Realism vs Stylised Graphics in VR

By Xander Townsend | Misfit Media Editor

As someone who specialises in developing virtual reality games/experiences for psychology and cognitive research, there is a conversation I find myself having more than I'd like to admit. What graphic fidelity is more effective? Do people actually find a realistic environment more convincing than a stylised one? What type of graphics make a person believe an environment is as close to 'reality' as possible? Would you be more inclined to believe Half Life: Alyx or Job Simulator?

 My current thesis is titled 'The Effect of Graphical Styles on Eliciting Emotional Responses in Virtual Reality'. My research is focused on comparing two VR environments of exactly the same content and visual stimuli, but of completely different styles. The most interesting element of research in this area has concluded that people will treat a VR environment as if it were real as long as it abides by the ruleset it presents. A ruleset within this context refers to the physical functions of an environment, things like gravity, physical properties of matter (colliders), and even social conventions of characters. So, if the graphics are consistently stylised and function within the ruleset presented (whatever the ruleset may be), the human brain will adjust what constitutes as effective reality to the stimuli being presented in VR. Therefore, theoretically any style of graphics can be effective for simulation purposes as long as a player can comprehend and acknowledge its ruleset.

So, why is it important to realise that a stylised VR environment can be as effective, if not more effective, then a realistic environment? Imagine being able to develop a VR simulation program in less than half the time, for less than half the budget, with the ability to run at higher frame rates and port to mobile headsets more seamlessly. Stylised graphics also allow for devs to utilise mechanics and functions that are outside of the scope of 'reality' (for example, magic casting or telekinesis) but imbed those functions into the simulated ruleset so it's still considered 'real' by the player.

Another important aspect that can be avoided when it comes to VR 3D asset development is the 'uncanny valley' predicament. 'Uncanny valley' is defined as the phenomenon involved with computer-generated/humanoid depictions bearing near-identical resemblance to a human being arousing a sense of unease or revulsion in the person viewing it. This is a massive problem especially with AAA game titles and the attempt at realistic characters/facial expressions. If 3D artists can instead showcase a unique and expressive style of character development, uncanny valley can be completely avoided.

So, in summary, utilising stylised graphics allows for cheaper, less time consuming, more optimised development processes, with the ability to imbed magic systems that players perceive to be real within the environment. Basically you can create a reality where magic exists.

Nintendo's art team and their hyper stylised graphics may be onto something, but ironically they chose not to invest in VR. Isn’t that interesting?