Unless you’ve been living under a rock for the past month, you’ve probably both heard and gotten extremely excited over one of the biggest announcements within the sim racing scene in recent memory – Sector 3 Studios and their sister team SimBin UK will be moving away from the trusty powerplant created by Image Space Incorporated, which has fueled all of their simulators dating back to their inception as companies, in favor of adopting the fourth iteration of the Unreal Engine for future projects. The announcements have always been supplemented with an array of proof-of-concept photos depicting RaceRoom Racing Experience assets within the Unreal Engine, and of course these are then promptly eaten up by a large portion of the sim racing community due to the enhanced visual quality compared to what sim racers are traditionally accustomed to with these types of games.
However, I’d really like to know why people aren’t asking basic questions about this announcement, as even the most preliminary investigation efforts in regards to information on Unreal 4 uncovers a much larger story developing right under our collective noses – both SimBin UK, as well as Sector 3, might actually be in big trouble when it comes to the future of their software, and these announcements are possibly part of a last-ditch attempt to bail themselves out of a very precarious situation.
I mean, let’s take a step back and look at this from a very general standpoint; here you have a group of passionate sim racers who have been working with a dedicated racing simulator engine for a little over a decade – an engine they have refined and tweaked on their own for years upon years in the pursuit of realism – suddenly dropping almost everything except the 3D models in favor of starting from scratch with the building blocks typically reserved for first person shooters and adventure games that occasionally make use of a lighthearted vehicle component.
Is this not ridiculously out of place? Why is everybody just going along with this and refusing to ask serious questions?
Time to go on a bit of a journey with this one.
Above, I’ve inserted footage of someone’s little drift game they’ve built within UE4; a tech demo of sorts to show off quasi-realistic vehicle physics. Passing over the fact that it’s someone’s pet project they’ve messed around with in their spare time, let’s directly address two major themes that will undoubtedly arise as you take a gander at the video clip:
The first, is that it doesn’t look all that great; as if someone made a custom map mod for one of the Just Cause games, or quickly threw a bunch of converted content and/or community mods into Grand Theft Auto 4, recording their escapades on a third party Ebisu track. This is the Unreal Engine 4 in action – there’s excessive bloom, depth of field, and motion blur that makes it look like those shitty Xbox 360 Kinect games we all had to suffer through at one point, or Triple-A titles that are simply trying too hard on the visual aspects. This is the last thing sim racers have been asking for.
Second, despite this guy’s best efforts to create a semi-realistic racing demo within UE4, the car floats around as if it’s making use of a third party handling mod for one of the PC Grand Theft Auto titles. Now sure, simplified, approximated physics may be fine for open world sandbox games where the driving element also has to co-exist with platforming, exploration and shooting elements, but GTR3, and eventually RaceRoom Racing Experience according to Sector 3, will supposedly be making use of this engine. We’re talking a genre of games where people spend years learning how to manage their tires in an endurance racing format, maximizing every last bit of the track for the final tenth of a second advantage on their opponents, or commanding the vehicle to the absolute edge of the tire – and holding it there.
What I see above does not indicate this kind of gameplay experience will be able to thrive in Unreal Engine 4; I’m instead looking at an engine where the driving portion is just one aspect of the entire sandbox. It seems designed for open-world games where there is a driving element sitting off to the side – so war games that make use of military vehicles, or open world free-roaming sandboxes – as opposed to a hardcore racing simulators that focus solely on the art of performance driving.
Next, I’d like to present our readers with gameplay footage of Moto Racer 4, the only dedicated racing game currently available for purchase powered by the fourth iteration of the Unreal Engine, as the other title is like this weird battle kart hybrid reminiscent of Mashed on the PS2. As you can tell by the title of the video, “worst game of 2016” in all capital letters, it’s obviously not very good at what it’s trying to be – a simple-minded arcade racer. An entire team of developers cannot build a budget-priced motorcycle racer with the Unreal 4 Engine, yet we’re supposed to expect this same engine to power a highly advanced racing simulation focusing on ultra-high tech GT3 and prototype entries?
