Why Do Films Judder on a PC Monitor? 24fps, 3:2 Pulldown & Refresh-Rate Multiples Explained

｜ HK GPU Fix

重點摘要

Films run at 24 frames per second, but a 60 Hz display cannot divide 24 evenly, so it repeats frames unevenly in a 3:2 pattern. That uneven rhythm, not weak graphics hardware, makes slow camera pans judder. Displays running at an exact multiple of 24 repeat every frame equally and restore an even rhythm.

A film can look perfectly sharp, the sound can be flawless, and yet every slow camera pan seems to stumble. The picture does not freeze, and it does not break up. Instead, movement advances, hesitates, advances, hesitates, in a rhythm so regular that it almost feels deliberate. Many viewers blame the graphics card, the player, or the cable. In most cases, none of those is the cause. The cause is arithmetic: the film delivers 24 frames every second, and the display refreshes 60 times every second, and 60 cannot be divided evenly by 24. Something has to give, and what gives is the evenness of the rhythm.

What 24 Frames Per Second Actually Means

Cinema and most drama series are shot and finished at 24 frames per second, often written as 24p. Each frame is therefore meant to stay on screen for about one twenty-fourth of a second, roughly 41.7 milliseconds, before the next frame replaces it. That interval is the heartbeat of film motion. It is also slow by modern display standards. A viewer used to a 60 frame per second video or a high refresh rate game may already find film motion a little steppy, because 24 distinct pictures per second simply carry less motion information than 60. That steppiness is part of the look of film, and it is not a fault.

The problem this guide covers is different. It is not the steppiness of 24 frames per second itself. It is what happens when those 24 frames are stretched across a refresh grid that does not fit them, so that some frames stay on screen longer than others. The motion then stops being evenly steppy and becomes unevenly steppy, and the eye is far more sensitive to unevenness than to slowness.

The Display Refreshes on a Fixed Grid

A display running at a fixed refresh rate redraws the picture on a rigid schedule. At 60 Hz, it redraws 60 times per second, once every 16.7 milliseconds. At 120 Hz, it redraws 120 times per second, once every 8.3 milliseconds. At 144 Hz, it redraws 144 times per second, once every 6.9 milliseconds. The display does not know or care how many frames the film contains. It simply shows, at each redraw, whichever film frame is current at that moment.

If the film frame rate divides evenly into the refresh rate, every film frame gets shown the same number of times, and every frame stays on screen for the same duration. If it does not divide evenly, some frames must be shown more times than others, and frame durations become unequal. That inequality is the whole story of film judder on a monitor.

3:2 Pulldown: How 60 Hz Handles 24 Frames

Divide 60 by 24 and the answer is 2.5. A display cannot show a frame two and a half times, so the classic solution, called 3:2 pulldown and sometimes written 2:3 pulldown, groups film frames in fours and distributes ten refresh slots among them. The first frame is shown twice, the second three times, the third twice, and the fourth three times, or the same pattern starting from the other phase, three, two, three, two, depending on where the sequence begins. Ten refresh slots carry four film frames, and the cycle then repeats.

The consequence is straightforward. A frame shown twice stays on screen for two refresh intervals, about 33.3 milliseconds at 60 Hz. A frame shown three times stays on screen for three intervals, about 50 milliseconds. The ideal duration is about 41.7 milliseconds, so the picture alternates between frames that are too short and frames that are too long. A camera pan that should move at a steady speed instead moves in a repeating pattern of shorter and longer steps. That repeating unevenness is judder, and its regularity is its signature: it is not random, it does not come and go with scene complexity, and it appears most clearly in steady, slow camera movement, where the eye can track the rhythm.

Commercial films are often actually encoded at 23.976 frames per second rather than exactly 24, a legacy of the NTSC television era, which applied a 0.1 percent slowdown. For the purposes of this guide, the pulldown logic is identical, and the distinction changes nothing about the diagnosis or the remedy. It is mentioned here only so that a player reporting 23.976 does not confuse the reader.

When the Numbers Divide Evenly, the Judder Goes Away

The remedy follows directly from the arithmetic. At 120 Hz, 120 divided by 24 is exactly 5, so every film frame is shown five times, a pattern called 5:5 pulldown. At 144 Hz, 144 divided by 24 is exactly 6, so every frame is shown six times. Some displays and televisions offer a dedicated 24p mode that switches the refresh rate to 48 Hz, where every frame is shown twice, or 72 Hz, where every frame is shown three times. In every one of these cases, each frame stays on screen for exactly the same duration, the rhythm becomes even again, and the pulldown judder disappears.

It is important to state what even repetition does not do. Repeating a frame does not create new motion information, and it does not make a 24 frame per second film look like a 60 frame per second video. The film remains a 24 frame per second film, with the motion character that frame rate carries, including motion blur and a certain steppiness in fast pans. Even repetition removes only the extra unevenness imposed by the mismatched grid. A viewer who expects film at an even rhythm to feel as smooth as a high frame rate game will still be disappointed, and that disappointment is a misunderstanding of what film is, not a remaining fault.

