Don’t be fooled, how can you tell if an astro photo is real?
Whether you’re a space fan or not, amazing nightscapes are admired the world over. They connect us to the cosmos, inspire greatness, and encourage an interest in science and nature. Astrophotography is without a doubt one of the most technical and difficult forms of photography. Whether it’s wide-field landscape or deep-sky photography, it requires a good technical know-how, perseverance, the sense of adventure to explore unseen landscapes, a good knowledge of the sky, and many hours of image processing experience.
Producing a good astro image in no easy feat…unless you’re faking it! And sadly, many people do.
So how can you tell a fake from the real deal?
One of the most obvious ways to spot a fake is knowing the sky and the seasons. The southern and northern hemispheres don’t share the same targets. The appearance of the Milky Way is also seasonal. Most astrophotographers know when and where things appear, which makes spotting an out-of-place sky in a photo easy. For example, seeing the Southern Cross in a photo captured in the USA — it simply isn’t possible to see this high in the sky in the northern hemisphere, it barely scratches the horizon in the Florida keys!
But it doesn’t stop there, the Internet and social media is riddled with fakes, from the outrageous to the very well produced false images. Early in 2016, even NASA’s prestigious APOD (Astronomy Picture Of the Day) also fell prey to two well-produced submissions. (PhotographingSpace was also caught by this problem)
To get some expert opinion on the subject we reached out to astro-watchdog @FakeAstropix.
Q & A with @FakeAstropix
What do you look for to identify a fake astro image?
First and foremost – what do we mean by a ‘fake’ astro-image? The dictionary definition says a fake is something having a ‘false or misleading appearance’. To this we would add images that are misidentified (often deliberately), where no credit has been given to the author of the image, or where it has been presented as the work of someone else. Sometimes an image ticks more than one of these boxes, for instance this impressive piece of artwork by A4size-ska, which habitually does the rounds every time there is a solar eclipse. It is usually tweeted as having been taken from the International Space Station – misleading Twitter users and denying the artist some well deserved credit for his work!
We started the @FakeAstropix Twitter account back in 2013 because we were irritated by the sheer volume of obviously unreal astro-images that were being touted on social media. Many images get posted again and again – and again! Most of these tweets are from spambots, so there’s no point trying to contact them. However, we do try to reach out to real people, and verified accounts, to put them right; and if we can think of something amusing to post with a RT we will – although it can be a real struggle sometimes.
Why do certain images catch your attention more so than others? Is there a tell-tale sign or something significantly different in certain posts?
The phrase “my friend took this” accompanied by an awesome image is a good indicator! Who needs friends who can’t even include your name in their tweet?
Occasionally an image just won’t ring true, it might be too perfect or seem unreal. But most of the time it’s a matter of testing what we know about the image to find out if it’s genuine. For example, we can use a planetarium program to look at the sky at the same time and location as the image to see if the star background matches up. In this case the Milky Way could not have appeared behind the lighthouse from the photographer’s direction: https://twitter.com/fakeastropix/status/425067592811028480
Perspective is also a good giveaway – and can be a good education opportunity as well. We had to work out the size of the Chrysler Building Crown, and how far away from the Empire State Building it was, to calculate by how much the Moon’s apparent size had been increased in this image: https://twitter.com/fakeastropix/status/448781577213206528
Sadly, media hype about ‘super’ moons seems to have made it almost obligatory to cut & paste an oversized Moon into an image for effect. This can sometimes lead to amusing results, for instance when someone pasted an image of the lunar farside onto his terrestrial foreground picture!
The timing can also be a clue – a widely reported fireball will bring a lot of old fireball images and the odd ATV re-entry out of the woodwork for us to have a go at identifying.
Not only is New Year a time for new resolutions, it is also a time for outing some familiar old pictures. Both of the following images have regularly been described as Europe at Midnight on New Year’s Day, ignoring the fact that they cover several different time zones!
Increase in light pollution 1993 to 2003: http://www.sciencephoto.com/media/160004/view
Map of Flickr & tweet locations: https://www.flickr.com/photos/walkingsf/5912946760
Perspective alone makes both views impossible in reality and, if the second image really was Europe at midnight, the Alps would be horrendously light polluted!
What’s your take on composites?
