Display response time and motion blur

Display response time measures a defined transition in pixel output, while motion clarity also depends on picture timing, illumination persistence and processing.

When a display is driven to a new pixel level, its output does not necessarily reach that level instantly. Response time measures a specified part of that transition.

The measurement needs a starting level, destination level and threshold definition. A quoted number without those conditions gives only a partial description of motion.

Transition, delay and persistence are different times

Quantity	What it describes
Processing/input delay	Time before a new picture begins affecting the output
Pixel transition time	Time spent changing between specified light levels
Illuminated persistence	How long the picture’s light remains visible within its presentation
Source-picture interval	Time between samples created by the content

A “one-ms response” does not establish one-ms input-to-light delay. A display can buffer a picture before driving a fast pixel.

A fast transition also does not imply that the picture is illuminated only briefly.

A curve needs a defined measurement interval

Consider an original normalized transition from zero toward one:

L(t) = 1 − exp(−t / 2), with t in milliseconds

The two-millisecond parameter is a time constant, not a measured gray-to-gray result.

The model reaches 10% at approximately 0.210721 ms and 90% at 4.605170 ms. Its 10%-to-90% interval is:

t90 − t10 = 2 × ln(9) = 4.394449 ms

It reaches 99% at 9.210340 ms, so reaching 90% is not the same event as completing the entire asymptotic transition.

The equation is a teaching model. Real response depends on the display technology, source/destination levels, drive policy and temperature, and need not have this shape.

A fast initial crossing can hide overshoot

An implementation can drive a transition aggressively so it reaches the target earlier. If its light output then exceeds the intended level before settling, the moving object can acquire a bright or dark inverse trail.

Use a separate original sampled trace:

Model time	Normalized output
0 ms	0
2 ms	1.15
4 ms	1.05
6 ms	1

The early crossing of one does not mean all later samples are correct. The maximum overshoot is 15% relative to this unit target.

VESA’s ClearMR material treats motion blur and overshoot/undershoot as relevant visual behavior. A transition-time figure alone cannot describe those artifacts completely.

The trace above is not an overdrive setting or a measurement of a panel.

Eye tracking exposes another blur mechanism

Suppose an ideal pixel changes instantly, then holds a picture’s light for the whole 16.667-ms interval of an exact sixty-Hz output.

As the eye tracks an object moving at an original declared 960 screen pixels per second, the object’s held position moves across the retina during that interval. A simple uniform-hold estimate is:

960 × (1 / 60) = 16 pixels of tracking smear

If illumination lasts only four milliseconds under otherwise ideal conditions:

960 × 0.004 = 3.84 pixels

This model excludes the camera’s blur, pixel transition shape, eye dynamics and multiple image pulses. It explains persistence as a separate contribution, not a complete prediction of perceived blur or a ClearMR score.

Black-frame insertion develops how the illuminated interval can be shortened.

Movies include their own motion sampling

A 24-fps movie can have substantial camera-exposure blur and only twenty-four distinct source samples per second. A fast display cannot reconstruct motion information that was never supplied.

Judder concerns another relationship: unequal presentation durations can add a repeated rhythm to those samples.

For Emby playback, compare the same scene while identifying source cadence, output mode and display processing. A panel response difference, an interpolation change and an uneven repeat schedule can all alter movement, but through different mechanisms.