Input Lag vs Response Time: Two Numbers People Merge

Two numbers appear on every monitor listing, and they are routinely treated as the same thing. One is response time, which describes how fast a pixel changes. The other is input lag, which describes how long the whole chain takes before that pixel is told to change. Merging them produces confident conclusions that measurement does not support.

We logged both on the same panels and the two figures barely correlated. A panel with an excellent response time could have mediocre input lag, and vice versa, because they are properties of different parts of the system. Knowing which one you care about changes what you buy.

What does response time actually measure?

Response time is how long a pixel takes to move from one value to another, usually quoted as a grey-to-grey transition. It describes the liquid crystal’s physical speed, and it determines how much a moving edge smears. It says nothing about when the pixel was told to change.

The quoted figure is also a best case. Manufacturers pick a favourable transition and a favourable overdrive setting, which is why two panels quoting the same number can look completely different in motion. The number is real but it is not a description of the panel in use.

And what does input lag measure?

Input lag is the delay from a signal arriving at the monitor to the corresponding pixels changing. It includes the scaler, any processing the panel does, and the time spent waiting for the next frame. It is a property of the electronics and the frame timing rather than of the liquid crystal.

This is the number that matters for how a display feels to use. It is also the number that is almost never published, which is why buyers substitute response time for it and end up confused by the result.

Why do the two not correlate?

Because a panel can have fast crystals behind slow electronics, or slow crystals behind minimal electronics. One of our samples had a modest response time and excellent input lag, because it did almost no processing. Another had a headline response time and noticeably more input lag.

That second case is common. Features that improve picture quality, such as scaling, uniformity correction or dynamic processing, take time, and that time lands squarely in the input lag figure. A monitor can be better looking and slower to respond at once.

How much of input lag is frame timing?

A large part. A panel refreshes at a fixed rate, so an input that arrives just after a refresh begins waits nearly a full frame before it can be shown. Averaged over random arrival times, that alone contributes half a refresh interval, which is substantial at any refresh rate.

This is why raising refresh rate helps feel even when nothing else changes. It shortens the average wait, and it shortens it by an amount that is easy to calculate and easy to verify. It is often the largest single contributor to the total.

Does the port change the measurement?

Sometimes. Different inputs can take different processing paths, and we measured panels where one port was consistently faster than another. The difference was small in absolute terms but repeatable, and it is worth testing if you have a choice of connection.

Scaling is the usual cause. A signal that needs to be resampled goes through more processing than one that arrives at the panel’s native resolution and timing, and that processing costs time. Feeding a panel its native signal is the safest way to avoid it.

What does overdrive do to the two figures?

Overdrive changes response time by driving the crystals harder, which makes transitions faster at the cost of overshoot. It has essentially no effect on input lag, because it acts on the pixel after the decision to change has been made. That separation is worth remembering.

The practical consequence is that overdrive settings trade smear against artefacts without changing how responsive the display feels. Someone chasing responsiveness who turns overdrive to its maximum is adjusting the wrong control entirely.

How do you measure input lag at home?

The accessible method is a comparison rather than an absolute figure. Display a moving reference on two screens at once, photograph both with a fast camera, and compare where each one is. The difference between them is real even if neither absolute number is precise.

A simpler version uses a test pattern that responds to a button press and counts frames between the press and the change. Frame counting is coarse but it separates a fast panel from a slow one reliably, which is the question most people actually have.

Why is published data so inconsistent?

Because there is no standard method. Some sources measure to the start of a pixel transition, some to the midpoint, some to the end, and each produces a different figure from the same panel. Without a stated definition, two published numbers cannot be compared.

Conditions vary as well: refresh rate, resolution, port, overdrive setting and whether any processing was enabled. A figure measured with processing disabled describes a mode most buyers will never use.

Which number should decide your purchase?

If you care about how responsive the display feels, input lag is the figure that matters, and you will usually have to find it from an independent measurement rather than a listing. If you care about motion clarity, response time matters more, though the published version is unreliable.

Most people care about both and can only find one. In that situation, refresh rate is a reasonable proxy, because it moves the frame-timing component of input lag and it is always published accurately.

What is the honest summary of the distinction?

Response time describes the crystals; input lag describes the chain. One is a property of the panel material and its driving, the other of the electronics and the frame timing. They are both real, they are both measurable, and they answer different questions.

Keeping them separate makes listings easier to read and reviews easier to trust. The moment a source uses them interchangeably, everything else it says about timing should be treated with caution.

What follows once they are separated?

Does panel type change the relationship?

It changes both figures and the gap between them. Different crystal alignments have different natural speeds, and the electronics behind them differ as well. We measured panel types where the response figure was excellent and the input lag ordinary, and others where the reverse held.

That is another reason the two numbers must be read separately. A generalisation that holds for one panel type will not transfer to another, and anyone comparing across types by a single figure is likely to be misled.

Is there a way to reduce input lag directly?

Sometimes, through a reduced-processing mode. Where a panel offers one, it usually measures faster, and the cost is some image processing. Testing whether you can see the difference on your own content is the only way to decide.

Feeding the panel its native signal helps as well, because it removes the scaling step. That is free, and on some panels it is worth more than any mode change.

What should you do with both numbers?

Use response time to judge motion clarity and input lag to judge responsiveness, and never substitute one for the other. They describe different parts of the system and conflating them produces decisions that do not match what you experience.

Where only one is available, prefer refresh rate as a proxy for the other. It is always published accurately, and it moves the frame-timing part of input lag in a predictable way.

How should you weigh the two when reading a review?

Find out which one the review measured, then ask whether it is the one you care about. A review focused on motion clarity will rank panels differently from one focused on responsiveness, and both can be correct for their own question.

Where a review covers both, prefer it. Two figures from one rig describe a panel far better than the best figure from each of two rigs, because the comparison is then meaningful.

Does a faster panel always have lower input lag?

No. The two are separate properties, and a panel with a very fast response figure can still have ordinary input lag if its electronics add processing. We measured exactly that combination more than once.

The reverse also happens, which is why relying on response time as a proxy for responsiveness produces wrong conclusions. Measuring or finding the input lag figure is the only way to know.

Does refresh rate appear in either figure?

Not in response time, which describes the crystals and is independent of how often frames are drawn. It does appear in input lag, because the wait for the next frame is part of that interval and it depends on the rate.

That asymmetry is another reason the two numbers should never be merged. One is a property of the panel material, and the other partly of the timing you chose.

The next question is how much of the input lag figure belongs to the display and how much to the machine feeding it. Separating those two is the natural continuation, and it usually reveals that the display was not the largest contributor after all.