Image Histogram Viewer — RGB & Luminance Distribution
Upload an image and instantly see its histogram: luminance plus R / G / B channel curves, with mean, median, standard deviation and shadow/highlight clipping percentages to judge exposure and color balance. Computed locally in your browser — no upload. Works on desktop, tablet & mobile.
A histogram is the health report of a photo’s exposure: the horizontal axis runs from shadows to highlights (0-255), the vertical axis counts pixels at each level. A distribution squeezed left means a dark image (underexposure tendency); squeezed right means bright (overexposure); a "cliff" at either end means shadow or highlight detail is already lost. This tool counts every pixel locally in your browser: the luminance histogram uses the perceptual 0.3R+0.59G+0.11B weighting, and separate R / G / B and overlay views reveal color casts. You also get mean, median, standard deviation (contrast) and clipping percentages at both ends, plus a plain-language interpretation. Everything runs locally — no upload.
When you need this
Photo review
Check exposure balance and end clipping after a shoot.
Before editing
Read the distribution before lifting shadows or taming highlights.
Color cast check
Strongly shifted RGB channels usually mean white balance is off.
Reading a histogram
- Whether the distribution covers the range — headroom at both ends is usually healthy
- Clipping percentages — whether shadow/highlight detail is already gone
- Whether RGB channels align — big offsets usually mean a color cast
- There is no "correct" shape: snow scenes lean right, night scenes lean left — both are normal
- It is a global statistic: a small blown-out area may barely show in the curve
- Large images are downsampled proportionally; the distribution shape is unaffected
Histogram FAQ
There is no single answer. Generally a wide distribution without cliffs at either end means balanced exposure and intact detail — but snow or high-key portraits naturally lean right, night scenes lean left, and both are fine. The histogram confirms your intent; it does not force every photo into one shape.
Clipping means pixels piled up at the extremes: a spike at 255 means blown "pure white" highlights, a spike at 0 means crushed "pure black" shadows — detail there cannot be recovered in editing. The tool reports the pixel share in the outermost 5 levels at each end; above roughly 2% deserves attention.
The luminance histogram blends the three channels with perceptual weights (green weighted highest) into one curve for judging overall brightness. The RGB view draws each channel separately — when the three curves are clearly offset, you likely have a color cast or white balance issue. Use luminance for exposure, RGB for color.
No. Pixel counting happens locally in your browser; large images are downsampled proportionally first (which does not change the distribution shape). Nothing is sent over the network or stored.