Mainmenu

Innovative software reveals unique patterns in the work of Vermeer

On 10 November, the RKD ─ Netherlands Institute for Art History will be launching the online publication Counting Vermeer, edited by Professor C. Richard Johnson Jr. from Cornell University. This book is the first to present the results of innovative and multidisciplinary research into all of the canvases used by one artist, in this instance Johannes Vermeer (1632-1675). Counting Vermeer demonstrates, for example, how computer-generated examination of the fabric support confirms that Vermeer is indeed the author of Young woman seated at a virginals (c. 1670-1672), an attribution that used to be regarded as doubtful. Advanced analysis of the canvases also supports the theory that Vermeer conceived at least two pairs of paintings as pendants, a fact that has been fiercely contested by art historians in the past.

In collaboration with an international team of scientists, curators and conservators, Professor Johnson developed software that can perform automated analysis of the threads in painting canvases, so that unique weave patterns, comparable to ‘finger prints’, can be recognised. In the past, thread counting was a purely manual process and based on samples taken from various parts of the canvas. In Counting Vermeer, prominent researchers consider how this new method can be applied in the analysis of the complete oeuvre of Vermeer. A significant addition to the research are the so-called weave maps, which allow us to ascertain whether the canvas supports of different paintings were cut from the same bolt of canvas. Weave maps are therefore an extremely useful tool for assessing questions of dating, pendant relationships, and authenticity.

Unique software now available

It has so far not been possible for others to use this tool. Via Counting Vermeer, on the RKD website, Johnson and his team are making the recently developed software available through open access for the first time. From now on, researchers will be able to apply this analysis technique themselves to paintings of their choice. As more people resort to this method and the results are shared, our chances of making new discoveries – such as establishing that two canvas supports used by different artists originate from the same roll – will increase. Use of the software on a wider scale will mean that the art-historical world can enter a new era of technical analysis through data sharing.

According to Professor Chris Stolwijk, general director of the RKD: ‘The software developed by Johnson and his colleagues to analyse canvases through automated thread counting which then identifies weave patterns has proven to be a highly innovative development for the examination of canvases of old as well as modern masters. Replacing the time-consuming, traditional method of thread counting by hand – a technique that is less reliable – automated thread counting allows us to create large, solid data files opening the door to new research into artists’ practices, the production of artists’ canvases as well as the authenticity of paintings.’