To elaborate on this topic a bit further and sort of drill home the point I’m trying to make for people not quite sure where I’m going with this one, Wikipedia has a phenomenal list of every single game either released or currently in development that’s powered by UE4, and there’s a very ominous trend that you can check for yourselves if you feel like killing five minutes of your day – race cars are nowhere to be found on this list. There are shitloads of adventure games, fighting games, shooters, role playing games, and even a pretty big list of survival horror titles, but auto racing titles are practically non-existent. To get any sort of driving game resembling a proper simulator, you have to go back to Unreal Engine 3 and take a peek at the indie rock crawling title Off-Road Drive – and even that game is unable to depict high speed vehicle behavior in a realistic fashion, with the commentator in the linked video noting the car comes to an almost instant stop the moment you let off the throttle, not to mention several instances of what’s quite frankly bizarre weight transfer and other miscellaneous bullshit.
Gee, maybe this engine isn’t very good for driving games?
KartKraft was first announced in September of 2011. At the time of this entry on PRC.net, it’s now February of 2017, and KartKraft has failed to materialize in any meaningful fashion, with major, non-abrasive sim racing outlets such as RaceDepartment openly questioning what has happened to the game. MotoRacer 4 launched in October of 2016 to abysmal critical reception. Obliteracers was a no-name arcade racer that scored a lowly 60% on Steam, and saw an embarrassing peak of 67 active users all playing the game at once. That’s the entire history of Unreal 4 powered driving games.
Three titles, one of which is vaporware and hasn’t been released to the public in seven years, one of which is a shitty battle kart racer, and the last is described as the worst racing game of 2016.
Despite this obvious red flag of developers unable to harness the power of Unreal 4 to create a captivating racing game, as well as the engine’s complete lack of history producing racing-oriented titles to the point where developers are using the engine to create anything but racing games, Sector 3 Studios, as well as SimBin UK, have decided that this engine is the future of their ultra-hardcore racing simulators, dropping an engine which was specifically built to function as a race car simulator in the process.
Sound absolutely ridiculous? Let’s keep going.
Several interviews with Chris and Allan Speed of the Sector 3/SimBin family have warranted lengthy explanations as to what sim racers can expect from both GTR 3, as well as RaceRoom Racing Experience in the future once both products make the switch to Unreal 4. Thanks to the excellent work of Paul Jeffrey over at RaceDepartment, we’re able to see the brothers making very concrete, specific statements regarding the future of simulators published under the Swedish umbrella of simulation studios, and upon dissecting their answers, it’s extremely cliche to say and I really need to stop using this phrase so often, but I’m left with much more questions that I’m shocked my fellow sim racers aren’t asking.
So let’s go through a few standout quotes – and please keep in mind I’ve chopped some up and cleaned up the grammar to focus on the core topic at hand:
In GTR3 we will have much improved car damage over what is industry standard at the moment, as well as a new particle system and a modern UI system.
First of all, the “new particle system” isn’t a new particle system; it’s simply Unreal’s particle system.
Second, manufacturers have placed more licensing restrictions on car damage than what we were accustomed to a decade ago – and this is something that teams like Kunos Simulazioni and Turn 10 have been very open about discussing with their users in recent times to kind of calm the complaints about very simplistic damage models. Car companies simply won’t allow heavy damage in modern driving games, so I’m confused as to how these guys intend to step it up in the fashion they’re talking about. The moment you sign a license agreement with, say, BMW, Ferrari, Porsche, or Lamborghini, they call the shots – and traditionally, they say “you can’t show our cars with massive damage.” So these guys are sitting here making claims they most likely won’t be able to back up once licensing deals are on the table, waiting to be signed.
With Unreal, we will have aquaplaning and water displacement that will affect the handling of the car, as realistic as possible here. Puddles will build up around the track, rain will occur on different parts of the tracks, and there will be as much variability in the weather as we can achieve.
While it’s true that Unreal 4 supports a vast array of weather effects, these effects are purely visual – elaborate particle systems, if you will. Their inclusion does not, by default, affect how a player traverses through the game world by making the terrain more difficult to retain traction on. A developer can manually adjust basic friction and overall grip variables for the surface of the game world – or race track, to be more specific – in an effort to fabricate what a wet surface would feel like in a race car, but these advanced dynamics alleged to be introduced in GTR 3, where cars naturally hydroplane over specific puddles of water that form on the racing surface… Unreal 4 doesn’t do that.
