- 70.0%Sample And Hold
- 30.0%Frame Hold Time @ 24 fps
Stutter is an artifact of motion that happens when a frame stays on the screen for too long. It can be bothersome while watching movies or other low-frame-rate content because the TV has to hold each frame on longer. Stutter has an inverse relationship with a TV's response time; the quicker the response time, the more stutter there is, while a slower response time results in less stutter.
Test Methodology Coverage
Our Stutter test was initially added as part of our 1.2 test methodology. It was completely revamped as part of our 2.2 test bench update, adding objective measurements using a high-speed camera both with and without motion interpolation enabled. This significant change means that most of our results are no longer directly comparable to TVs that are still on older methodologies. We still include the Frame Hold Time @ 24 fps results, though, and those are directly comparable to older reviews. Learn how our test benches and scoring system work.
| Test | 1.12 | 2.0 | 2.1 | 2.2 |
|---|---|---|---|---|
| Frame Hold Time @ 24 fps | ❌ | ✅ | ✅ | ✅ |
| Frame Hold Time @ 60 fps | ❌ | ✅ | ✅ | ❌ |
| Sample And Hold | ❌ | ❌ | ❌ | ✅ |
| Sample And Hold (Interpolation) | ❌ | ❌ | ❌ | ✅ |
| Interpolation Consistency | ❌ | ❌ | ❌ | ✅ |
When It Matters
Not everyone will notice stutter, but if you do, you'll notice that camera movement looks jarring and uneven during long panning shots. Stutter has an inverse relationship with response time, so it's often a trade-off between having little stutter or a quick response time. Since there are ways to help reduce stutter, having a TV with a quick response time is almost always a better choice.
Above, you can see an example of stutter. The TV on the left has a quick response time, so there's no motion blur behind fast-moving objects, but the shot looks more jarring as each frame is held longer. On the right, the response time is much slower, so there's a blur trail behind the ball, but the overall movement in the shot is smoother. You should easily see the difference if you pay attention to the plant in the foreground. Although response-time stutter is more pronounced on TVs with very fast response times, such as OLEDs, most modern TVs exhibit at least some stutter since response times have gotten pretty quick on TVs overall. You won't notice stutter during most scenes, but it's most noticeable during scenes with slow camera movements, like a slow panning shot across a landscape.
Our Tests
We measure stutter with the TV in its most accurate, post-calibration settings, with brightness set to maximum to reduce backlight flicker. We warm up the TV for 30 minutes, as response time can be sensitive to temperature, so a cold TV can have a slower response time and doesn't represent real-world usage. We use a Kron Technologies Chronos 2.1-HD high-speed camera and a 24p video of balls moving diagonally across the screen, and record a slow-motion video and an aggregate image showing how the balls moved across the screen. We repeat this test twice, once with all motion processing settings disabled on the TV, and then a second time with motion interpolation adjusted to output a 30 fps signal. We do multiple passes on each TV to find the best setting for each model.
Using image analysis software, we measure the position of the green ball as it moves across the screen and create a graph showing how it moves (shown above). The ball moving against a non-uniform background showcases the variety of transitions that might occur in real content. The blue line in the graph shows how a ball would move across the screen if it could move continuously, with an infinite frame rate and an instantaneous response time.
| LG G5 OLED | LG QNED92A | |
|---|---|---|
| Motion Interpolation Disabled |  |
 |
| Motion Interpolation Enabled at ~30fps |  |
 |
The aggregated image shows every ball position over the full length of the video, with the background removed. This approach removes the impact of response time, as we're taking the image after the ball has transitioned into each position. In the Stutter box, with motion interpolation disabled, the image should be very similar across all TVs. In the motion interpolation box, though, we can see more significant differences between each TV, depending on how the motion interpolation feature behaves on each model. Looking at the LG G5 and the LG QNED92A above, we can see that the QNED92A takes a more aggressive approach, generating up to five intermediate frames.
There are some limitations to this approach, however. The algorithm used to generate the aggregate image sometimes gets things wrong. In the examples above, we can see that there's a bit of noise on the top white ball, and the center of the green ball is slightly erased in some positions.
Sample And Hold
The Sample And Hold stutter test looks at the difference between the measured ball position and the expected reference trajectory. The faster the response time, the longer the frame will remain in a static state, increasing the difference between the measured position and the reference trajectory. In real-world terms, this means the ball appears to jump more from one position to the next, with less blur between frames, causing more noticeable stutter.
The relatively slow response time of any LCD TV creates phantom objects. As the ball transitions from one position to the next, it creates a secondary image between the two frames. This ghost image reduces the movement gap, so the object doesn't appear to jump from one position to the next as it would on an OLED.
