Feature
July 4, 2010

Extended Play: GT5 Physics

Introduction

SimHQ's initial E3Expo 2010 report on Gran Turismo 5 for the Sony Playstation 3 was based on a short, two minute period behind the wheel in the opening hour of E3. That time increment was based on Sony's E3 demo having a timer attached to it — after two minutes of driving the game demo simply faded to black and reset to one of the game interface pages. With the two minute demo complete, each of us were politely shuffled off to the exit line and told to have a nice day.

Well, that two minute taste of GT5 simply wasn't good enough for SimHQ's team of die-hard sim racers, in large part because our brief exposure the game seemed pretty promising but didn't provide a good feel for the game's physics.

During the final hour of the show, with all of our other appointments and obligations met, the majority of the staff simultaneously migrated back to the Sony display and got back into the queue for some extended play on the GT5 demo. Although the line was long, Sony's GT5 handlers took a liking to the men in crimson and allowed SimHQ unprecedented access to three GT5 cockpits to do some extensive, subjective evaluating of the game demo's physics. And by extensive we mean that ConManly, Chipwich, Joe and Chunx each got nearly a continuous full hour of seat time with the new title — albeit in an endless string of 2 minute rides. So despite our extended session with GT5, these short rides, combined with the E3 demo's highly restricted access to settings and functions, made GT5 difficult to evaluate, other than the aspects we were shown. With that in mind, each of the car types were driven by the SimHQ team using various game settings, such as those shown at the bottom of the page. With that long list of cavaets, here are our subjective thoughts on Sony's GT5 E3 demo.

WRC

On gravel, the rally cars felt strangely disconnected from the road. Even before Chunx started the demo, one of the Sony handlers advised him to not use any more than "half-throttle, maybe less". Despite several years of playing Richard Burns Rally, driving rally cars in GT5 instantly stirred visions of trying to drive an air hockey puck for Chunx, who found the grip level on gravel less than RBR's wet stages and more like driving on damp grass than on dry gravel. Perhaps that's just due to our limited exposure to the demo, or maybe the WRC cars (or the gravel tire model) are still being tweaked before the game's release - but at E3 the game demo seemed to portray an almost frictionless surface on unpaved roads. We're hoping that's just a pre-production glitch, because the SimHQ staff really, really likes rally driving.

On tarmac, the rally car story is just the opposite. Using a Citroen C4 with tarmac tires (no, not R3 tires...) around the Top Gear Test Track, the WRC car stuck like glue, and offered almost no power-on oversteer or lift-throttle oversteer. That's very similar to how RBR feels on tarmac stages, with tons of grip until the tires reach their limit and break away violently — similar to real racing tires. Since Polyphony Digital has been planning on WRC cars for some time (and possibly paid the FIA some substantial coin for the license), we suspect and hope that there will be further development of the gravel grip modeling prior to the game's release this November.

NASCAR Cup

Chunx put a 2009 Impala Cup car to the test around the oval at Indianapolis during a 2 lap race with the AI. Although he didn't have to steer to the right in order to go straight (as in a real cup car), the cornering and AI performance was very convincing and made for a fun racing experience that also looked very good. The cars looked and sounded very good, and the grip level and handling on corner entry and exit seemed to be pretty accurate. Sony also reported that they've programmed a lot, if not all, of NASCAR's track rules into the game when using Cup cars, which is great news for oval racers. It's certainly understandable that the GT5 demo at E3 would offer no way to alter or tweak a car's chassis setup, and while we suspect that stagger or asymmetric camber or wedge can't be programmed into GT5's setup engine, it sure would be a nice addition if it did — that one addition might make GT5 one of the better NASCAR games out there.

Super GT 500

Chunx and Joe tried out a couple GT racing-class cars from Japan's wildly modified, 500-horsepower Super GT series around Le Mans and a fictional street circuit in Rome. The real world cars feature tons of downforce and a lot of grip, and that was well-replicated in the game, with predictable handling and performance that mirrored that found with the same cars in Forza Motorsport 3 and even in a freeware mod for GTR2. There's not much to say here - the game felt really good, the JSGT cars had gobs of power, close gearing ratios and a lot of grip from their R1 and R2 racing tires, until the grip limit of the tires was reached, at which point the tires gave up abruptly (just like real racing tires). One nice-to-see physics aspect was the movement of the visual scene in conjunction with undulations in the pavement on the Mulsanne straight, which also upset the taut suspension and caused some wandering of the front end as the downforce and load level on the front tires varied. Great stuff!

Performance / Exotic Street Cars

Joe tried out the Corvette ZR1 at Le Mans and Chunx the Ferrari 458 Italia on the Top Gear Test Track. Both cars had less grip than their racing counterparts, in part due to their street tires with lower operating temps and grip levels. Of course, you have about a half-dozen tire compound options in GT5, ranging from street tires all the way up to three types of racing tire compounds. Just to get a better flavor of the handling, different tires were chosen, as they'd be fitted to the real cars — Chunx chose high-performance street tires for the Ferrari . Joe noted that while the powerful ZR1 could approach 200 mph on the Mulsanne, the car caused the driver a great deal of trepidation as the nose started to get light and the car darted around on the imperfections in the pavement.

The Ferrari could easily be over-driven into a corner, with the result being massive understeer. Conversely, an abrupt right foot could snap the rear end loose and turn the car around in the corners. Pretty much what you'd expect from a car like the Ferrari. Best of all, Chunx had some real "Jeremy" moments as he performed a lurid power slide out of Hammerhead that were very difficult to maintain, easy to over-cook and felt very realistic (really, more realistic than FM3, where that edge of the physics model seems to be toned down a bit). We think that the street cars are where GT5 really shows its strong physics engine, because of the reduced grip level of the street tires and minimal downforce on these high-powered cars.

Closing Comments

After the team hogged three GT5 2D demo cockpits for over an hour (and drew a crowd that put more people into Sony's GT5 3D demo queue), the SimHQ racers came away with some very positive impressions of the demo's AI, physics and tire modeling, at least on paved surfaces. In addition, after longer exposure to the game, the force feedback and wheel resistance from the Logitech DFGT seemed to be quite good — certainly more robust than in GT5-Prologue. The Sony handlers did inform us that the physics engine of GT5 is a complete re-work over the Prologue edition, and it certainly seemed that way. We hope that the in-game experience with a G25/G27 or Fanatec wheel is even better — especially if in-game options for H-pattern, sequential or paddle shifting are offered, and those products' clutch pedals are still supported. It would be interesting to do a side-by-side comparison with Forza Motorsport 3. Or the just announced Forza Motorsports 4. Or both.

We're all looking forward to testing the finalized GT5, as the short taste we got at E3 seems to indicate that fans of PC racing simulations could soon have a compelling reason to put a PS3 on their Holiday wish list.

We want your Feedback. Please let us know what you thought of this article here.

Click here to go to top of this page.