Correcting Brightness and Contrast in a Clinical Photo

A before photo is a bit dark and the after is a bit bright. The natural impulse is to open an editor and match them. Often that is reasonable. Other times, it is the start of an edit that changes what the photo shows. The difference between correcting exposure and manipulating the result lies in how, why and on which photos the adjustment is made.

This page shows what is acceptable, how to do it equivalently on both photos and why it is better to avoid needing to correct at all.

Why do photos come out with different brightness?

Brightness and contrast vary for several reasons, and almost none relate to the patient:

• different ambient light (time of day, bulb, window);
• the camera's automatic exposure shifting with the frame;
• different skin and clothing colors in the frame;
• distance and angle of the light;
• modes such as HDR and portrait, which redraw the image;
• differences between the devices used at each visit.

If you have not stabilized these factors, the before and after already come out different, and post-shot correction tries to fix what the routine did not. See how to avoid it in locking exposure and focus on the phone and HDR and portrait mode in clinical photos.

What is an acceptable correction?

A correction is acceptable when its only goal is to make photos comparable, without changing what they show.

Examples of adjustments usually tolerated in standardization:

• matching overall exposure between two photos;
• correcting color based on a gray card (see white balance in clinical photos);
• cropping and aligning identically (see cropping and rotating clinical photos).

What is not acceptable?

The line is clear: if the edit changes what the photo shows about the patient's face, it stops being correction and becomes manipulation. The limit is explored further in to edit or not to edit a clinical photo.

How to make the adjustment with care

1. Keep the original in a separate location and work on a copy.
2. Choose a reference that appears in both photos and does not change: the neutral background, the gray card or neutral clothing.
3. Measure that reference's value in both images (many editors have an eyedropper with a numeric readout).
4. Apply the same type of adjustment (global exposure) to both photos until the references match.
5. Do not use curves, shadows or local highlights.
6. Check side by side and by flipping between them, to see whether the face looks the same.
7. Note in the chart that an adjustment was made and what it was.

Can you avoid the correction?

Almost always. The most effective way is to standardize capture:

• diffuse, symmetrical light that is always the same, as in lighting for clinical facial photos;
• locked exposure and focus;
• same background, same distance and same height;
• no HDR or portrait mode;
• the same device, whenever possible.

With that, the brightness of the two photos ends up close enough to skip the editor. The time spent on the capture routine is less than the time spent correcting each photo later.

The case of swelling and redness

In photos taken shortly after a procedure, the face may show swelling and redness. Correcting brightness or color to "hide" that is editing the result. The right approach is to photograph at the interval set by your routine and, when the photo has to be taken earlier, explain to the patient that it records the recovery phase. See result photo with swelling still present.

What about Harmy?

Harmy's projection keeps the lighting, skin, background and hair of the original photo and changes only the marked areas. That means a photo that is very dark, bright or contrasty produces an illustrative simulation with the same flaws. If you correct the photo before using it in the app, make only a light global exposure adjustment, never retouch the skin, and keep the original. Generally, it is better to retake the photo with proper light. See bad photo for AI projection.

Summary for your routine

Get the light and exposure right at capture, keep the original, correct only globally and equally on the photos being compared, never beautify and record the adjustment. The less you need to correct, the more reliable the photo is as a document.