How to Reverse-Engineer a Photo

Monday, October 22, 2018


Although it sounds like a highly technical term, ‘reverse engineering’ is something you’ve done many times. Any time you’ve asked questions like “What camera did you take that with?”, “What settings did you use?” or “Where was this taken?”, you’ve been trying to reverse engineer a photograph.
lake how to reverse engineer photos
We’ve all looked at a photo and tried to figure out how it was created. I do it every day. Whether you’re conscious of it or not, when you see a photo you admire you try to analyze it. You’re asking, “How do I take photos like that?”
The truth is, if you ask the photographer about their camera or settings you’re asking the wrong questions. By all means, ask questions. After all, that’s how we learn. But you can also learn a lot from studying an image – if you know what to look for. When you can visually deconstruct an image, you’re one stop closer to being able to create something similar.
This isn’t a lesson in plagiarism. It’s simply a way to learn from other photographers whose work you admire. No successful artist would be where they are today without learning from the works of others they look up to.

Light and Shadow

The most important photographic lesson I ever learned is that it’s all about the light. Reverse-engineering photos is no different. Analyzing the light in an image is the simplest and most effective way to learn how a photo was made.
When you look at the image, ask yourself a few questions.
  • Which direction is the light coming from?
  • Is there more than one light source?
  • Is the light hard or soft?
  • Is there reflected light in the photo?
  • What about the color temperature? Is it warm or cool?
Sometimes the answers will be obvious. Sometimes they’ll be impossible to answer. But the more often you ask them, the better you’ll get at answering them.
If you’re looking at a landscape photo, you can almost always assume there’s only one light source – the sun. But that doesn’t mean you can’t deconstruct the light. The direction, hardness and temperature of the light will tell you a lot about the conditions the photo was taken in. Even though beautiful landscape lighting isn’t as technical as portrait or product lighting, you can still learn a lot from analyzing it.
aerial how to reverse engineer photos
The sun striking this landscape, along with the warm light on the right side of the image, clearly show where the light is coming from.
If you’re reverse-engineering a portrait, it’s more likely to have more than one light source, as well as reflected light. When a photographer starts balancing multiple light sources, reverse-engineering a photo can become more difficult. But there are still ways to analyze the light if you know what to look for.
Start by asking yourself, “Where are the shadows?” It may seem a little backwards, but one of the best ways to analyze light is to look at the darker parts of the image. Where is there no light? Do you see any hard shadows? Are there areas where you can see the light dropping off gradually? Studying the shadows will tell you about the direction of the light as well as how large it is relative to the subject.
bear how to reverse engineer photos
The illuminated fur around the outside of the bear show that this image was backlit. And its shadow on the ground shows the exact direction the light was coming from.
Interpreting a photograph’s light becomes more difficult as the lighting gets more complex. As more light sources or reflectors are added, the shadows become less obvious. If the shadows are very light or non-existent, it likely means either the light is very diffuse and bouncing all over the place, or there are multiple light sources.
If you’re lucky, you can sometimes see exactly what light source was used by looking for reflections. Look at the eyes, glasses, windows, water surfaces, and anything that reflects light. Sometimes you can see a perfect reflection of the light source, but at the very least you’ll be able to see its direction.
cabin how to reverse engineer photos
The soft light on the subject, combined with the reflection in her glasses, show the window as the light source. The very dark shadows tell us there are no other light sources.

