So what's new Physics wise?

[JDMKING13]

Best Motoring vs GT5 👍

CoolColJ I love it man!

 

[objp]

GT5 better have a tire model 100 times better than any other GT game if they want to be mentioned in the same breath as iRacing and Netkar Pro. I don't see it happening, especially since there have already been other disappointments.

But seriously now, iRacing has what...21 cars? Some of which are similar. Netkar has less than that. GT5 has over 1000. There's simply no way that PD could model 1000 cars, and god knows how many different tires, to that level. It simply isn't possibly with their dev team.

 

[Bigbazz]

The BMW vid above certaintly shows a MASSIVE improvement in physics over the old games, just from what i can see and hear, it reminds me of how the Solstice acts in Iracing, though GT always did it good with the slower cars and the road cars, it was the faster cars and racing tyres where it was off.

Looking forward to next week to try this for myself.

 

[mykem]

GT5 better have a tire model 100 times better than any other GT game if they want to be mentioned in the same breath as iRacing and Netkar Pro. I don't see it happening, especially since there have already been other disappointments.

But seriously now, iRacing has what...21 cars? Some of which are similar. Netkar has less than that. GT5 has over 1000. There's simply no way that PD could model 1000 cars, and god knows how many different tires, to that level. It simply isn't possibly with their dev team.

As a matter of fact it was, Stefano Casillo (nKar Pro's developer) who recently mentioned GT Academy TT in the same breath as iRacing and nKar Pro in a recent interview. In his word "I think there is a fundamental difference in how FF is calculated in the group netKar PRO, iRacing, Live For Speed, VirtualGP3, even Gran Turismo Academy felt "netKar-ish" to me." And the word so far is that GT5's physics is even better than the one in GT Academy. Go figure.

You can read the rest of the interview here: http://www.radicalsonline.com/conte...peaks-about-his-life-simracing-and-netKar-PRO

 

[Mech0z]

Do you become a better driver from this game (IRL) just wonder if its worth for a student like me to buy a Fanatec GT3 RS v2 wheel and a PS3 just to game this single game I have had a GT25 before and raced grid and allthough its fun its so damm unrealistic, and I want to become a better driver

 

[bjgil2]

GT5 better have a tire model 100 times better than any other GT game if they want to be mentioned in the same breath as iRacing and Netkar Pro. I don't see it happening, especially since there have already been other disappointments.

But seriously now, iRacing has what...21 cars? Some of which are similar. Netkar has less than that. GT5 has over 1000. There's simply no way that PD could model 1000 cars, and god knows how many different tires, to that level. It simply isn't possibly with their dev team.

The physics model works with a set of parameters, they don't create a 4 new tyre models for each car x 1000 cars (or even 1 physics model per car), each car has parameters which are fed into the physics algorithm to produce feedback about the dynamics of the car.

Once the car's parameters are fed into the algorithm, along with current speed, steering angle, braking, etc. the algorithm produces a set of outputs which are fed back out to adjust the behaviour of the car, which then (along with the players new controller movements) become the new inputs for the next iteration of the algorithm

 

[boxox]

I really only have one request for advancement of the GT5 tyre model over previous iterations:



Don't do that.

 

[objp]

The physics model works with a set of parameters, they don't create a 4 new tyre models for each car x 1000 cars (or even 1 physics model per car), each car has parameters which are fed into the physics algorithm to produce feedback about the dynamics of the car.

Once the car's parameters are fed into the algorithm, along with current speed, steering angle, braking, etc. the algorithm produces a set of outputs which are fed back out to adjust the behaviour of the car, which then (along with the players new controller movements) become the new inputs for the next iteration of the algorithm

So wait...when all the cars are fitted with a...S3 tire for example, it acts the same on every car? That isn't right if thats the way they do it.

 

[mykem]

So wait...when all the cars are fitted with a...S3 tire for example, it acts the same on every car? That isn't right if thats the way they do it.

It should- it's similar to fitting, for example, a Bridgestone RE050, on a Ferrari 430 Scuderia or a Nissan R35 GTR. The tire modeling should be the same regardless of the cars. The difference rest in the modeling of the car's inherent physics (eg. mass + suspension) which would then impact the way the tires behave.

 

[objp]

It should- it's similar to fitting, for example, a Bridgestone RE050, on a Ferrari 430 Scuderia or a Nissan R35 GTR. The tire modeling should be the same regardless of the cars. The difference rest in the modeling of the car's inherent physics (eg. mass + suspension) which would then impact the way the tires behave.

