Reading the histogram
Left is black, right is white, and the height is how many pixels sit at that brightness. That is the whole thing. What people get wrong is believing there is a correct shape.
- A spike on the right edge means clipped highlights — those pixels are pure white and no editing recovers them.
- A spike on the left edge means crushed blacks. Usually less costly, and sometimes what you wanted.
- A gap at one end means you left range unused. On raw files, gaps at the right end are wasted signal-to-noise.
- A comb pattern — regular gaps through the middle — means the file has been stretched in editing and is running out of levels. It is the visual signature of banding to come.
- Colour channels apart at one end means one channel clipped first. In raw that is often recoverable; in JPEG it usually is not.
A histogram describes distribution, not correctness. A snow scene should pile against the right. A low-key portrait should pile against the left. The only real error is unintentional clipping — which is what the percentages above measure.
Zebra, clipping map, peaking
Zebra stripes
Broadcast monitors have marked over-bright areas with diagonal stripes for decades. IRE is a video brightness scale where 0 is black and 100 is nominal white, so 100 IRE zebras mark genuinely clipped pixels, 95 warns you a fraction of a stop earlier, and 90 is the classic setting for placing lit skin tones. The advantage over a histogram is spatial: you see which part of the frame is going.
Clipping map
Marks blown highlights in red and crushed shadows in blue at once. Faster to read than zebra when you are checking a finished frame rather than setting exposure.
Focus peaking
Detects high-contrast edges and tints them green. Sharp regions have strong local contrast; soft ones do not, so the outline traces the plane of focus. This page runs it on a still, but its real home is a live viewfinder during manual focus.
Common questions
How do I read a histogram?
Black on the left, white on the right, pixel count as height. Spikes at either edge mean clipped tones. Everything between is distribution, and distribution is a creative decision rather than a right answer.
Why does the camera say I clipped when the raw file looks fine?
Because the camera's histogram is built from the embedded JPEG preview, contrast curve and all. Raw typically holds between half a stop and a full stop more highlight information than that preview shows.
What percentage of clipping is acceptable?
There is no threshold, only intent. A frame with the sun in it can clip several percent and be perfect. A portrait clipping half a percent on the forehead is a problem. Use the overlay to see where before you judge how much.
Does this work on raw files?
Not directly — browsers cannot decode most raw formats. Export a JPEG or TIFF first. Do note that the export carries whatever curve your converter applied, so it will clip earlier than the raw did.
Is my photo uploaded?
No. It is drawn into a canvas in your browser and analysed in JavaScript on your machine. There is no server on this domain to receive it.
Related tools
All three, live in the viewfinder.
Lucid puts a real-time RGB histogram, selectable zebra stripes and focus peaking on your iPhone's preview — before the shot, where they actually help.