Image Exposure

Correct image exposure in stops, the way a camera does

Exposure is the amount of light that reached the sensor. When a shot comes out too dark or too bright, the most natural fix is to change exactly that — multiply the light, not the stored pixel numbers. The image exposure tool decodes your image to linear light, scales it in EV stops (+1 EV doubles the light, −1 halves it) and encodes it back to sRGB.

Two helpers make larger corrections usable: Highlight recovery rolls bright tones off smoothly instead of letting them clip, and Shadows opens up or deepens dark areas without moving the highlights. A live histogram reports clipping as you work, and you can process a batch at full resolution in your browser and save it as a ZIP.

How to use the Image Exposure

1. Open the photos that need correcting — drag them in, browse for them, or paste from the clipboard.
2. Set Exposure in stops: try +0.5 to +1.5 for dark photos and −0.3 to −1 for bright ones. Type values such as 0.33 in the box for ⅓-stop steps.
3. If skies or white clothing turn flat white, raise Highlight recovery until the clipped-highlights figure drops.
4. Use Shadows to reveal detail in dark areas (positive) or add depth (negative).
5. Compare with the split preview and download one image, or a ZIP of all of them.

Image Exposure features

Why exposure has to happen in linear light

Image files store gamma-encoded values: the sRGB standard spends more of its 256 levels on dark tones because our eyes are more sensitive there. Level 128 is not half the light of level 255 — it is about 21.6%. Doubling the stored number therefore does not double the light, and it upsets the balance between shadows and highlights.

This tool decodes each value with the sRGB formula lin = ((v ÷ 255 + 0.055) ÷ 1.055)^2.4 (or v ÷ 255 ÷ 12.92 for the very darkest values), multiplies by 2^EV, and encodes the result again. Photographic mid-grey, 18% reflectance, is stored as level 118. At +1 EV it becomes 36% light, which encodes to level 162 — not 236, which is what simply doubling the number would give.

Exposure at a glance

How typical levels change at different settings, with highlight recovery off:

Highlight recovery and shadows, exactly

• Highlight recovery sets a knee k = 1 − 0.6 × h ÷ 100 in linear light. Values above it are compressed with k + (1 − k) × (1 − e^(−(x − k) ÷ (1 − k))), which joins the straight line smoothly and approaches white without reaching it. At 100 the knee is 0.4, so even light four times brighter than white keeps some separation from light twice as bright.
• Shadows multiplies each pixel by 2^(1.5 × s ÷ 100 × (1 − Y′)²), where Y′ is the pixel’s encoded luminance from 0 to 1. Black gets the full ±1.5 EV, mid-tones about a quarter of it, white nothing.
• An 8-bit JPEG cannot bring back what is already 255: a blown-out sky stays white. Recovery prevents new clipping when you brighten; it cannot invent lost detail.

Tips

• Correct exposure first, then contrast and colour — the contrast and saturation tools work best on a well-exposed picture.
• Phone photos get noisy when lifted by more than about 1.5 EV.
• Exposure brightens highlights faster than shadows. For a gentler midtone lift, the Midtones mode of the brightness tool may suit better.
• Save as PNG if you plan more edits, to avoid repeated JPG compression.