The table also carries a warning that surprises many buyers of gaming displays. A higher refresh rate is not automatically better for film. A 120 Hz display gives film an even rhythm, while a 165 Hz display does not divide evenly by 24, and a 180 Hz display gives 7.5, which is also not whole. Only 240 Hz, among the common high refresh rates above 144 Hz, divides evenly by 24. For film viewing, the question is never how high the number is. The question is whether it is a multiple of the content frame rate.

The Same Test Applies to Other Frame Rates

The divisibility test is general, and two further frame rates make it concrete. Some European television and online content runs at 25 frames per second. At 60 Hz, 60 divided by 25 is 2.4, which is not whole, so 25 frame per second content on a 60 Hz display suffers the same kind of uneven repetition as 24 frame per second film. Content at 30 frames per second, by contrast, divides evenly into 60 Hz, with every frame shown exactly twice, so it has no pulldown judder at 60 Hz. That contrast is worth remembering, because it explains why one programme can judder and another, on the same display and in the same evening, can look perfectly steady. The display did not change. The arithmetic did.

This also means diagnosis must always begin by confirming the frame rate of the actual content being watched. Not everything labelled as a film or a drama is 24 frames per second. Some series and variety content are produced at 25, 30, or 60 frames per second. Assuming that every juddering programme is a 24 frame per second film will misdirect the diagnosis whenever the content frame rate is something else.

Repetition Is Not Interpolation

A separate technology is often confused with even repetition, and the confusion matters, because the two do opposite things to the picture. Frame interpolation, marketed on televisions as motion smoothing and sometimes described as a judder reduction feature, estimates and inserts brand new frames that never existed in the film, between the real frames. Motion becomes smoother, but the picture is no longer the film as made. The result is often described as the soap opera effect: film starts to look like video, and the estimation can go wrong, leaving halos or smearing around fast-moving edges.

Even repetition, by contrast, changes nothing in the picture. It shows each real frame more times, without inventing anything. A viewer who wants the film as the director made it should therefore choose an exact multiple refresh rate with interpolation switched off, rather than reaching for motion smoothing. Interpolation is a preference about how motion should look, and it can be enjoyed by viewers who like it. It is not a repair for pulldown judder, and it should not be presented as one.

On a PC, the equivalent of a television 24p mode is usually manual: set the display output refresh rate to a multiple of 24, such as 120 Hz or 144 Hz, in the display settings, or use a player or graphics frame rate matching feature where one is available. Not every display or player offers such a feature, and this guide makes no claim that any particular model does. The point is the principle: match the output rhythm to a multiple of the content frame rate, and keep interpolation out of the comparison while testing.

The Diagnostic Sequence, in Order

Run these steps in order. Each step changes exactly one thing, and the pattern of which changes matter is what identifies the cause. Changing several settings at once destroys the evidence.

Step 1: Establish which kind of stutter this is. Does the stutter appear as a regular, repeating hesitation during camera pans, only while watching video, while gaming stays smooth? That pattern points to a frame rate and refresh rate mismatch, and this guide continues. Random stuttering, stuttering across the whole system, or stuttering that also affects games points elsewhere first: the frame rate collapse checks in article 136, the shader compilation checks in article 161, or the hardware decoding checks in article 173. Do not force those problems into the pulldown explanation.

Step 2: Confirm the content frame rate. Use the player information or statistics display, where the player provides one, to confirm the frame rate of the programme actually being watched. Record whether it is 24, 25, 30, or something else, because the divisibility test in step 4 depends on it.

Step 3: Confirm the display refresh rate at the moment of viewing. Check the current refresh rate in the display settings, including advanced display settings. Note that a laptop or a display may change its refresh rate automatically for power saving, or through dynamic refresh rate behaviour covered in article 231, so the rate at the moment of viewing is what matters, not the rate the display is capable of.

Step 4: Do the division. Divide the refresh rate by the content frame rate. A whole number means an even rhythm. A non-whole number means uneven repetition is suspected. This single calculation separates pulldown judder from most other causes, and it costs nothing.

Step 5: Run the comparison test. Watch the same slow panning shot twice, once at 60 Hz output and once at 120 Hz or 144 Hz output, with motion smoothing switched off in both cases. If the judder clearly lessens or disappears at the multiple refresh rate, the pulldown explanation is confirmed. Then, as a second comparison, switch motion smoothing on and off at the same refresh rate, and notice the difference between the look of even repetition and the look of interpolation.

Step 6: Exclude the remaining layers. If changing the refresh rate changes nothing, look at decoding, covered in article 173, including hardware decoding of the codec in use and browser hardware acceleration, at streaming quality drops, where article 229 separates network problems from graphics problems, and at dual display interference, covered in article 240. A judder that survives every refresh rate change is not pulldown judder, whatever it feels like.

Boundary Cases Worth Knowing

First, variable refresh rate technology for games must not be written up as a cure for film judder. Variable refresh in games works by letting the display follow the rhythm at which the graphics card delivers new frames. Film playback usually passes through a fixed rhythm desktop composition path, which is a different route with a different rhythm. The game mechanism and the film playback path are not the same thing, and one must not be assumed to fix the other. Any guide that promises that switching on a game variable refresh feature will solve film judder is overclaiming.

