Authentic Photography
A Journalist’s Diary
Photography is at risk. AI improvements elevate smartphone photography, post-processing, and image generators. Dis- and misinformation through visual documents get easier to spread and more difficult to spot.
That is a problem for photography and journalism.
But it’s also a problem for all of us. We all must care about how photography evolves and how we manage new forms of photography and new types of images. Today, I’ll try to explain why a raised awareness of the technological side of photography is important.
First, I’ll define what a photograph means for me and look at journalism’s rules for photography. Then, I’ll add a few thoughts about modern smartphone cameras and AI editing/generators. Based on that, I’ll write about new technology that might help journalists produce authentic visuals that the public can trust with (more) confidence.
What is a photograph?
One picture. One moment. That’s what a photograph means to me.
The photographer witnesses a moment and captures it. The shutter freezes that moment, and at the end of the process stands an image, an accurate and authentic representation of life. Accurate and authentic as in what is in the image is what the photographer saw and how they decided to compose it.
There are many ways in which this traditional idea of photography can still misrepresent or tell a different story to the truth, of course.
As Kenneth Kobré writes in his guide “Photojournalism: The Professionals’ Approach” on the idea of taking a print to the location of capture and comparing it with what you see: no finished image will ever be absolutely “identical in color, tone, brightness, contrast, etc.”
More so, the framing can include or exclude elements vital to an accurate, unbiased representation of reality.
Editing includes many more pitfalls. More on that below.
But despite those caveats to authenticity, it’s that understanding of photography that builds the foundation for humanity’s trust in photography as a tool to record history and tell accurate stories.
Photography in journalism and the use of smartphones
I wrote about this part in October 2022 already. Instead of repeating or rewriting that post, I’ll leave you with the bottom line, a link to that post, and one of the photojournalism ethics guidelines.
Journalism is an industry without one overarching guarding institution. There’s not that one universally accepted standard of how journalism should be practiced. Nor is there one certificate that proves journalists are well-versed in ethically sound reporting.
The path to journalism can look different from person to person and from country to country. Not to mention the added complexities of the digital age that allows anyone to publish anything.
However, several industry organizations offer codes of ethics that agree on a framework within which the human element in journalism is supposed to operate. Those guidelines include personal conduct, i.e., what is allowed or expected in reporting practices. For photographers and editors, it also delineates the most important aspects of the photographic process, from gaining access and photographing to managing relationships and editing.
When you peruse educational staples, you’ll find that cropping, for example, is accepted. But be aware of the potential to change a picture’s meaning.
When it comes to special effects, they always require an explanation. Ken Kobré’s guide demonstrates that effectively through an image of a farmer next to a large watermelon. The watermelon appeared taller than the human due to a wide-angle lens, a technique that should be explained in the caption. We had a recent example of such an occurrence when President Biden and his wife met former President Carter and his wife.
Kobré’s section on ethics describes an incident when a photographer darkened the background behind two firefighters attending the funeral of a colleague to emphasize the scene. The picture won an award, which was eventually stripped because a later investigation found the photographer not guilty of trying to deceive but left those in charge “uncomfortable” because he had gone “over the line in the use of some techniques.”
The publication of a composite in The Washington Post in 2012 (a sunset scene, which was generated by merging multiple exposures for more dynamic range than the camera could capture in a single file) shows the problems of defining the limits of what journalism should do clearly.
The then President of the NPPA said that HDR, just like any other form of digital manipulation, combined different moments and created an image that does not exist.
I recall reading something similar about panorama stitching, for example.
Note that composites can still find their way into journalism through illustrations, which are visuals that don’t depict reality and are easy to identify as such. They are crafted to visualize a concept or abstract idea. But that’s probably a standalone topic.
What we need to remember: composites are a step too far for photojournalism.
The impact of smartphones on photography
I’ll come back to that last point in a moment. But first, a few lines on smartphones and their camera functionality.
Smartphone photography is more popular than ever before. The image quality phones deliver is stunningly good. It’s the only camera in the lives of many people. And even news agencies celebrate the use of phones as tools of the trade.
I believe that we are entering difficult terrain in that last regard.
I’m not debating the image quality or whether a phone is good enough. I care about the distinction between photography and composite photography.
Whether someone accepts composites for personal images and memories is one question. But if “handmade” composites are too much for journalists, how can automatically created composites be okay?
And composites are all you get with modern phones.
Photonic Engine, Deep Fusion, HEIF Max, RAW Max, Night Mode, HDR. The list of technobabble grows longer with each new generation of phones.
