RunTheTests
/

Gamma Calibration Chart

Estimate your display's gamma by matching a grey patch against a 50% dither pattern.

Which square blends into the lines around it?

Sit back from the screen, or squint, until the striped background looks like flat grey. One square will disappear into it. Judge it straight on — on most panels the match shifts if you look from an angle.

How this test works

A fine pattern of alternating pure black and pure white pixels emits, on average, exactly half the display's maximum light. That is an average of photons, and it holds no matter what transfer curve the panel applies. A solid grey square looks identical to that pattern only when it emits half the maximum light too — which happens at the encoded value 0.5^(1/gamma). For the sRGB target of 2.2 that value is 186, not the 128 most people expect; 128 would be correct only on a display with gamma 1.0, which nothing has. Five squares are drawn, one for each candidate gamma, each surrounded by the dither. Whichever disappears into its background identifies the curve your display is actually using.

What the results mean

Estimated gamma
The transfer curve implied by the patch you matched. 2.2 is the sRGB standard that web and desktop content is authored against.
Matching patch value
The 8-bit level that emits 50% luminance at that gamma. Worth seeing because it is nowhere near 128.
Device pixel ratio
Must be a whole number for the dither to land on real pixels. Anything else and the browser averages the lines away.

Common problems and fixes

Every square looks like it matches
The line pattern is being resampled, which happens at any zoom level other than 100%. Press Ctrl+0, or Cmd+0 on a Mac, and they will separate.
The match changes when I move my head
That is panel viewing angle, and it is normal on TN and many VA panels. Judge from directly in front, at the vertical centre of the screen.
My gamma reads 2.4 and I mostly watch films
2.4 is the BT.1886 target for video in a dark room and is correct for that. It is too dark for general desktop and web use, which assume 2.2.

Frequently asked questions

Why is the correct grey 186 and not 128?

Because displays are not linear. Encoded 128 is about 22% of maximum light on a gamma-2.2 display, not 50%. To emit half the light you need 0.5^(1/2.2) = 0.73, which is 186 in 8-bit. The gap between 128 and 186 is the gamma curve made visible.

Is this as good as a hardware calibrator?

No, and it is not trying to be. A colorimeter measures the actual light coming off the panel at many points and builds a correction profile. This reads one point on the curve through your eye, in steps of 0.2, and tells you whether you are near the standard.

Does my monitor's gamma setting have to be 2.2?

For web browsing, photo work and general desktop use, yes — that is what the content assumes. For film in a dark room, 2.4 is the deliberate standard. There is no single right answer independent of what you are viewing.

More in Device