What this calculator does
Frame rate is the wrong unit for thinking about smoothness, because it is not linear. Going from 30 to 60 fps saves 16.7 milliseconds per frame, while going from 120 to 150 saves only 1.7. The first is transformative and the second is almost imperceptible, and the frame rate numbers give you no clue that this is the case.
Frame time is the honest unit. It converts the comparison into milliseconds, which is the same unit as your input lag and your network latency, so you can finally add them up and see which part of the chain is actually costing you.
The formula
Frame time is 1000 ÷ frames per second, in milliseconds. The refresh interval is the same calculation on your monitor refresh rate. Because a monitor can only show a new image when it refreshes, the frames actually displayed per second is the lower of the two figures, and anything your card renders beyond that is discarded unless adaptive sync is doing something about it.
| Term | Meaning |
|---|---|
| Frame time | Milliseconds between one frame and the next. |
| Refresh interval | Milliseconds between monitor refreshes. A 60 Hz panel refreshes every 16.67 ms. |
| Frames displayed | The lower of your frame rate and your refresh rate. |
| Wasted frames | Frames rendered that the monitor never shows. |
The inputs explained
| Field | What to enter |
|---|---|
| Frame rate (fps) | Your average frame rate. If you have a 1% low figure, run that too, since the lows are what you actually notice. |
| Monitor refresh rate (Hz) | Your monitor refresh rate in hertz. Check it in display settings rather than assuming, as panels often ship set to 60 Hz. |
| Session length (min) | How long you play, in minutes. Used only for the total frames figure. |
When to use it
Deciding whether an upgrade is worth it
Convert both frame rates to frame time before comparing. A jump from 45 to 60 fps saves 5.6 ms per frame; one from 120 to 144 saves 1.4 ms. The second sounds bigger in fps terms and is four times smaller in the unit that matters.
Working out if your monitor is the bottleneck
If your frame rate is well above your refresh rate, the wasted frames line tells you how much work is being thrown away. That is the case for either capping the frame rate or buying a faster panel.
Adding up your input lag
Frame time is one term in the chain alongside input polling, display response and network latency. Having it in milliseconds lets you add it to the others instead of guessing at its share.
Worked examples
Every figure in the tables below is produced by this page’s own calculator at build time, so the numbers and the tool always agree. Select any row to load that scenario.
What does each frame rate cost in milliseconds?
Frame rate climbs steadily. The time saved each step does not.
| Frame rate | Frame time | Frames displayed per second | Frames wasted per second |
|---|---|---|---|
| 30 fps | 33.33 ms | 30 | 0 |
| 45 fps | 22.22 ms | 45 | 0 |
| 60 fps | 16.67 ms | 60 | 0 |
| 90 fps | 11.11 ms | 60 | 30 |
| 120 fps | 8.33 ms | 60 | 60 |
| 144 fps | 6.94 ms | 60 | 84 |
| 240 fps | 4.17 ms | 60 | 180 |
How much does the monitor refresh rate matter?
The frame rate is fixed here, so only the panel changes.
| Refresh rate | Refresh interval | Frames displayed per second | Frames wasted per second |
|---|---|---|---|
| 60 Hz | 16.67 ms | 60 | 84 |
| 75 Hz | 13.33 ms | 75 | 69 |
| 120 Hz | 8.33 ms | 120 | 24 |
| 144 Hz | 6.94 ms | 144 | 0 |
| 240 Hz | 4.17 ms | 144 | 0 |
Questions
Is 60 fps enough?
For most single-player games, comfortably. The honest way to answer it is in frame time: 60 fps is 16.7 ms per frame, which is below the threshold most people notice in slower games, and above it for fast competitive ones.
Why cap my frame rate below my refresh rate?
With adaptive sync, capping a few frames below the refresh rate keeps you inside the variable range and avoids the latency penalty of hitting the ceiling. Without it, uncapped frames above the refresh rate just produce tearing and heat.
Does frame time tell me about stutter?
Average frame time does not. Stutter is variance, so a run of 6 ms frames with an occasional 40 ms one feels bad despite a good average. Look at 1% low figures and put those through here as well.
How does this relate to input lag?
Frame time is one component. Your total lag also includes input polling, game logic, render queue and display response. Having the frame contribution in milliseconds lets you compare it with the rest instead of guessing.
The other half of aim consistency is sensitivity, which cm per 360 makes comparable across games, and time to kill puts the milliseconds in context.