VRHow / XR Technology / What is mixed reality?

What is mixed reality?

Short answer: Mixed reality (MR) puts digital objects into a view of your physical surroundings and gives those objects spatial behavior: they can stay on a real table, disappear behind a sofa, respond to your hands, or use your room as part of the experience. It is not simply “VR with a camera”; the important step is the device understanding enough of the space to make digital content share it.

Last checked October 11, 2026. This is a technology explainer, not a hands-on product review. Capabilities vary by device, software version, lighting, room and region; VRHow has not independently tested the devices described here.

MR is about relationships, not just see-through video

In everyday product language, “mixed reality” is used inconsistently. Microsoft’s documentation defines it broadly as a blend of physical and digital worlds, while its historical “virtuality continuum” places experiences between physical reality and a fully digital one. That broad definition can include both see-through holographic devices and immersive headsets that show the room through cameras. Microsoft’s overview explains the spectrum and its device distinctions.

A useful practical test is interaction. An AR overlay may put a label over a real object. An MR experience goes further when the content is anchored to a location, understands surfaces or boundaries, and reacts when you move around it. A virtual chessboard that remains on your coffee table is spatially placed; pieces that hide behind the table edge or collide with its surface are spatially aware. The boundary is not perfectly fixed, so treat the label as a description of behavior rather than a guarantee of quality.

How a mixed-reality headset builds the illusion

Most current consumer MR headsets combine several systems. Each solves a different problem, and a weakness in any one of them can make the experience feel unconvincing.

  1. Passthrough or optical see-through shows the real world. Camera-based passthrough captures the room and displays it on the headset screens; optical see-through lets light from the room pass through a transparent display. They are not equivalent: camera passthrough adds processing, exposure and latency considerations, while optical see-through does not give the same kind of camera view. For example, Meta’s developer documentation describes passthrough as a live camera feed with virtual content overlaid.
  2. Tracking estimates the headset and input pose. Head and hand/controller tracking tells the application where the user is looking and reaching. This is what makes a virtual panel stay put as you walk around it rather than slide with your head. For background, see VR tracking explained and six degrees of freedom (6DoF) explained.
  3. Environmental understanding supplies surfaces and depth. A system can detect or reconstruct floors, walls, furniture and other geometry. It can then place content on a surface, calculate collisions or hide a virtual object behind a real one. Microsoft’s spatial-mapping documentation describes triangle meshes attached to a world-locked coordinate system, while also warning that mapped surfaces can change and that an unscanned surface cannot be used for placement.
  4. Anchors preserve location. An anchor ties virtual content to a physical point or feature. Persistence can allow a screen or game board to return to a similar place in a later session, but “persistent” should not be read as infallible: room changes, tracking quality and platform support affect the result.
  5. Input and rendering make the interaction believable. Hand tracking, eye gaze, voice, controllers, haptics, lighting and spatial audio all contribute. A digital object can be correctly positioned yet still look wrong if its edges, shadows, occlusion or audio do not agree with the room.

These are not purely marketing concepts. Meta’s current developer documentation lists passthrough, depth-based occlusion, scene understanding and spatial anchors as separate building blocks. The OpenXR Working Group has also published spatial-entity extensions for plane and marker tracking, anchors and persistence, but describes implementations as rolling out across runtimes. That is evidence of an emerging portability layer—not proof that every headset supports every feature. Meta’s MR documentation and Khronos’s OpenXR announcement provide the relevant context.

What MR feels like in practice

Static MR might be a virtual monitor fixed to a wall, a board game on a desk, or a guided 3D model on a workbench. It can be useful without filling the entire room. Dynamic MR uses the room as part of the action: a game may route an enemy around furniture, or a training application may position instructions beside a real machine. The latter demands better mapping and safety design than simply floating a video window.

MR does not make virtual objects physically solid, and a headset’s cameras are not a safety certificate. A virtual barrier may fail to match a chair; a hand may be tracked poorly in shadow; and passthrough can make real-world text or distance harder to judge than unaided vision. Never rely on MR for collision avoidance, medical decisions, industrial safety or supervision of a child without the appropriate safeguards and human oversight.

MR, AR and VR: the useful distinction

ExperienceCore relationshipQuestion to ask
VRThe digital world replaces most of your view.Does the experience need the real room at all?
ARDigital information is added to the real view.Is a simple overlay enough?
MRDigital content is spatially anchored and responds to the environment.Does the room’s surfaces, depth or objects change what the content does?

For the shorter, side-by-side introduction, use VR vs AR vs mixed reality. The distinction is intentionally functional: companies may call a product “MR” even when a particular app only provides a floating 2D window.

Should you care about MR?

Choose MR when the real room adds value: a fitness app that needs awareness of your play area, a design review that benefits from life-size placement, or a game that turns a familiar room into a play space. Choose ordinary VR when isolation, visual control and a fully authored world matter more. Choose a simple AR overlay when you only need directions or labels and do not need room-scale interaction.

Before buying or installing an MR experience, ask:

The best MR experiences are therefore not the ones with the most virtual effects. They are the ones that make a clear promise about the physical environment, meet it reliably, and tell the user when they cannot.

Sources