Problems with photos of the heavens

Around 200 photographs on average are used in the Guardian every day, and a similar number published daily on the website. If one of them is of a star, or stars, it all too often leads to problems.

On 1 February we published a photograph that was captioned “A meteor storm over Stonehenge”. A chorus of readers said: “Oh no it isn’t.” A correction was published on 3 February: “A photograph used to illustrate an article about stargazing was wrongly captioned as showing ‘a meteor storm over Stonehenge’. In fact it was a photograph of stars taken with a long exposure which, due to the Earth’s rotation, produces the effect of trails (The best places to see stars in Britain, 1 February, page 3, G2).”

Accurate insofar as it went, but there were even more problems with the picture than that. As one of our less angry readers put it: “Not for the first time, the Guardian has got a star picture quite wrong. That was not a meteor storm. It was a simple time exposure, showing the tracks of stars over a period of 10 minutes or so. And it was a composite construct, an artificial picture rather than a true one.

“Those stars are not in the right places for a British night sky. They appear to have been taken from a much lower latitude, and the Stonehenge image then merged … From the aspect of the sarsens [silicified sandstone boulders of a kind used in the construction of Stonehenge], it’s clear that this agency photo was taken looking south-west.

“With the help of colleagues from Navlist, a mailing list devoted to traditional navigation, the star constellations shown can now be identified. These relate to the picture shown in the online version on 31 January, rather than the printed version shown on 1 February, because in the latter the picture has been substantially cropped, making identification more difficult. But the stars in the two are the same.

“All the stars shown are members of northern constellations. In real life, they wheel around a common centre, near to the pole star, high in the northern sky at 51° altitude. However, the picture in the Guardian shows those stars wheeling around a point well below the horizon, in the southern sky. That would be quite impossible.”

He concludes by asking whether the Guardian still has a policy of not using photomontages unless clearly labelled. The answer is yes (and we have a policy of not “flipping” pictures, ie reversing them, although our use, twice within recent days, of a flipped agency picture of John Wayne in the role of Rooster Cogburn with his eye patch over the right instead of the left eye, where it should have been, might suggest otherwise).

I have quoted extensively from the letter because the reader sums up the situation quite accurately, as this email from the photographer to the agency that supplied the digital composite makes clear: “I picked a star trail exposure from my files, so that ‘clear skies’ could be illustrated above the (Stonehenge) monument. I usually use northern hemisphere stars to get the more circular concentric stars around Polaris, but the higher elevation in the sky with less curvature worked better with the telescopic view of the comet (separate exposure). This particular sky image is possibly at high elevation looking south.”

A second email gives more detail: “I located my original star trail exposure, and it does look as though the camera was facing NE with the constellation of Cassiopeia near left of centre. However, the photo was made in a vertical format and I shifted it to horizontal to compose the Stonehenge image. This seemed to work best for the overall composition.”

The agency has conceded that the photograph should have been labelled as a digital composite and will now go through its files to amend all similar picture captions clearly.

It should have been spotted at the Guardian, but I have a great deal of sympathy with a picture desk that nearly a decade ago – on the day of the 9/11 terrorist attacks, to be exact – received just 2,000 images a day into their computer system. Now every day they receive 15,000 to 20,000 into that system. A highly specialised digital composite that is not labelled as such is difficult to spot.

This column was originally published in The Guardian on Feb. 14, 2011.

Comments are closed.