A 360 camera is a camera that records the entire sphere of the surrounding scene, 360 degrees horizontally and 180 degrees vertically, using two or more ultra-wide or fisheye lenses whose images are stitched into one continuous spherical frame. Among video cameras, it stands apart because it does not frame a shot at the moment of capture: the viewer, or the editor later, chooses where to look.
The output is usually stored as an equirectangular video, a 2:1 flat map of the sphere, which players wrap back onto a virtual sphere for viewing on screens, phones or headsets. Consumer models are compact, often resembling an action camera, while professional rigs combine many synchronized sensors.
How a 360 camera captures the full sphere
Most consumer 360 cameras use two back-to-back fisheye lenses, each covering slightly more than a hemisphere, often around 190-200 degrees. The overlap between the two images gives the software matching detail to align and blend them along a stitch line that runs around the camera body.
Stitching combines lens calibration, geometric remapping and blending. Each fisheye image is projected onto the sphere using a model of its lens, then the overlapping zones are aligned and feathered together. Optical flow stitching, used in higher-quality pipelines, analyzes motion and depth in the overlap to reduce visible seams on nearby objects.
| Component | Role | Quality factor |
|---|---|---|
| Fisheye lenses | Capture overlapping hemispheres | Sharpness and flare control at extreme angles |
| Image sensors | Record each lens image | Size and dynamic range |
| Synchronization | Aligns exposures in time | Prevents seams on moving subjects |
| Stitching software | Merges images into a sphere | Seam visibility and parallax handling |
| Stabilization | Keeps the horizon level | Gyroscope data drives reframing |
Resolution and image quality in 360 video
Resolution figures for 360 cameras are easy to misread. A 5.7K or 8K spherical file spreads its pixels over the whole sphere, so any single view the viewer actually sees contains only a small part of them. A headset view of roughly 90-100 degrees uses about a quarter of the horizontal width, which means that effective viewing resolution is far lower than the headline number.
| Spherical resolution | Typical frame size | Detail in a 90-degree view | Typical use |
|---|---|---|---|
| 4K | 3840 x 1920 | Roughly SD-like sharpness | Older consumer cameras |
| 5.7K | 5760 x 2880 | Below Full HD sharpness | Current consumer standard |
| 8K | 7680 x 3840 | Close to Full HD sharpness | High-end consumer and prosumer |
| 12K and above | Multi-sensor rigs | Above Full HD sharpness | Professional VR production |
Small sensors and extreme lenses also limit dynamic range and low-light performance. Because the camera sees both the sun and the shadowed side of the scene at once, highlight clipping and noise are common, and many models offer log recording or HDR modes to cope with the contrast.
Spherical, reframed and stereoscopic outputs
The same 360 recording can be delivered in different forms, which is one of the main reasons the cameras are popular. The choice determines how the audience experiences the footage.
| Output | Description | Best suited to |
|---|---|---|
| Monoscopic 360 video | Single spherical image viewers can pan around | Web players, tours, documentation |
| Stereoscopic 360 video | Separate spheres for each eye, giving depth | VR headsets, immersive storytelling |
| Reframed flat video | Editor chooses a standard 16:9 or 9:16 view | Social media, action sports |
| Tiny planet and special projections | Stylized remaps of the sphere | Creative and promotional clips |
| Still 360 panoramas | Single spherical photographs | Real estate tours, mapping |
Reframing, also called overcapture, has become a major use case. The camera records everything, and the editor later animates a virtual camera inside the sphere, choosing angles and pans after the fact. Combined with horizon-locked stabilization, this produces smooth footage even from handheld or helmet-mounted use.
360 camera compared with other capture methods
A 360 camera is one of several ways to record wide or immersive imagery. Each method trades field of view, image quality and workflow complexity differently.
| Method | Coverage | Strength | Limitation |
|---|---|---|---|
| 360 camera | Full sphere in one capture | Simple, framing chosen later | Lower detail per view, stitch seams |
| Action camera | Wide single view | Rugged, high detail in frame | Must be aimed during capture |
| Panorama stitching from stills | Any coverage, often full sphere | Very high resolution | Static scenes only, slow capture |
| Multi-camera VR rig | Full sphere, often stereoscopic | Highest quality and depth | Costly, complex stitching |
| Cinema camera with fisheye lens | Hemisphere or less | Large sensor quality | Needs a second capture for the full sphere |
Shooting techniques and common pitfalls
Working with a 360 camera changes basic production habits, because there is no behind the camera. Crew, lights and microphones are visible unless they are hidden or removed in post-production.
- Parallax at the stitch line: subjects closer than about 1 m can break or double at the seam; keep important action facing a lens rather than the side of the camera.
- Visible support: use a thin monopod or invisible selfie stick aligned with the lens gap so it disappears in stitching.
- Camera height: placing the lens around eye level, about 1.5-1.7 m, feels natural in headsets.
- Movement sickness: fast or tilted motion in headset viewing causes discomfort; keep the horizon level and movement smooth.
- Lens care: the protruding fisheye front elements scratch easily; use lens guards and clean them before each take.
Audio also deserves attention. Spatial audio formats such as ambisonics record sound from all directions, so that what viewers hear rotates with where they look, which strongly increases immersion compared with ordinary stereo.
The idea in brief
A 360 camera records everything around it at once: in front, behind, above and below. It usually has two very wide lenses pointing in opposite directions, and software joins their pictures into one round image.
When people watch the video, they can drag the picture or turn their head in a headset to look anywhere. Alternatively, an editor can pick the best view afterwards, so nobody has to aim the camera correctly while filming.
A practical scenario
A mountain bike guide wants promotional clips of a downhill trail but cannot ride and aim a camera at the same time. The guide mounts a consumer 360 camera on a thin pole attached to the backpack, about 40 cm above the helmet, recording 5.7K spherical video with horizon locking enabled.
In editing, the guide reframes the footage into 16:9 clips for a website and 9:16 clips for phones, choosing a following third-person view on fast sections and a forward view through tight corners. The pole vanishes in the stitch line, so the footage looks as if it was filmed by a drone close behind the rider.
From one descent, the guide produces several differently framed clips without reshooting, and the smooth, level horizon keeps the footage comfortable to watch despite the rough terrain.