Monday 30 May 2011

Richard Mosse and Aerochrome

I was away for a couple of days so it was only today (Monday) when I opened the Guardian newspaper Saturday magazine to find some interesting colour infrared photographs taken in the Congo by Richard Mosse. It's a (probably) unique application of infrared photography to reportage. He was unaware of it but he was taking these photographs almost exactly a century after the first infrared photograph was published early last year.

The best place to explore the photographs is his web site (natch) where you follow links either to the project itself (called Infra) or to the text/press page.

He's interested in the aesthetic of the false-colour images, having seen it on a Grateful Dead album cover (and elsewhere ... more on IR rock photos here). The idea is to make us take a second look at things when using this film. It's very difficult to get good results with the film for a whole bunch of reasons, and I am impressed technically with what Mosse has achieved. Partly this is because he worked in medium format, using stock cut down from the 70mm Aerochrome rolls, which provides a sharpness the 35mm Ektachrome rarely achieved.

The stock, KODAK AEROCHROME III Infrared Film 1443, is now discontinued, which would make it the last mainstream infrared film. Apart from some obscure Soviet Russian film, Kodak seem to have been the only company to experiment with infrared and colour-shifting. In case you are unaware, Infrared Ektachrome and Aerochrome (essentially the same film) have three sensitive layers, one each for near infrared, red and green although all layers are sensitive to blue which must be filtered out. A yellow (minus-blue) filter was the 'correct' filter to use with this but may people also used orange and other filters to play with the resulting palette of colours. The dyes in the film were coupled in such a way that infrared in the original scene was rendered as red, red as green and green as blue.

It was originally produced as camouflage detection film. The first incarnation appeared in 1942 in the form of 'Aero Kodacolor Reversal Film, Camouflage Detection'. The idea of combining infrared and panchromatic emulsions with colour processing (to produce a false colour result) had originated in Kodak's laboratory in the UK in 1941. The plan was to have two layers which were sensitive each side of 700 nm, where the Wood effect kicked in. Early tests of the new film, used in conjunction with conventional colour film, were positive although the film was considered to be too slow. By 1945, a faster version of the film had been standardised and was recommended for operational use in the Pacific at the close of the second world war.

Eventually, the film became widely used to remotely detect the health of foliage, even from space. Since healthy foliage reflects a lot of near infrared (due to the Wood Effect) the relative amounts of green and infrared that a plant reflects are an indicator of its health. With this infrared film the colour shifts from magenta to red as the foliage gets healthier. Camouflage detection is a subset of this since green paint and dead leaves could easily be distinguished from the real thing. That's until the 'enemy' started making special infrared camouflage paint.

So one thing that strikes me about Richard Mosse's photos from Congo are how magenta the landscape is: stressed as the botanists would say. Healthy foliage would be bright red but there is little of this to be seen, although there is some.

According to an interview in the BJP, which is reprinted in the text/press page of Mosse's web site, he's going to experiment with infrared movies, shot with a Red One without its infrared blocking filter. This can be easily done, as I discussed this very process with Red at IBC about four years ago. There may be issues with any colour-splitting filters in the light path (splitting RG and B to the three sensors) since the dyes in the filters will have their own variable amounts of infrared response. Richard says that "It will be a kind of blown-out monochrome palette", apparently expecting it to look like a night scope. I would expect it to look more like the usual output of a DSLR with an infrared filter in front of it but those pesky colour-splitting filters are a bit of a wild card. I did an experiment with Ikegami for last years' IBC and while the monochrome infrared output was poor (because of the splitter filters) we did get a good result using a minus-blue filter. I've written about this approach before, when I went to the BBC Natural History Unit and we tried a minus-blue filter on a modified DSLR to emulate infrared Ektachrome.

So it may be that, with a bit of suitable channel mixing, Richard Mosse's infrared movies will continue the look of his aerochrome stills. I hope we'll see the results.