If cameras are prioritized, one of the first specifications to control a new phone is image stabilization. You have probably been told that optical image stabilization (OIS) ranks highest, but this is not the case if you take many video clips. Instead, we would argue that electronic image stabilization (EIS) works better for video.
While OIS is superior in still images, the same cannot be said for videos. Videographers will benefit from OIS, true, but EIS actually does a better job of moving images in succession. Both stabilization methods help minimize camera body shaking, but due to their way of working, EIS is a better choice for videos.
How OIS Works
In phones, OIS works by suspending the lens over camera module with springs and electromagnets. The phone detects the frequency and amplitude of any shaking, then uses the electromagnets to match it and adjusts the lens angle. The lens moves in the opposite direction to the shake so that the image can again be centered on the image sensor.
All this happens before the image hits the image sensor, which means that there is no image degradation. OIS uses the phone's gyroscopic sensors to help detect shaking. The gyrosensors measure movements and send these data to a "micro center", which then moves the lens to compensate for the movement.
In short, the OIS mechanically moves the image sensor in response to any shakes detected in the phone. However, this physical movement takes time, which is a valuable product in high-definition video. Another disadvantage of OIS in the case of video is the movement of the lens relative to the image sensor may result in distorted perspective, which is consistent with a "jello effect" in videos.
] Electronic image stabilization is different. The system still detects camera shake with its sensors, but adjustments are made when the image hits the image sensor.
This is achieved by zooming the image until it is larger than the image sensor and scanning in the image for each motion, which is then counteracted by moving the image in the opposite direction. However, this method will reduce video quality and resolution due to zooming.
Another method is to use an oversized image sensor where the image only uses about 90% of the chip area. The additional 10% of space is then used to recapture the image without any "cropping".
Since the lens does not move with the EIS, the process of capturing each image in your video is never lowered by mechanical parts. The EIS just needs to adjust the image, resulting in a much smoother video.
In addition, EIS systems have the ability to predict switching of the next frame, which makes a big difference in video quality. As you move, the camera determines the direction you are moving, and if it determines that you are moving in a single direction (as you pan the camera over an object), it can adjust the next shift accordingly. This enables a much smoother video.
According to Google, when Pixel's law had to decide between OIS and EIS for the first Pixel, they agreed on EIS due to of its video performance. According to Google's camera product management, Isaac Reynolds:
"EIS and OIS have very different goals, so you cannot compare them to ask who is better / worse. OIS mainly improves low light photography by physically compensating for handshaking within OIS is primarily for photo and EIS are just for video. "
While Google added OIS to Pixel 2, it was able to achieve fantastic stabilization with only EIS on the original pixel thanks in part to their algorithms. This leads to another advantage of EIS, its ability to improve over time with software updates. Because algorithms are what determines how the system compensates for specific movements, since the OEMs take in more data, they can improve the function.
This article has been produced under the Gadget Hack's special coverage on smartphone-based video creator tips for filming and editing. Check out the entire Videography series.