Yes, puddles do form, and you can inject a variety of rain effects into your game which react naturally to physical objects in the environment – for example, matches played in the rain in Rocket League – but they don’t actually do anything to the racing surface or your car’s behavior. It’s an immensely detailed water splash animation; or in the case of a puddle, displacement animation.
In fact, merely running a Google search on “Unreal Engine Aquaplane” only brings up about a page of links re-directing back to the GTR 3 interview originally conducted on RaceDepartment. So we’ve got a developer saying they’ve switched to the Unreal engine partially to make use of its advanced aquaplaning simulation and treacherous wet weather driving conditions, when this feature has never actually existed in any iteration the Unreal engine and doesn’t warrant any relevant search results of people even talking about it, aside from Sector 3 and SimBin themselves.
We will mirror the full weekend structure, rules and regulations, types of classes and individual driver strengths of that series, different weather attributes, day/night cycles, animated pit stops everything you would expect from that series will be included.
Again, Unreal 4 supports the ability for a talented group of content creators to inject day/night cycles into the sandbox, and even compose their own weather patterns if they’re wanting to go to that extent. That’s not the issue here. However, these elaborate visual effects and atmospheric conditions are not tied into vehicle physics; injecting a fancy third party weather plug-in will not suddenly make your vehicle’s engine generate more or less horsepower based on the humidity, air temperature, and elevation – which I assume is what the duo mean by “different weather attributes.”
In fact, vehicle editing as a whole is extremely basic, a far cry from the elaborate HDV files isiMotor enthusiasts are used to obsessing over. So not only has the complexity of vehicle editing been reduced dramatically compared to the simulation-oriented engine Sector 3 and SimBin plan to depart from, believing you can somehow tie weather into vehicle performance is nothing more than a pipe dream.
Additionally ,we are going to improve on the physics found in RaceRoom Racing Experience, and take over the best bits of the audio from RaceRoom as well.
The Unreal Engine is almost a closed toolbox of sorts; you simply can’t take part of one game engine, and throw it into another without years upon years of work – as the engine was never designed for that to begin with. It has been designed so you open up the toolbox and create a game with Unreal. So how are you going to improve on physics created in a purpose-built race car simulation engine by starting from scratch in a toolbox where car physics are greatly simplified and considered only a fraction of the entire experience, and still manage to have an internal playable demo within six months time created by a team of anywhere from four to seven individuals?
Unreal is an engine used to create first person shooters. The AI has been built for humans and bots, the collisions are built for characters, and the ballistics are built for guns. You can’t remove these characteristics from the core engine, so it’s no wonder that no major racing simulator – or even the odd arcade racer, for that matter – has been constructed using Unreal as a base. The ones that try, fail spectacularly.
After looking at some of the stuff above, and the overall storyline fueling this blog entry – a hardcore racing simulator developer announcing they’re dropping a trustworthy sim engine in favor of something that traditionally powers first person shooters – one important question still remains: Why are Sector 3 Studios and SimBin UK choosing to pursue this route when it’s absolutely nonsensical for them to do so?
Though we’re obviously not privy to all the inner-workings that would undoubtedly help flesh out the conclusion a story like this, one thing we do know for certain is that ISI’s baby, rFactor 2, is now living comfortably under Studio 397 banner, with Marcel Offermans and Luminis in charge of the project to a certain degree. I’m under the impression that there was more to this deal than most originally thought, as it’s extremely suspicious that after Sector 3 and SimBin have spent over a decade using the isiMotor engine in all of their software, merely months after rFactor 2 has changed hands and there’s been a shakeup of sorts at ISI, one of the main developers powered by ISI simulator technology is suddenly expressing their desire to jump ship to a first person shooter engine. The timing of that is a little too impeccable to be a coincidence.
You do not develop hardcore auto racing simulations for over a decade using a purpose-built auto racing engine, only to suddenly abandon everything except your physical assets and run to an engine that quite frankly has no purpose creating racing games let alone simulators, while struggling to explain the benefits this new engine will provide your upcoming games. This is sketchy as fuck, and I’m disappointed I’m the only one pointing this all out.