An Overview of VR headsets

The discussion about virtual worlds and VR experiences is incomplete without understanding the hardware used to deliver these.

Let’s dive into the different concepts.

Device classes

Display devices are essential for viewing virtual worlds. Historically, VR worlds were CAVE-based. CAVE stands for cave automatic virtual environment.

The CAVE implementation featured 3D views, similar to the modern-day 3D cinema, displayed on three to six walls. User’s movements were tracked by sensors and users wore 3D glasses. Since 2010, improvement in tracking and display technology has led to a boom in VR headsets as consumer products.

Modern-day VR headset

Modern-day VR headsets come in many shapes and sizes.

These headsets consist of the following elements:
a) Display and optics to generate two separate views, one for each eye.
b) Graphics processor to generate the graphics shown as the two separate views.
c) Input controllers to capture input commands by the user.
d) Sensors to infer the position and movement as input.

Let’s explore these individually.

Displays and optics

VR headsets contain high-resolution displays which present the content of the virtual world. The headsets use a special type of lens called Fresnel lens.

The lens focuses on the content for the user. Most modern-day smartphones can work as VR headsets by adding optical elements as shown by Google Cardboard (see Figure 1).

Figure 1: Google Cardboard

Higher-resolution VR headsets are perceived to have better view quality and higher immersion. Higher display resolution creates more processing load on the device.

FoV (Field of View) (see Figure 2) is another important characteristic of VR headsets. Modern-day VR headsets support at least a 110-degree horizontal view.

Figure 2: Field of view (FOV) is the open observable area a person can see through their eyes or via an optical device

Refresh rate is the number of frames that the display can output per second. High refresh rates (75Hz or higher) alleviate motion sickness and reduce flicker-induced fatigue.

However, higher refresh rates also require powerful processors to deliver content in a timely manner. Thus the cost of the VR headset increases with increasing resolution, refresh rate, quality of optical elements and the FoV.

Graphics Processing Capabilities

Graphics processing is a resource-heavy task and the two main classes of VR devices, tethered and standalone, approach this differently.

Tethered devices like the Oculus Rift S (see Figure 3) and HTC Vive offload this computation to a dedicated computer with a powerful GPU through USB 3.0 and HDMI. However, the tether cables can be unwieldy and limit the user’s movements.

Figure 3: A typical tethered VR headset with controllers (Oculus Rift S with wireless controllers)

Standalone devices like Oculus Go, Oculus Quest and Google Daydream/Cardboard compatible smartphones use the onboard mobile GPU for graphics processing.

This mobile GPU is not as powerful as a PC-based GPU. Also, the unit has to rely on inbuilt battery power which is also limited.

Input Controls

To interact with the world, the device has to capture inputs provided by the user. Most devices support a combination of buttons, trackpads and joystick controls bundled into a hand-held unit that communicates wirelessly with the VR headset.

In addition to this, certain devices support the user’s gaze, head-pose and simple hand movements as inputs. Eye-facing infrared cameras, gyroscopes and accelerometers sense the inputs.

The input controller units are self-contained and support 3DoF or even 6DoF (see Figure 4).

Figure 4: Degrees of Freedom (DoF) in input and tracking detection: All 3 DoF-capable devices can detect the orientation-based changes to yaw, pitch and roll around the three axes.

6DoF-capable devices detect 3DoF input and also track limited movement along the three axes

Tracking Sensors

Tracking sensors go beyond the capabilities offered by input controllers. Tracking-based input communicates the orientation and position of a tracked entity to the VR headset.

The tracked entity can be hand-held controllers, the user’s fingers or hands, the user’s body or even the VR headset itself within a room. Tracking enables the user to perform more natural-feeling actions to interact with the virtual world.

The tracking sensors can be room-scale and external (e.g. the original Oculus Rift and HTC Vive) or internal (i.e. inside-out tracking) on the device (e.g. Oculus Quest). Inside-out tracking is very useful for the portability of untethered devices.

Software eco-systems

To create content for VR headsets, a creator needs to develop and deploy software customized for the target device. The two leading software ecosystems, SteamVR and Oculus (PC or Mobile), enjoy multi-platform support. Google Daydream supports VR on smartphones.

WMR (Windows Mixed Reality), OSVR (Open Source Virtual Reality) and proprietary platforms like PSVR (PlayStation Virtual Reality) also allow VR development. Unreal and Unity engines provide SDKs (Software Development Kits) and integration for all the above platforms.

However, these development tools are complex and involve a steep learning curve. On the other hand, the browser-based WebXR has been emerging as an alternative VR/AR tool and does not require an SDK to develop.

Frameworks such as A-Frame let developers create and deliver virtual worlds through WebXR-supported browsers on a range of devices. We will exclusively rely on A-Frame for exploring the concepts of the virtual world in this course.