Rolling shutter is one of those technical terms that gets thrown around a lot. If you are starting out in cinematography, you may have heard the phrase “jello effect” or seen images where straight lines bend during a pan.
What you need to know is this is not a lens issue or a codec glitch. It comes from the way most digital cinema sensors work. Let’s go through it step by step so you can understand what it is, how it shows up, and what to look out for when picking a camera.
What’s the big deal about a shutter?
Let’s start with the second word in the name. What is a shutter?
In cinematography, a shutter is a device that controls how long the camera sensor is exposed to light. It essentially determines how much light hits the sensor. In modern digital cameras, the shutter is usually electronic.
However, traditional cinema cameras exposed film using a mechanical rotary disc shutter. The open portion of the disc was measured in degrees, known as the shutter angle. A 180-degree shutter meant the film was exposed for half of the frame interval.
If you want to know more about how shutter angle relates to shutter speeds, check out this article:
Traditionally you shot at 24 frames per second.
In other words, the camera shows 24 pictures every second. This means, each picture has 1/24th of a second exposed to light. This is 41.7 milliseconds. That’s the maximum time a frame can be exposed at that frame rate.
This means the shutter is fully open, or at 360-degrees. For cinematic motion, the shutter angle is 180 degrees (half a circle). This means the frame is now exposed for half the time.
360/180 = 2. This means the time the frame is exposed is 41.7/2 = 20.8 milliseconds.
If the shutter angle is smaller, the frame is exposed for even less time.
- At 90 degrees, it’s only open for a quarter of the frame time, about 10.4 milliseconds.
- At 45 degrees, it’s one-eighth, about 5.2 milliseconds.
If the shutter angle is bigger, the eye is open longer.
- At 270 degrees, it’s three-quarters, about 31.3 milliseconds.
- At 360 degrees, it’s the whole frame, 41.7 milliseconds.
Different shutter angles directly changed motion blur. A wider angle gave smoother blur but less crisp motion. A narrower angle gave sharper, staccato motion but also reduced light.
Cinematographers used shutter angle creatively. To know more, read this article:
What rolling shutter is
Most cinema cameras today use CMOS sensors and electronic shutters. The two primary types of electronic shutters used in video cameras are rolling shutter and global shutter.
A rolling shutter captures the image by scanning the sensor line by line, from top to bottom. Think of it like a scanner slowly moving across a document, capturing one row at a time.
The top of the frame is recorded first, and the bottom is recorded last. That time difference is called the scan time or readout time.
If nothing in the frame moves, you will not notice the difference. But if you pan the camera or if the subject moves, each row of pixels is exposed at a slightly different moment. The result is distortion. Check out the image at the top of this article and you’ll get the idea.
If we imagine each orange box is a pixel, our image is made of only three pixels (each on top of the other). As our hero climbs the stairs, the pixels are read one at a time, thereby causing a lag in the final frame.
Don’t let the simplicity of my example fool you. Rolling shutters are more complicated than this, but the analogy holds.
Since the image isn’t captured all at once, fast-moving subjects or quick camera movements can cause a few common problems:
- Jello Effect: Objects wobble unnaturally, as if they’re made of jelly. This often happens when the camera shakes or moves quickly.
- Skew: Straight lines (like buildings or poles) appear slanted or tilted when the camera pans quickly.
- Smear: Fast-moving objects appear stretched or smeared across the frame.
- Banding: A flash or strobe may only light part of the frame. Lightning might look like a strip across the image.
These artifacts can be distracting to the audience and cause problems for visual effects.
Numbers that matter
At 24 frames per second, each frame lasts 41.7 milliseconds. If a sensor takes 20 ms to scan from top to bottom, the bottom is nearly half a frame late. If it takes 8 ms, the delay is less than a quarter of the frame time. Shorter scan times mean cleaner motion.
Most labs measure scan time in milliseconds. The lower the number, the less rolling shutter you will see.
The common method is to film a strobe or spinning fan. The distortion pattern reveals how long the sensor takes to read the frame:
Sensor designs
Different sensor types change how rolling shutter behaves.
- Conventional CMOS: Slower, often 15–25 ms.
- Stacked and Partially-stacked CMOS: Faster because of extra circuitry, often under 10 ms. This is because the processor is “stacked” along the sensor.
- Dual-gain sensors: Capture highlights and shadows at once, balancing range and speed.
- Partial scan: Cropped or windowed modes read fewer rows, so they are faster.
Global shutter
Global shutter exposes all pixels at once. This removes skew, wobble, and banding. Straight lines stay straight, and flashes cover the frame.
