In the Eye of Ivanoh: The Limits of Digital Cropping

The high resolution of digital cameras is transforming our profession. New camera models now offer digital files that hover around 50 million pixels. This abundance of data allows us to crop considerably within the same file. A technique that sometimes allows us to discover a picture within the picture.

This good old trick was given to me by the photographer Patrick Sanfaçon of The Press. After choosing a first framing that we consider final, we will zoom into our image about 30% to see if our eye does not capture something else. This reflex will often change our original perspective and occasionally give us nice surprises.

However, there is one big red flag. Resolution has its limits! There are factors to consider before doing such a thing. Basically, enlarging the image by half is often the maximum. You can cheat with pixels, but not too much.

Then there is a simple visual analysis to be done. If we perceive the shape of the pixels, we have gone too far. This is what must be avoided at all costs.

I have chosen some photos taken during the US presidential election debate between Donald Trump and Kamala Harris to illustrate this subject.

I noticed that photographer Saul Loeb of the Getty news agency cropped different images from the same digital file.

During the debate, the two candidates were a good distance apart, so a wide shot was needed to illustrate the dynamic between the two politicians.

Thanks to the high resolution, if a person’s expression is interesting, we can crop and isolate one of the two speakers in post-production.

This information is not trivial, because when shooting, we are not constantly obliged to change lenses, knowing that we can reframe if necessary in post-production.

During a commercial break, Donald Trump left his lectern. He seemed rather upset. The photographer wanted to isolate the expression of the Republican candidate. So he allowed himself to frame more tightly, but from the same file.

At the very beginning of the debate, Kamala Harris approached Donald Trump to shake his hand. It was an important moment, the only one during the evening where the two candidates were next to each other.

The image is successful. The moment of the handshake is well captured and the eyes of the two candidates meet.

The photographer, or sometimes a photo editor, has cropped the handshake into a very close-up. The result is unfortunately disappointing, if not unpublishable. We have gone too far, the pixels are omnipresent.

The appearance of pixels is due to several factors:

• The most common problem is enlarging the image too much
• The compression format of images during transmission may cause a loss of quality
• When the lighting is low, you should increase the sensitivity (ISO) of the camera. The higher the sensitivity, the more pixels will be visible, giving a grainy effect.
• When there are very deep blacks, pixels will show up quickly when you enlarge the photo. In this case, Kamala Harris’s black coat and the black background in the background add to the problem.

It is surprising that the agency chose to put this photo on its site. I have trouble understanding this decision.

Floods caused significant damage in southwestern France earlier this month.

Part of a bridge has been moved by the force of the water. The debris around the structure is omnipresent in the shot. We can easily spot the man with his red sweater in the center of the image.

An informative, well-composed photo that gives us a good understanding of the state of affairs.

In order to better understand the enlargement issues, I cropped the photo excessively.

The grass on the bridge has a funny texture, the outline of the shapes is jerky, a little blurry. The texture of the water is not very nice. The digital file is insufficient, giving a disappointing result.

In addition, the loss of visual information is too great, so it is desirable to return to the original image.

My wink of the week

I chose this image from all those I captured during Géhane Kamel’s press conference because the coroner’s facial expression was appropriate for the nature of the subject.

However, there is a problem: the photo was poorly framed. It gave the impression that the desk where Géhane Kamel was sitting was leaning to the right.

Fortunately, when cropping, it is possible to change the overall tilt of the image. This feature allows you to adjust the orientation of the frame to the left or right.

By changing the tilt of the desk, I corrected the horizon line and fixed my problem.

When the resolution allows it, I keep a little space (25% more) to be able to crop as I wish later. In this case, I could easily make a vertical or square photo.