Gear and Settings

In many cases, you don’t need to ask what equipment or settings were used to create a photo. With practice, you can learn to guesstimate the technical details such as focal length, aperture and shutter speed.
Figuring out what focal length was used isn’t too difficult once you know how focal length affects a photo. As a general rule, the shorter the focal length (wider angle), the more distortion you’ll see and the more of a scene will fit in the frame. As the focal length gets longer (normal or telephoto), you’ll see more compression in the image and less of the scene in the frame.
car how to reverse engineer photos
Only a very wide-angle lens can capture everything in a scene like this from the ground to the sky. The lens distortion makes closer objects like this car look much bigger.
While this won’t tell you the exact focal length used, it will give you a ballpark figure. With practice, you’ll be able to tell if a photo was taken with a wide-angle (<35mm), normal (35-85mm) or telephoto (>85mm) lens. The exact number doesn’t matter. What does matter is getting a rough idea where your focal length needs to be to create the same look.
As with focal length, you can figure out roughly what aperture was used by understanding how it affects an image. As the lens aperture opens and closes, the depth-of-field (DOF) of the image changes. The wider the aperture (smaller f-number), the narrower the DOF.
Again, the exact number doesn’t matter. What matters is understanding how aperture affects DOF and how to interpret the DOF of a photo. If the image is sharp and in focus from the foreground right through to the background, a smaller aperture (f/11-22) has probably been used. If everything but the subject is soft and out of focus, a larger aperture (f/1.4-5.6) has probably been used. If the DOF is somewhere in between, the aperture is probably around f5.6-11.
Finally, these principles can also be applied to shutter speed. You probably know that shutter speed affects the way movement appears in an image. If objects you would expect to see moving are frozen still, you know a faster shutter speed was used. If there’s some motion blur in the image, you know the shutter speed was slower.
bay how to reverse engineer photos
You can see that a longer shutter speed has been used here to create the milky water effect, common with long-exposure photography.
With a landscape photo, any time you see silky-smooth water or clouds common with long exposures you know it has a shutter speed of at least a few seconds. If you’re seeing some movement, it’s more likely to be less than one second. To freeze movement, you’d expect shutter speeds of at least 1/100th of a second.
rocks how to reverse engineer photos
Very short shutter speeds are required to capture moving water, as in this seascape photo.
If the photo doesn’t include any moving objects, it’s much more difficult to figure out the shutter speed used. But if there’s no movement then shutter speed doesn’t really matter. It just needs to be fast enough to avoid any blur caused by camera movement to ensure a sharp image.

Post-Processing

Reverse-engineering the post-production that’s been applied to an image is the trickiest part. There’s almost no limit to what can be done in Photoshop today, which makes it difficult to figure out how a photo has been processed.
You can get a rough idea of how much post-processing has been applied by looking at the photo. Does it look realistic? Do the colors and tones appear the way you’d expect in real life? Is the whole image well exposed from the darkest shadows to the brightest highlights? Are the light and shadows consistent across the image as you’d expect? Do the people look real, or impossibly perfect?
Asking these kinds of questions will help you know what to look for. It’s easy to look at the image as a whole and get frustrated trying to analyze it. As you break it down and look at the individual details of a photo, it becomes easier to see the edits that have been applied.
I’ll admit this isn’t exactly my strong suit, being colorblind. Picking out the color grading or effects that have been applied is never easy for me. But with practice I’ve become much better. And if I can do it, so can you.
london how to reverse engineer photos
You can see by looking at the church that extra warmth has been added in post-production to emphasize the warm late afternoon sun.
Keep in mind that photographers and retouchers that are highly skilled in Photoshop, are very good at making their images look natural and unedited. Just because an image looks real doesn’t mean it is. A photo that’s been edited by a Photoshop ninja will be very difficult to reverse-engineer.

Exif Data

When all else fails, and you desperately want to know the settings used to take a photo, you may be able to access the image’s exif data. When a digital photograph is created, a bunch of data is embedded into the file. This includes focal length, shutter speed, aperture, camera model, and often a bunch of other information.
A photo’s exif data is often stripped out by the photographer or the website it’s uploaded to. But if it hasn’t been stripped, you can easily access the data by either:
  • downloading the image and reading the data on your computer
  • using one of the many websites that will analyze a photo’s exif data for you.
Some websites, such as exifdata.com, can even analyze a photo from the image’s URL.
cuba how to reverse engineer photos
The shadow of the tree clearly shows the direction of the sun, while the light reflecting off the concrete has filled in the shadows on the subject.

 

No comments:

Post a Comment