Grey-to-Grey and the Transition They Choose
That single response-time figure in the spec sheet is the best measurement from a 65,536-entry table. Here is how manufacturers pick it, and what the rest of the table looks like.
The Number on the Box Is Not an Average
LCD response time is quoted in grey-to-grey (GtG) milliseconds — the time for a liquid-crystal cell to rotate from one grey level to another. The full picture is a 256×256 transition matrix: every combination of starting grey (0–255) and target grey (0–255), measured individually. That is 65,536 entries, and manufacturers report exactly one of them.
The one they report is the fastest. Transitions between mid-grey levels — typically around the 128-to-64 or 64-to-128 range — rotate the crystal the least per unit time and respond best to overdrive pulse shaping. Near-black and near-white transitions move through a much larger angular range and are sluggish by comparison; the nematic material's viscosity and elastic restoring force both fight the transition at extreme tilt angles. A panel quoted at 1 ms GtG with overdrive may take 8–12 ms on a dark-to-bright transition, and the spec sheet says nothing about this.
Overdrive — applying a voltage higher than the target level to accelerate crystal rotation, then pulling back — compresses the fastest transitions dramatically, which is why the marketing number has dropped from 8 ms to sub-2 ms over the past decade without the underlying material changing much. Overdrive is tuned per-transition in the panel's look-up table (LUT), and the tuning is optimised for the transitions that produce the headline number. Poorly compensated transitions produce coronas and inverse ghosting, artefacts that are the direct cost of aggressive overdrive on transitions the LUT handles badly.
Reading a Full Transition Matrix
Reputable third-party reviewers publish full GtG heatmaps — a colour-coded 256×256 grid where the x-axis is the starting grey level, the y-axis is the target grey level, and each cell encodes transition time. Reading one reveals the actual distribution:
The diagonal is irrelevant. Staying at the same grey level requires no transition; diagonal cells are zero by definition and should be ignored when forming any impression of panel speed.
The corners are the worst case. Top-left (black to white) and bottom-right (white to black) are the full-range transitions. High-quality panels get these below 5 ms with overdrive; budget panels may sit at 15–20 ms. These transitions dominate perceived motion clarity on high-contrast scene cuts.
Mid-range is where the spec lives. The cluster of fast values around the matrix centre is where the headline figure is measured. On most VA (vertical alignment) panels, even this region runs 4–8 ms without overdrive enabled; the 1 ms figure requires aggressive overdrive that often introduces visible overshoot on those very transitions.
VA versus IPS asymmetry. VA panels are substantially slower than IPS (in-plane switching) in the dark regions of the matrix — transitions involving grey levels below roughly 30 are notoriously slow because the crystal must overcome a steeper potential well near its rest orientation. IPS panels are more uniform across the matrix, though their absolute best is not always as fast as the best VA claim. This structural difference is the main reason VA panels have historically exhibited more visible trailing on dark objects against dark backgrounds, regardless of what the spec sheet says.
Average GtG is the honest metric. Some reviewers compute a weighted or simple mean across the full matrix, excluding the diagonal. This number is usually several times higher than the headline figure on fast gaming panels, and the ratio itself is informative: a smaller ratio indicates better overdrive LUT coverage; a larger ratio indicates the LUT is highly optimised for a narrow set of transitions.
When evaluating a panel for motion-critical work — esports, simulation, video production monitoring — the matrix mean and the worst-case corner values matter far more than the advertised figure. A panel with a 1 ms headline and a 6 ms matrix mean beats one with a 0.5 ms headline and a 10 ms mean in every real-world frame transition that is not the cherry-picked one.
The spec sheet is not lying, exactly. It is reporting a measurement correctly labelled as grey-to-grey. It just neglects to mention which grey, or which direction, or that overdrive was engaged at its most aggressive setting. Reading the full matrix is the only way to know what the panel actually does.