We calculate the average sample-and-hold distance for each frame and the standard deviation. Since this version of the test is done with motion interpolation disabled, the standard deviation should be close to 0, and each frame should have the same sample and hold, but due to the margin of error of the test and the sensitivity of the equipment, there's some minor variation between frames. In real-world usage, there can be a greater deviation between transitions, as most LCD TVs have different response times at different gray levels. So you might see more stutter in a bright outdoor shot than you would with a dark scene.
Frame Hold Time @ 24 fps
Whereas the Sample And Hold metric above shows us how long the object is perceived to be in the same position, the Frame Hold Time indicates how gradual the transition occurs. For 24p content (like movies), the signal sends 24 frames every second. Since there are 1,000 milliseconds in a second, each frame is held on for 41.67 ms. To calculate the frame hold time, use the TV's average first response time, which is the time it takes for the TV to first change from one shade to another, without necessarily settling at that shade. We subtract the first response time from the frame time mentioned earlier, and that gives us the frame hold time.
Taking the Sony X90L as an example, it has a first response time of 6.1 ms. Subtracting that from the frame time of 41.67 ms gives us a frame hold time of 35.6 ms. This means that when watching 24p content, the image is static for approximately 85% of the time, resulting in noticeable stutter in scenes with slow camera movements.
Sample And Hold (Interpolation)
For this test, we repeat the Sample And Hold test above, but with the TV's motion interpolation features enabled to output as close as possible to a 30 fps signal. This is a low amount of motion interpolation that can reduce stutter without introducing too much soap opera effect, so it's a common middle ground that both home theater enthusiasts and average consumers can enjoy. We capture a new slow-motion video and aggregate image with the final settings, and again use image processing software to measure the ball position as it moves across the screen and compare that to the reference trajectory. The final number posted in the review is the mean of each measurement.
Interpolation Consistency
As part of the above test, we also look at how consistent the interpolation is. If the sample and hold time varies between each frame, it can create an inconsistent viewing experience that can be very distracting in some cases. Using the same data captured in the above test, we simply calculate the standard deviation of the results.
Why Do Modern TVs Have Stutter?
We mentioned above that stutter is directly related to a TV's response time, but why is that? It's all about the panel technology. Modern TVs all use a technique known as sample-and-hold to display an image. This means that once the pixels that make up a TV's image have drawn a frame, they remain static until it's time to draw the next one. It takes some time for a pixel to change from one color to the next; that time is known as its response time. The faster the response time, the longer that pixel remains fixed until it's time to change to the next color. The human eye doesn't expect motion to jump this way from one frame to the next; it expects motion to be fluid, so the longer the frame is held static on the screen, the more it'll appear to stutter.
This differs from older TV technologies like plasma and cathode ray tubes, which used short light bursts to produce an image instead. These older technologies are completely stutter-free, as they don't use sample-and-hold at all. This means that if you're upgrading from a plasma TV, for example, you'll likely be even more sensitive to stutter, and TVs with fast response times like OLEDs may actually appear to have choppier motion when watching movies and shows than your much older plasma.
You can learn more about stutter and motion in general through our series of videos on our Home Theater YouTube channel, starting with the first entry linked below.
How To Get The Best Results
While you can't change the TV's response time, there are a few settings that can help reduce stutter if it bothers you.
Motion interpolation
Motion interpolation is probably the best way to reduce stutter. This feature increases the frame rate of the video signal to a higher one. If you have a 120Hz TV and you're watching 24 fps content, increasing the frame rate can help reduce stutter because each frame isn't held on for as long. It doesn't change the response time, but it reduces the total time each frame remains on the screen. However, using motion interpolation too aggressively leads to the dreaded soap opera effect and distracting artifacts in the image.
Backlight strobing
Some LED TVs have a feature called backlight strobing, which is also known as black frame insertion (BFI) on OLEDs. LED TVs can flicker their backlight to help reduce the amount of blur trail behind each object. Although this isn't a direct way to decrease the amount of stutter, it decreases the amount of time each frame is shown on the screen. Overall, the point of the BFI feature is to improve the appearance of motion.
Judder removal
Judder and stutter, although they sound similar, aren't the same thing. Stutter is about the amount of time each frame is held on for after the pixels transition, while judder is an inconsistent frame time. There's also micro-judder, which affects the appearance of motion, but is caused by dropped frames during scenes with complex motion
Conclusion
Response-time stutter can be distracting while watching low-frame-rate content, such as movies. It happens when a TV has a quick response time, as the pixels transition colors quickly, but then the TV has to hold the frame for a few milliseconds before showing the next. We measure stutter using a simple calculation from the response time, as stutter and the response time have an inverse relationship. There are a few settings to help reduce the amount of stutter, like motion interpolation, but there's no perfect solution to eliminate response-time stutter on modern displays.