Many modern images are composites – multiple frames stacked, combined and processed in Photoshop or equivalent packages. Ruling out composites would rule out most of today’s planetary and deep-sky imaging. The same applies to Photoshop – just because an image has been Photoshopped doesn’t mean it has been ‘faked’. It all depends on how it’s been applied.
So, at what point does a digitally enhanced image become a ‘fake’? We would argue that it comes down to an intention to deceive; that is, to present an altered view as if it were ‘real’ or unaltered. Ideally the creator of the image should explain how it was produced – although we appreciate that many professional imagers like to guard their techniques. But as long as they are open and honest about what has been enhanced, and in what way, then we have no problem with this. The problem usually comes when an image is stolen by spam feeds and circulated as something it’s not.
Some people have labelled it a fake, but the photographer has been completely honest about the mathematical techniques he used to produce it – and even wrote some of the software himself. The grey area is really about whether you call it an astro-image, or a work of art. In our view – almost uniquely – it’s both!
Can you list some of the worst fakes you’ve ever seen?
One of the images that started us off was this one, tweeted as the Northern Lights in Alaska
It took us a while to realise that the ‘aurora’ was a hugely over-processed Hubble image of M42 behind some CGI mountains. It still surfaces on a regular basis and is like an old friend! A nice GIF showing the transition between images can be seen here.
The star trails in the lake is a favourite. The reflection has a touch of Van Gogh about it!
This image is totally bizarre; an inverted fisheye view of the horizon being used to represent the Earth from orbit!
We also have a soft spot for a certain globe-trotting meteor.
The mosque image is interesting as we had to figure out where it is and check that the light pollution is epic. We were also amused by the fisheye distortion to building but not to the sky!
Have you ever encountered a fake which turned out to be real?
Sometimes we’re suspicious of an image, but research fails to prove conclusively that it’s a fake so we give it the benefit of the doubt. Occasionally we are happy to be proved wrong, as with this image of an annular eclipse by Colleen Pinski.
How do you find these astro fakes?
There are a lot of them about, so it’s not difficult. We have a list of ‘sources’ which includes a number of ‘spam’ twitter feeds some of whom attract hundreds of thousands of followers by tweeting pretty pictures, interspersed with advertising. We also search on stock phrases – for instance ‘Eclipse from the ISS’ – to track down image subjects that are doing the rounds.
Quite often our followers alert us to suspect images, which is invaluable – and we are not alone. There are a number of other twitter feeds, and several blog sites, that are dedicated to debunking all sorts of fakery. @picpedant and @badastronomer are very good at identifying fakes and great people to follow on twitter.
What methods of verification do you use to ensure that the accusations are correct?
We try not to accuse anyone – we just point out any errors and inconsistencies in as interesting and entertaining a way as possible.
Google is a fabulous boon to the modern astro-image sleuth! The reverse image search can help to track down original images, and Google Earth can pinpoint locations and help us work out where an image was taken from. As mentioned earlier, free planetarium programs like Stellarium can reproduce the night sky at a specific time and location, and help us pick out the star patterns in pictures.
The rest of it is down to experience and common sense!
P.S.: We also have a network of very knowledgeable people who we can ask if we get stuck!
What’s your take on the 2 recent APOD fakes? Why is APOD getting inundated with fakes?
Until a couple of months ago we would have regarded APOD almost as a guarantee of authenticity; if we could find the image and description on APOD – job done!
Recent events have reminded us that APOD is actually run by two guys in their ‘spare’ time. Although very knowledgeable, they almost certainly don’t have time to check accuracy of every image submitted to them in depth, and the website makes it clear that you use the information on it at your own risk.
Gradually more & more people – professional photographers hoping to sell images and amateur imagers who just want the attention – see sites like APOD as a good way to publicise their work. It’s inevitable that some less scrupulous imagers will submit dodgy pictures that occasionally slip through the net. However, the site has operated for over 20 years with very few issues, and it would be a great pity if they were to stop using amateur images because of the risk of being taken in by a few plausible fakes.
What advice can you give to photographers to help protect their images?
Be careful what you post – and where. Put a watermark or credit on the image itself, and always upload low resolution images, never raw or TIFF files.
If you have a Flickr page you can disable downloads – and be vigilant. We recently spotted this image by @Dewbow from 2012.
@HighTechPanda didn’t only plagiarise the picture but also his text! Talk about lazy!