The beginning of a new project is tremendously exciting. At first, it feels like the sky is the limit. Grand ideas for innovative game-play, compelling experiences and immersive visual styles abound. Freed from the constraints of a mature codebase, intellectual property and art theme, game developers can run wild.
It is during this heady time that the infamous divide between artist and engineer must be ruthlessly bridged. Inside an effective game developer, such a divide is only ever superficial. It may be expressed as animators sitting on one side of the office and engine programmers on the other, the intervening space regularly crossed for collaboration, socialisation and coordination of effort. However, the technical divide does remain.
That divide could be realised thus: The team agrees a project will adopt a level of graphical detail that allows fluid performance across devices with lower levels of processing power, such as consoles and tablets. Artists agree that this will necessitate a less realistic, more characterised art style. Engineers proceed to work on a renderer that prioritises more basic features over expensive ones.
Initial concept art is appropriately toned down from Crysis levels of visual acuity. Upon the team reviewing it, engineers agree that the style is achievable within envisaged resource restrictions. The first models are created and textured, and a mock up scene is rendered in an out-of-engine tool such as V-Ray. It looks great!
Oohing-and-aahing ensues as the team enjoys the prettiness of it all. The effective, cross disciplinary team notes that V-Ray has applied subtle, expensive, high quality effects such as ambient occlusion, complex reflections and extremely detailed shaders. They make sure the art does in fact look good when those effects are removed and simplified, as they will be when the assets are introduced to the game engine.
Another team that operates in silos may not realise that the art is being previewed in a way that is not necessarily achievable within engineering limits. A cohesive, consistent style is not achieved and the game does not attain its optimal artistic potential nor performance targets.
The ideal response attained by the former team does not necessarily require engineers and artists to literally be crossing over and sharing work. It could also be achieved by ensuring that each have a depth of talent and knowledge that transcends traditional disciplinary boundaries – An artist with knowledge of the engineering implications of ambient occlusion will be able to concept more achievable styles. An engineer with an eye for design will be able to write shaders that sit well with other rendering elements.
Traditional labels of ‘artist’ and ‘engineer’ are, in the best cases, simply indicators of primary priorities. They should not indicate limitations of work, or responsibility, or interest. At the point of project inception, blur the lines between them as thoroughly as possible. The resulting game will be better for it.
– Hugh