Second, the display refresh rate is not the only number that matters. A streaming service may deliver a lower frame rate or a lower quality stream when the network struggles, and a player may drop frames when decoding is not hardware accelerated. Those produce irregular stuttering, not the regular rhythm of pulldown judder, and step 1 exists to separate them before any settings are changed.

Third, even a perfectly even rhythm does not remove the inherent character of 24 frames per second. Fast pans in film will still show stepping and motion blur, because only 24 distinct pictures exist each second. That is the original look of the film, not a fault to be chased, and no refresh rate setting can add motion information that was never shot.

Five Myths, Dismantled

Film stutter means the graphics card is too weak. Wrong. Playing a 24 frame per second film is a very light load. When the stutter is regular and confined to video playback, the cause is almost always the rhythm allocation between frame rate and refresh rate, not computing power. Upgrading the graphics card will not change the arithmetic.

Switching on interpolation fixes the problem. Wrong. Interpolation changes the picture by inventing frames, and it can introduce estimation artefacts. Viewers who want the film as made should choose an exact multiple refresh rate instead, which removes the unevenness without altering the content.

A higher refresh rate is always smoother for film, so 165 Hz must beat 120 Hz. Wrong. For 24 frame per second film, 120 Hz and 144 Hz give an even rhythm, while 165 Hz does not divide evenly by 24. The right question is whether the refresh rate is a multiple of the content frame rate, not how large the number is.

Changing the cable or the driver will fix it. Wrong. Pulldown judder has nothing to do with cable bandwidth, covered in article 190, or with driver versions, covered in article 210. Replacing hardware at random wastes time and money, and it leaves the arithmetic untouched.

All stuttering is judder. Wrong. Game stuttering, decoding frame drops, and network buffering each have their own rhythm and cause. Step 1 of the diagnostic sequence exists precisely to keep them apart, and a diagnosis that calls everything judder will recommend the wrong remedy for most of them.

When It Is Worth Having the Machine Checked

Most cases end at step 5 of the diagnostic sequence: a confirmed arithmetic mismatch and a refresh rate setting that follows from it. A check is worth considering when the tests point in conflicting directions, for example when the judder persists at every refresh rate, when it appears in games as well as in video, or when the display behaves oddly under other content. Those patterns point beyond settings to the graphics card, its cooling, or its configuration, and that is diagnosis work rather than settings work. Published inspection and repair pricing for this kind of work falls within the shop three published tiers, from HK$400 to HK$2,800, depending on what the diagnosis finds; no figure outside that range should be quoted for this category. Our shop is at Room 1209, Dali Plaza, Shek Mun. The telephone number is 2186 8010 and WhatsApp is 6558 6806. There is no Mong Kok shop, and there is no on-site service; diagnosis is done at the shop. Before bringing anything in, write down the results of steps 1, 2, and 4, the symptom pattern, the content frame rate, and the division result, because those three observations narrow the investigation more than any description of the stutter alone.

Frequently Asked Questions

Why does only the camera pan judder, while dialogue scenes look fine?

Because pulldown judder is a rhythm problem, and rhythm is easiest to see in steady, continuous motion. A slow pan gives the eye a constant movement to track, so the alternating short and long frame durations show up plainly. In a mostly still dialogue scene, there is little movement for the uneven rhythm to act on, so the same mismatch stays hidden.

Will a 120 Hz display make my films look like a high frame rate video?

No. A 120 Hz display showing a 24 frame per second film repeats every frame five times, which makes the rhythm even, but it adds no new motion information. The film still moves at 24 distinct pictures per second, with the motion blur and stepping that frame rate carries. Evenness is restored, but the film does not become a 60 frame per second video.

Is 144 Hz better than 120 Hz for watching films?

For the specific question of pulldown judder, no. Both divide evenly by 24, 120 Hz at five times per frame and 144 Hz at six times per frame, so both give an even rhythm. Other display qualities may differ between models, but the film rhythm itself is even in both cases, and neither is more correct for 24 frame per second content.

Why does one programme judder and another look steady on the same display?

Because the programmes may have different frame rates. A 24 frame per second film on a 60 Hz display suffers 3:2 pulldown, while a 30 frame per second programme on the same display divides evenly, with every frame shown twice. Confirm the content frame rate before concluding that the display behaves inconsistently.

Should I just leave motion smoothing on and stop worrying?

That is a viewing preference, not a repair. Motion smoothing invents intermediate frames and changes how the film looks, and it can leave halos around fast movement. Viewers who prefer the film as made should use an exact multiple refresh rate with smoothing off. Viewers who like the smoother look may keep it on, but they should know it is altering the picture, not restoring its original rhythm.

I changed the refresh rate and the judder stayed exactly the same. What does that mean?

It means the judder is probably not pulldown judder. Return to step 6: check hardware decoding, streaming quality, and dual display interference. A judder that survives every refresh rate change belongs to another layer, and continuing to cycle refresh rates will not find it.