I’m several phone generations behind at the moment but like to stay in the loop, and I like to think I’m somewhat technology-minded, at least, even if I can observe myself being less and less inspired or interested in the annual technology race.
But even reading a well-written explainer on the latest iPhone camera technology left me unsure of what exactly the device does.
The writer makes one thing clear: no matter what you do, iPhone hands you a composite, an amalgamation of various images that were optimized for a specific technical parameter and then assembled for the best possible image quality.
Take the resolution as an example. 12 MP, 24 MP, or 48 MP?
Explaining how the standard 24 MP files are created, the article explains that the Photonic Engine “captures multiple versions of a scene” and “then compresses those images into a single layer.” In an additional (computational) step that “optimized result” is combined with a “48 MP image to add detail, resulting in the default 24 MP image.”
All clear?
The bottom line: smartphones deliver composites, or put differently, a manipulated image. That’s why the image quality is so convincing.
And the quality is astonishing. It’s terrific what phones can do these days.
But computational photography, or composite photography, or whatever name you want to apply, is a genre in itself. It’s not photography.
Some might think the difference is philosophical.
For personal use, it may be (I still believe people should know the difference and then make an informed decision if they want to rely on that).
But the question remains where this is going for journalism. Can journalism afford automated composites when “manual” composites are considered “too much?” Can journalism afford that layer of non-transparency when trust in journalism is already in acute danger?
Other AI impacts
AI’s influence in photography doesn’t stop with smartphones, of course. AI culling, AI keywording, AI toning, generative fill, and other editing tools are continuously being introduced and refined.
And photographers seem to love that AI is ready to take over the laborious post-processing work, as any stop on photography sites or video channels suggests. A recent look around the forums of one popular editing tool for anyone who’s dealing with large collections and/or quick delivery times, Photo Mechanic, includes impatient queries for AI inclusion.
I could add my two cents to all of that, but I also don’t want to make this longer than it is already. The topic is vast and has many tangents one can get lost in debating.
Maybe just this: while culling, toning, and the writing of metadata (such as keywords) are indeed arduous tasks, they are also necessary parts of photography, especially journalistic photography.
Handing culling duties over to an AI reduces the learning effects and introduces an unknown into the process of selecting images for a story. The former may “only” limit the photographer’s growth, but the latter directly impacts the quality of journalism and trust in journalism. Editing has its own ethical aspects that are at risk when an AI is allowed to pick images.
Can you spot the fake?
Finally, AI image generators are becoming more powerful, leading to results that are difficult to spot, even for people who know what to look for.
That makes spreading dis- and misinformation through generated images dangerously easy and difficult to spot.
Here’s a look at the use of AI in spreading wrong information about the latest crisis in the Middle East, here is a poignant piece on AI’s impact on “photographic truth,” and here is a discussion about the challenges for fact-checking images in times of AI generators.
The Value of Trust & Authenticity
With all that in mind, it’s not hard to see why people lose trust in journalism and news photography.
The more technology advances, the more important it becomes that journalism is transparent, knows how it uses new technologies, and owns honest mistakes (which we also need to learn to differentiate from “fake,” by the way).
At the end of the day, technology is not ethical or unethical. Humans are.
A year ago, I shared news about an initiative focussing on new technologies to help responsible human behavior.
The Content Authenticity Initiative was a good start to ensure we can tell authentic visuals from altered ones.
Now, Adobe and Leica have made the next step. And it could be a big step forward for photography if more brands incorporate it and if the remaining kinks get cleaned up.
You can find a detailed explanation in this video, which starts at the appropriate moment (the first 13 minutes deal with other Leica-related news).
In essence, Leica recently announced a new camera, which includes a chip that attaches a certificate to each file. That certificate allows anyone to see where the file comes from and what happened with it between the moment of capture and the moment it was published. The history of what happened to the file includes post-processing steps, including AI tools. And if, at any point, someone tempers with the file, the certificate becomes invalid.
I love the direction and hope all camera manufacturers implement that technology soon, for it adds a security layer that makes trusting photographs easier again.
I hope you found value in today’s post. While trust in visuals may be a question for journalism and how photography changes may be seen as a philosophical question, I do believe protecting photography is a task we all should care about.
Smartphones and computational photography have their place. But it’s us photographers who should be aware of the technology and how we want to use it. And maybe the companies driving these technologies just need to be animated to do a better job of taking us along for the ride instead of dazzling us with technobabble. Photography is an essential part of the human experience, a valuable tool to record history and express ourselves. Both are qualities of higher importance than devices or the technology that makes photography possible.
