How much creativity is there in HDR photography?

In the wave of discussion about artificial intelligence, some have prematurely hailed an August ruling by a US court as ‘groundbreaking’, when in fact it has only confirmed what has long been known in copyright law: only a human being can be a creator. However, in the debate over copyright to ‘works’ produced using artificial intelligence, it is easy to forget just how many ‘AI-created works' exist around us already.

A photograph can be a work

In 1888, when Kodak launched a camera for amateurs, advertising it with the slogan "You Press the Button, We Do the Rest", a revolution began in the availability of photography. Back then, the popular view was that a photographer was not a creator, but only the 'operator' of the device. In 1948, however, the Berne Convention, which laid out the basis for copyright protection internationally, designated photographs as works, although whether particular individual photographs qualify as a work is still being discussed today.

Not every photograph enjoys protection

Only human activity, and only that which is individual and creative in nature, results in a work. In photography, the photographer’s individuality may manifest itself in a number of choices: the moment of taking the photograph, the point of view, framing, lighting, depth-of-field, choice of focus and perspective, the use of special effects or procedures aimed at endowing the photograph with a specific character (this according to the Polish courts, e.g.,  in the judgement of the Court of Appeal in Lublin of 12.09.2016, case No. I ACa 942/15). If such elements bearing the stamp of the creator's personality are missing from a photograph, there is no work to which copyright protection can be granted. This is the case, for example, with passport photographs, whose parameters are strictly defined by law. A portrait photograph may be judged differently, as there are no specific requirements for it, and therefore the photographer has greater scope for creativity (see the judgment of the EUCJ in case C-145/10, Eva-Maria Painer).

Digital photography as analogue

Even before the ink on the new version of the Berne Convention was dry, pioneers of technology were already thinking about a new technology - digital photography. By the end of the 20th century, not only was the first fully digital camera on the market, but there was even a digital camera module integrated into a mobile phone, and probably no one gave much thought to the question of whether a photograph taken with a new digital camera can be a ‘work’. Digital photography very quickly replaced analogue photography (which is now considered a vintage process, like the daguerrotype). In the early days of the digital photography market, there was a race to increase the number of megapixels in a camera to increase the quality and precision of the images. At the same time, miniaturisation continued, and cameras became equipped with further advanced functions professional photographers rely on (control over exposure, depth-of-field, focus, and many others).

Software takes pictures

However, a limit was eventually reached, and the number of pixels was no longer key. Then, software entered the battle for customers. Today, what determines the quality of the photographs a camera produces is the software used by the producer. Photos can be significantly improved just by updating this software., and the Internet is full of tutorials on how to upload newer software to older camera models to enjoy better photo quality. Today, these programs know better than we do which photo is the "best". 

HDR is a snapshot...

The technique known today as 'HDR' already existed in the days of analogue photography. Gustave Le Gray was the first to try his hand at combining two analogue photos with different exposures. Today, every (or almost every) smartphone takes photos in HDR mode by default. When the user, having consciously chosen the moment of shooting, point of view, framing, lighting, depth, focus and perspective, finally presses the shutter button (or touches the screen in the corresponding place), the camera takes from several to a dozen shots, which the camera software then combines into a single image, selecting the best parts from each shot. The photographer has no control over what the algorithm chooses. All this for even better quality. So, who is the author of such a photo? The smiling user, who may not even know how advanced the technology behind his or her photo is - or the trained algorithm that does the post-production work for them and in fact chooses which photo is worth saving and showing to the world? How, then, is individual creativity apparent in a photograph over whose final form the photographer had significantly limited influence?

... but is an HDR photo a work?

Is it not the case, then, that under current legislation a photograph taken in this way cannot be considered a work? After all, the decision on the final appearance of the image is taken by artificial intelligence. The photographer, of course, has the option to take a photograph without using the HDR mode, thereby retaining full creative freedom. And it can be argued that even when taking an HDR photo we still determine many aspects of the photograph (i.e. frame, angle, perspective, moment). However, it cannot be denied that the scope of creative freedom here is limited – and in addition, not always in a way that is understandable to the photographer. This is because some of the work is done by AI.

The role of the algorithm in the creation of the piece

Software that works similarly to that used in HDR images has also been successfully used in the audio and video industry to enhance recordings without human intervention, and this has not been controversial. In the debate over whether the products made using generative AI can be considered works, the prevailing view is that, under current legislation, this is not possible. On the other hand, it is permissible to grant copyright protection ‘hybrid works’ created by a human being with the use of various AI tools (e.g., using Deep Dream) - provided, of course, that certain requirements are met with respect to the author's influence on the final form of the work. It seems that it should be no different regarding the products of generative AI, or at least those where generative AI is used as a tool in the hands of a creator - like the analogue camera used to be.

This is the English version of the article published on the Polish legal portal Prawo.pl on 9 September June 2023.