Quick Read. Image stabilisation is a feedback system. Motion sensors detect small rotations or translations of the camera, a processor estimates how those movements would shift the image on the sensor, and movable optical elements or the sensor itself are displaced in the opposite direction. The goal is not to freeze the world. It is to keep the projected image from wandering because the photographer moved.
One-sentence answer: Image stabilisation works by measuring camera movement and moving part of the imaging system in the opposite direction so the scene stays more nearly fixed on the sensor during the exposure.
Camera Shake Is Mostly Angular
When photographers handhold a camera, the most damaging movement is often tiny rotation: pitch, yaw and roll. A small angular change can move distant scene detail significantly across the sensor, especially with long focal lengths. That is why telephoto photographs are particularly sensitive to shake.
Stabilisation systems place motion sensors—typically gyroscopic and accelerometer-based—inside the lens or camera. These sensors do not wait for blur to appear in the photograph. They measure movement as it happens.
Optical Stabilisation Moves Lens Elements
In lens-based stabilisation, a floating optical group shifts laterally so the light path bends in a compensating direction. The camera moves one way; the stabilising group moves the projected image the other way. The subject appears steadier on the sensor and often in the optical or electronic viewfinder as well.
This approach can be optimised for a particular focal length and lens design, which is especially valuable in long telephoto lenses where even tiny angular errors become visible.
IBIS Moves the Sensor
In-body image stabilisation—IBIS—moves the sensor instead of the optics. The sensor platform can shift horizontally and vertically and, in many systems, rotate or compensate across several axes. This lets stabilisation work with many lenses, including older lenses that contain no stabilising group.
Modern cameras can also coordinate lens stabilisation and sensor stabilisation. One system may handle some axes or frequency ranges better than the other. Together they create a larger control loop.
Why Stabilisation Helps at Slow Shutter Speeds
Without stabilisation, small hand movements accumulate during the exposure and smear detail. If the projected scene is held more nearly stationary, the shutter can remain open longer before camera shake becomes visible. Manufacturers express this advantage in “stops”, meaning the system may permit shutter times several powers of two slower than a non-stabilised baseline under test conditions.
Those ratings are not guarantees. Technique, focal length, stance, breathing, resolution, subject distance and the definition of acceptable sharpness all affect real results.
It Cannot Freeze the Subject
This is the most important boundary. A stabilised camera can produce a sharp wall at 1/8 second while a walking person becomes blurred. The system corrected camera movement; the person still moved through the exposure.
Stabilisation therefore buys freedom only when camera shake is the limiting factor. Sports, wildlife and children may still require faster shutter speeds because subject motion owns the problem.
Panning Requires a Different Kind of Correction
During panning, the photographer intentionally moves the camera along one axis while trying to suppress unwanted movement in others. Some stabilised lenses and cameras detect this pattern or offer dedicated panning modes so the system does not fight the intended motion.
This is an elegant control problem: stabilisation must distinguish desired movement from accidental movement.
Tripods Change the Situation
Older stabilisation systems could sometimes hunt or introduce tiny corrections when the camera was perfectly rigid, leading photographers to disable stabilisation on a tripod. Many modern systems detect stable support better, but behaviour varies. The safe practice is to understand the specific camera and lens rather than treating a twenty-year-old rule as universal.
Stabilisation Can Move the Sensor for Other Jobs Too
Once a sensor can move precisely, manufacturers can use that mechanism for more than shake correction. Some cameras shift the sensor by sub-pixel amounts across several exposures to improve colour sampling or resolution. Others use the mechanism for horizon correction or dust-removal routines. A stabilised sensor is therefore also a programmable imaging platform.
Three Experiments
- Static-subject ladder. Photograph one stationary object handheld at progressively slower shutter speeds with stabilisation off and on.
- Subject-motion test. Repeat the experiment with a moving person and observe where stabilisation stops helping.
- Telephoto comparison. Compare the same shutter speed at short and long focal lengths to see how angular camera movement becomes more visible.
Common Misconceptions
- “Stabilisation freezes motion.” It primarily corrects camera movement, not subject movement.
- “Five stops means every photograph can be five stops slower.” Ratings describe controlled test performance, not a universal guarantee.
- “IBIS makes lens stabilisation obsolete.” Lens-based correction can remain advantageous, especially with long focal lengths; coordinated systems can use both.
- “Stabilisation is only useful in darkness.” It also helps telephoto framing, video and any situation where camera movement limits image quality.
The Final Idea
Photography often asks the photographer to hold still. Image stabilisation asks the camera to help. It measures your imperfect steadiness in real time and moves part of itself against you—not because your hands failed, but because engineering found a way to make small human movement negotiable.