Epic’s Unreal Engine 5 has had a long gestation period. First unveiled about three and a half years ago, the best UE5 efforts – The Matrix Awakens and Fortnite – have come from Epic Games itself, with early third-party releases largely falling short of expectations. There’s a palpable sense that the engine hasn’t gotten a proper third-party workout, or at least one that doesn’t come with certain caveats and reservations.
Enter RoboCop: Rogue City. This first-person shooter from Polish developer Teyon packs the full suite of UE5 features – including Lumen GI and reflections, virtual shadow maps, and Nanite geometry – while targeting 60fps on console platforms as well. Initial impressions are positive, but how do the various techniques come together to produce high-quality final imagery – and do these features scale gracefully to home consoles, from PS5 and Series X to the lower-powered Series S?
I think it’s fair to say that Rogue City packs very impressive visuals, and even ranks as one of the strongest showcases for real-time rendering this generation. To understand it properly, I think it’s worth highlighting each implementation of the key UE5 technologies that it uses so well.
Lumen global illumination is used in RoboCop in place of the typical precomputed GI solution, as a way to process indirect lighting in real-time. Essentially, the software Lumen tech used here traces lighting against a very simplified version of the world – using signed distance fields, in addition to tracing in screen-space – in order to draw accurate indirect lighting and also reflections with relatively low cost.