Is editing photos cheating?

Editing isn't new. Darkroom printers were doing it decades before Lightroom existed. What matters is what the shot is claiming to be, and whether the edit changed that.

21 August 20268 minute read

Cropping, dodging, burning, pushing the contrast and spending an afternoon on a single print: darkroom photographers did all of that, and argued about how much of it was allowed, decades before anybody had a slider. Digital made the same work faster and put it in more hands, but the work itself didn't change. Pretty much every photograph is edited, so it's better to ask what kind of image this is supposed to be, and whether the editing changed that claim.

Is editing photos cheating?

No. Cropping, colour, dodging and burning, sharpening and local exposure are the last part of making the shot. If you refuse to touch them on principle, you're letting the camera's default settings make those decisions for you. The argument worth having is about what the shot is claiming to be. A personal shot made for yourself has lots of room. A wildlife competition entry or a news photograph has almost none, however many sliders moved.

A grey Belgian Air Force A400M on the runway with all four propellers turning, trees and a cloudy sky behind. — Mark Rycroft's A400M on RIAT departures day. His caption lists the edit: cropped to get rid of distractions, foreground desaturated a little and sky enhanced. That's normal editing, and it has an Edit Reveal on Kōlab if you want to see where it started.

Editing has always been part of the finished shot

A raw file is a set of measurements waiting for somebody to decide what they mean. The camera's own preview is one version of that, decided by an engineer who has never seen your scene. Exposure, contrast, colour, white balance and crop get decided by somebody either way, and it may as well be you.

An F-22 Raptor almost in silhouette at sunset, with orange light glowing through the cockpit canopy — At ISO 100 and ƒ/9 the sky keeps its colour and the F-22 comes out dark, so everything after that is deciding how dark to leave it. That's dodging and burning, just with a slider instead of a bit of card.

That still leaves the interesting question: how far can the decisions go before the shot starts showing an evening that didn't happen? The fair objection is that the claim is almost never stated. Nobody captions a landscape "this is an interpretation", viewers assume a photograph shows something real, and a composite can ride on that assumption without the photographer telling a single lie. That's why the rest of this is about the claim rather than the sliders.

The line depends on the kind of photography

Personal and creative work

Your shot, made for yourself, and you can do pretty much anything: composite it, replace the sky, take the colour somewhere no evening has ever been. The one thing we'd ask is that if somebody could reasonably assume they're looking at a real moment, you say enough that they aren't misled. A line in the caption covers it.

Documentary and journalism

Here the claim is out in the open, and it's the strongest one photography makes: this happened, and this is what it looked like. Removing something edits the event, and adding something is worse. What's left is tone and exposure, and even there the test is whether somebody standing next to you would recognise the scene.

Wildlife

In wildlife the problem is behaviour rather than colour. Warming up a shot changes how it feels. Combining two moments so an animal appears to do something it didn't do changes what the shot says, and behaviour shots are valued because the behaviour is hard to see. Edit the light as much as you like and leave the animal doing what it did.

Competitions

Read the rules, and read them before you edit rather than after you've placed. Competitions draw the line in very different places, and theirs is the only one that counts. Many ask for the original file if they're interested in your shot, so keep it whether or not you ever plan to enter anything.

A useful test

The question to ask

Would somebody feel misled if they knew exactly what you did? If not, you're probably dealing with ordinary editing. If so, the fix might just be saying what you did. You don't necessarily need to edit less, you may just need to describe the shot more accurately.

Forks of lightning across a purple night sky over tower blocks and cranes in Istanbul — Thirty seconds of a storm over Istanbul in one shot, by Can Karayel. Nobody standing there saw it like that in a single moment, and the shot isn't claiming they did.

Why photographers worry about this

Underneath the ethics there's usually a smaller worry: if my raw file doesn't look amazing, did I get the shot wrong? No. A flat raw file is what a well exposed capture looks like. The photographers whose finished work you admire mostly aren't getting better files than you. They go out knowing what the file is for, and finishing it afterwards is most of the job. Treating the edit as a consolation prize has put plenty of people off learning the half of photography you do sitting down.

A KC-135 tanker departing Phoenix against a golden winter sunset — Joe Copalman's KC-135 out of Phoenix Sky Harbor, with the downtown skyline reflecting a golden winter sunset behind it. Most raw files come out of the camera flatter than this, and knowing where to take them is most of the job.

Where editing goes wrong

Almost nobody gets into ethical trouble with editing. Visual trouble is far more common, and it tends to come from the same few habits.

Trying to rescue everything

Some shots need editing and some need shooting again, and learning to tell the two apart quickly is worth more than any editing technique. An hour dragging a weak file towards acceptable gets you one acceptable file. The same hour spent going back and shooting it again gets you something you'll keep. If the problem is the light or where you were standing, close the laptop, as no amount of Lightroom will move you three steps to the left.

Giving every shot the same treatment

Presets are fine for getting an edit started, but you'll nearly always need to adjust from there. Put the same look on a misty morning and a harsh midday scene and both end up looking like the preset rather than the day. A portfolio where every shot has the same shadows is the giveaway. Match the edit to the light instead, even though it takes longer.

Carrying on too long

There's a point in every edit where you stop making decisions and start fiddling, and you can usually feel it coming a while before you admit it. Past that point you're mostly just getting used to the shot. Leave it and open it again tomorrow with fresh eyes. Half the time you'll undo most of the last hour (and wonder what you were thinking with the clarity slider).

Never checking another screen

Most people will see your shot on a phone, held at arm's length, quite possibly on a bus. Deep blacks block up, strong warmth turns orange, and saturation that looked great on a good monitor turns garish on a screen adding its own contrast. Check every edit on a phone before you put it up. It only takes a moment and it catches most of these.

AI has made the question more complicated

All of that was settled long before generative tools came along, and they have moved the argument on. Denoise, masking and tonal work operate on what the sensor recorded, however clever the maths. Generating a subject, extending the scene or replacing a sky adds information that was never in front of the camera, and that's a different kind of edit. Whether photographers should disclose AI editing goes through where the line sits and how to say what you changed.

Show the process

The quickest way past the worry is looking at other photographers' starting files. Kōlab's Edit Reveal puts the original behind the finished version so anybody can drag between them, and after a dozen of those the worry mostly goes. A strong edit shows that somebody made decisions and knew which ones to make, whatever the original looked like. There's more on that in why before and after is the fastest way to learn editing.