The trade-off is lower dynamic range and higher noise. RED Komodo and Raptor X cameras use global shutter. The future is definitely global shutter, because no one likes rolling shutter artifacts. Here’s an example of rolling shutter skew in the same camera in different recording modes:
The HQ version is higher quality. In other words, more data to read in the same time.
Faster sensors need more processing lanes and power. That means more heat and cost. ARRI focuses on dynamic range. Sony pushes speed. RED chases resolution and frame rates. Blackmagic targets affordability.
Each choice creates different rolling shutter performance.
How much is too much?
Here is a simple scale that’s not an absolute but a good rule of thumb:
- Under 10 ms: Excellent. Artifacts are minimal.
- 10–15 ms: Acceptable. You will see skew in whip pans, and certain wobble effects in handheld work.
- 15–20 ms: Noticeable. Handheld action shows distortion that’s hard to watch on a large screen.
- Over 20 ms: Problematic. Best for static or slow shots.
Use these ranges to judge if a camera fits your needs.
Real-world numbers
Approximate scan times from labs and tests across the web (don’t hold me to it, I didn’t get these numbers myself so they could be off):
- Sony
- Sony FX3 / FX6: ~8.7 ms
- Sony FX9: 22.2 ms (it’s far worse being the more expensive camera!)
- Sony Venice 2: <3 ms
- Sony Burano: 16.9 ms full frame, 14.8 ms S35 (one of the reasons this camera was panned by many for its price)
- Blackmagic Design
- Blackmagic PYXIS 6K: 15–25 ms depending on mode (You get what you pay for)
- Blackmagic URSA Cine 12K LF: ~5.5 ms
- Blackmagic Pocket 6K Pro: ~20 ms
- ARRI
- Arri Alexa 35: 5.7 ms (16:9), 7.9 ms (Open Gate)
- Arri Alexa LF / Mini LF: ~7.4 ms
- Arri Alexa Classic: ~7.2 ms
- RED
- Red V-Raptor 8K VV: 8 ms (full frame), 6 ms (Super 35)
- Canon
- Canon C400: 9.5 ms (full frame), 6.9 ms (Super 35)
- Canon R5C: 15.5 ms
- iPhone 15 Pro: ~5 ms
Look out for camera tests for the camera model you’re interested in and the rolling shutter scan time will tell you a lot.
You still have to test the camera. Don’t just use the number to judge!
How to test for rolling shutter
Here are three simple rolling shutter tests anyone can do with just a lens and everyday subjects. No special gear required.
1. The Fence or Window Pan Test
Find a fence, railing, or window blinds – anything with strong vertical lines. Do a fast horizontal pan at different speeds.
Review the footage and look for vertical lines tilting into diagonals. The more tilt, the slower the sensor readout. The example I’ve given above is a pan test.
Watch this video for an explanation of how to test:
2. The Ceiling Fan Test
Point your camera at a spinning ceiling fan or bicycle wheel – anything spinning. Use a medium shutter angle so the blades are crisp.
Check if the blades look bent or warped. Straight blades should stay straight. Any curve or skew shows rolling shutter distortion.
3. The Flashlight Pulse Test
This is a harder test. In a dark room, flash a small bright flashlight quickly across the frame. Record with a normal shutter angle.
If only part of the frame lights up at a time – like a horizontal band – you are seeing how long it takes the sensor to scan. You’ll see these kinds of issues crop up in nightclub or concert scenes, or in lightning strikes, etc.
How to work around rolling shutter
What if you’re stuck with a camera with higher than 10ms rolling shutter scan time? You can manage rolling shutter on set in these ways:
- Avoid fast pans.
- Use stabilization for handheld work.
- Be careful with strobes and LED lights.
- Use crop modes or reduced resolutions to reduce scan time.
- Use lower frame rates if that’s an option.
- Use a lower mode (4K normal vs 4K HQ, for example).
- Always test before production. If you catch it in the shot, change the shot. There’s no getting away from bad rolling shutter artifacts. On a large screen, they are very obvious.
Closing thought
Rolling shutter is not a defect. It is a design choice. Each camera balances speed, range, and cost differently. Your role is to know what your camera does, test it, and use it within its limits.
Sub-10 ms feels natural. 15 ms is still usable. Over 20 ms is when you start to get concerned. Global shutter eliminates the problem so if you can get one of those you don’t have to worry about rolling shutter artifacts, that’s for sure.
And don’t forget: Test. Numbers are only half the story. Even worse, the people measuring it might have made a mistake. How would you know for sure?
So, test!
