Wedges should fade out smoothly, not abruptly
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
Near White
Click for Fullscreen

Test Pattern

Healthy Result

Near-white wedges separate step by step until they merge only at the very tip; near-black wedges stay distinguishable down to the darkest step; the B/W bars show crisp, evenly spaced transitions with no blown-out or crushed zones.

OSD Calibration

Lower brightness until the near-black steps separate, then raise contrast until the near-white steps separate. Back off slightly from clipping on either end, and balance the two so midtones stay neutral — contrast first, brightness second, then fine-tune.

Dynamic Contrast Warning

Disable dynamic contrast (DCR / "auto contrast") before testing. It re-maps the image in real time, crushing highlight and shadow detail so wedges merge or pop unpredictably — the test then measures the monitor's processing, not its real contrast.

Contrast Test: Judge Shadow and Highlight Detail and Calibrate Contrast Online

Contrast is where monitors quietly fail: a panel can look bright and punchy yet crush every step below code 32 and clip everything above 224. This free contrast test gives you a near-black wedge (0-32), a near-white wedge (224-255) and pure black/white bars so you can see exactly where your display loses detail — then fix brightness and contrast in the OSD until both ends survive.

How the wedges and bars work

The near-black wedge stacks eleven steps from 0 up to 32; the near-white wedge stacks the mirror-image steps from 224 up to 255. Each step is only a few code values apart, which is exactly where panels give up: cheap scaling, aggressive contrast settings and 6-bit+FRC dithering panels merge neighboring steps first. On a healthy SDR display in a dim room you should count most of the steps as separate blocks on both ends.

The black/white bars do a different job: alternating full black and full white columns expose what happens when the panel switches between extremes. Watch for glow around the white bars (IPS/VA glow, or blooming on edge-lit backlights) and for blacks that look gray while white bars sit next to them — that pair of artifacts is your practical contrast, regardless of what the spec sheet claims. For a cleaner look at dark-screen uniformity, cross-check with the black screen page.

Reading shadow and highlight detail

Work in a dim room after the backlight has warmed up for a few minutes. Start with the near-black wedge: if the first two or three steps melt together at the left, your shadows are being crushed — the usual causes are contrast set too high, brightness (backlight) set too low, or a "dynamic contrast" mode re-mapping the curve behind your back. Then check the near-white wedge: if the top steps merge into one white slab, highlights are clipping and the contrast control is pushing the panel past what it can resolve.

Ambient light changes the verdict. In a bright office a slightly crushed near-black wedge is invisible and harmless; in a dark home-cinema room the same monitor shows only 30-32 as one block and every night scene looks muddy. Judge the wedge in the lighting you actually use, and note where the boundary sits — that number is your personal calibration target for this display.

Calibrating contrast, and static vs dynamic contrast

In the OSD, contrast only shapes the top of the signal (how 224-255 maps to the panel) while brightness is really the backlight. Practical order: set brightness for a comfortable white luminance first, then raise contrast until the 255 step stays clearly separate from 250 in the near-white wedge — the moment they fuse, back off one or two clicks. Re-check the near-black wedge afterwards, since contrast changes can darken low steps too. If steps keep refusing to separate, compare against the grayscale test — merged steps across the whole range point to a panel bit-depth limit rather than a settings issue.

Always disable "dynamic contrast" (also sold as DCR, auto-dimming or contrast enhancer) before judging anything. It re-maps the tone curve frame by frame, so the wedges breathe and shift under your eyes and the numbers mean nothing; it also causes banding you can confirm on the gamma and banding page. Static contrast is the honest figure — roughly 1000:1 on IPS, 2500-3000:1 on VA — and it is what you improve with the calibration above. HDR panels play by different rules; run the HDR test separately. For profiling beyond eyeballing, continue on the color calibration page.

Frequently Asked Questions

What is a contrast test?+

It shows fine steps just above black (0-32) and just below white (224-255) plus black/white bars. Where steps merge, your display is crushing shadows or clipping highlights; the bars reveal glow and how blacks hold up next to white.

What should contrast be set to?+

Whatever keeps the 255 step separate from 250 in the near-white wedge while the near-black wedge still shows its first steps — usually 70-85 on a PC monitor, not the 100 default. Brightness should be set for comfortable white luminance first.

What is the difference between static and dynamic contrast?+

Static contrast is the real ratio the panel delivers on one frame, around 1000:1 for IPS and 2500-3000:1 for VA. Dynamic contrast dims or brightens the backlight depending on the scene, inflating spec-sheet numbers into the millions while making gradients and tests unstable.

Why can I not see separate steps in the near-black wedge?+

The room is too bright, the backlight is set too low, contrast is crushing blacks, or dynamic contrast is enabled. If steps stay fused after fixing those, the panel itself lacks the bit depth — verify with the grayscale test.

Does a higher contrast ratio number always look better?+

No. A million-to-one dynamic figure tells you nothing about real image quality; a panel with a solid 3000:1 static ratio and clean steps will beat a poorly calibrated 5000:1 display. Judge steps and glow on screen, not on the box.

Can a browser measure my real contrast ratio?+

No — true contrast needs a meter, because browser output depends on your OSD settings and room light. What this test does measure reliably is where detail is lost at both ends, which is what calibration actually needs.