study of Van Gogh both as a man engrossing in his own right and an artist of primary importance to the history of the development of post-impression art rightly never ceases to enthral art lovers; the combination of his vivid colours and the spontaneity of his brushwork required enormous courage at the time and continues a century later to amaze on close study. In this of course he shares a close affinity with his illustrious 17th century Dutch predecessor Rembrandt.

    On a more practical point: we depend upon Vincent’s letters to definitively authenticate and date many of his pictures – the letters often contain very specific detailed descriptions of works he is either engaged upon or has just completed. The details of a picture Vincent was `absorbed in’ contained in his last letter to his mother -undated but late July, 1890- are equally specific yet do not match any found in pictures comprising his established oeuvre thus indicating this picture in all likelihood is missing – yet apart from a tentative reference by Dr Jan Hulsker in his Catalogue Raisonne `The Complete Van Gogh’ 1977 explaining his difficulty reconciling this description with any of the known pictures in Van Gogh’s oeuvre an odd silence has enveloped the whereabouts of this picture: why should that be?

    My reason for raising this strange conundrum is that I’m quite convinced I know the whereabouts of this picture – it is good for these descriptive details (and goes further since Vincent was still engaged painting it), and for period, palette and style despite the letter containing the description not being published by the Van Gogh family until 1953 making a fake practically impossible. No physical examination of this picture by experts has ever taken place.

    Just in passing, an interestingly point I’ve never see referred to in the literature, apart -as mentioned above- from buying framed canvases from art material dealers when his meagre finances permitted Vincent in his determination to continue his creative endeavour more generally was obliged to buy rolls of canvas of varying quality -including on one occasion we know of painting on a kitchen cloth while as mentioned not infrequently reusing canvases – and there is good reason to believe he employed another form of support than the final stretcher -a flat board perhaps?- when painting out of doors since these pictures frequently extend around the stretcher by a inch or more.

    In all respects I find the Van Gogh Museum a place one simply must visit when in Amsterdam.

    • Jolein van Kregten says:

      Dear Terence Kelly,

      Thank you for your reaction to the blog on automated thread count and in particular your observations on Van Gogh’s use of alternatives to ready-stretched artists’ canvas. Indeed he is known to have painted several works on French household cloth, known as torchon. An example on torchon cloth patterned with red stripes is in our museum, The Garden of Daubigny (June 1890). For Wheatfields in a mountainous landscape (late November-early December 1889), now in the Kröller-Müller Museum collection in Otterlo, he used a red-checquered cloth instead. These are just two examples illustrating the diversity of the picture supports used by the artist, which will form a theme within the 2013 Studio Practice exhibition in our museum. On the topic of Van Gogh’s reused canvases, see the current exhibition.

      kind regards,

      Ella Hendriks
      senior conservator Van Gogh Museum