Yes but an S3 can be put on a Civic, or on a Ferrari. Those tires likely are not going to be the same size, therefore they'll behave much differently.

GT5's physics just seem like a half ass way of doing things.

 

[mykem]

Yes but an S3 can be put on a Civic, or on a Ferrari. Those tires likely are not going to be the same size, therefore they'll behave much differently.

GT5's physics just seem like a half ass way of doing things.

I'm sure PD considers the sizing of the tires as part of the car's inherent physics. Being able to know the tire's specifications doesn't make it a sim. A good example is that games like NFS and Midnight Club allow you to choose not only the brands of tires but also the size/width of front and rear tires which amount to practically nothing. I believe Forza and SHIFT have these features which again makes little if any difference in how the driving actually feels.

While PD's approach may be half-assed but if you've read my earlier post, the result as Stefano Cassilo put it, can feel "netKar-ish". That is if you're willing to be more flexible.

re: the NSX tire lock steering. I don't normally drive the car in GT5P without ABS since I spend considerably more time in TT mode (which doesn't allow you to switch off ABS while keeping in at a low 1 (10 being the max)). Instead of tire modeling, I believe it's mainly the fault of the GT5P braking system. It would e interesting to see how non-ABS cars- for example, the Ferrari F40 would behave in the same situation. As far as I know the Honda NSX-R (in real life) comes with (non-switchable/defeatable) ABS system.

 

[CoolColJ]

The tyre lock thing shouldn't be an issue. Prologue and GT5 share nothing in common physics wise. Anyone who played the Time Trial would have realised that and GT5 is further refined and different from that tyre model wise. You can do drifts now with a 900 degree wheel, whereas before you had to be a freak to pull it off

All aids off, with a gamepad - first turn, tyre lock, into the sandpit. You can see the tyre meter on the bottom left go red on the front tyres

 

[alucard0712]

Just tell me how it feels compared to TT physics?To me TT physics felt best.

 

[objp]

Guess I just don't have high hopes for a tire model that doesn't allow pressure to be adjusted, and cannot give us temperatures. That's ridiculous. Its 2010.

 

[Vspectra]

Collision physics in the rally stages still look kind of weird.

 

[mykem]

Guess I just don't have high hopes for a tire model that doesn't allow pressure to be adjusted, and cannot give us temperatures. That's ridiculous. Its 2010.

At this point you're verging on that precipice that turns someone into... but I hate that word anyway. Simple logic: GT5 supports tire wear therefore temperature is intrinsically part of the tire physics. Yes, pressure and being able to change tire size would be nice but unless it can be properly implemented which at this point (2010) is entirely the domain PC Sims: nKar Pro + iRacing. If it's half-assed as in FM3, then I think it should be left out.

re: NSX tire lock steering. I was told the problem was fixed in GT5P's Spec II or III physcis update.

 

[objp]

Pressure is so basic though, and in fact, pressure increases and decreases along with the temperature of the tire. So if you say that temperature is part of the tire model (it is possible to model wear without temperature), then pressure has to be as well or its inaccurate.

 

[CoolColJ]

M3 messing around on the Top Gear track - Donuts are easy as pie now

 

[boxox]

Yes but an S3 can be put on a Civic, or on a Ferrari. Those tires likely are not going to be the same size, therefore they'll behave much differently.

GT5's physics just seem like a half ass way of doing things.

In terms of actually assessing a tyre as a physical object you will most often see it represented as a collection of dimensionless numbers that only really do anything once they are attached to a tyre radius and width and given a load. There's nothing stopping an almost identical formula being 'right' for two different tyres attached to different cars.

 

[toMsonsLLC]

X1 videos look stupid fake, no wonder it isn't a real car. Doesn't use real physics.

Sorry, but what a waste of time on PD's part..

 

[WolfRacer543]

X1 videos look stupid fake, no wonder it isn't a real car. Doesn't use real physics.

Sorry, but what a waste of time on PD's part..

It looks how a car with epic downforce, extremely light weight and extremely powerful should look.

 

[Perkel]

X1 videos look stupid fake, no wonder it isn't a real car. Doesn't use real physics.

Sorry, but what a waste of time on PD's part..

lol, no wonder why fan was banned from F1

 

[pervert]

X1 videos look stupid fake, no wonder it isn't a real car. Doesn't use real physics.

Sorry, but what a waste of time on PD's part..

Wow another constructive post to add to...

"GT5 is stupid. You have to buy a car BEFORE you take any tests. So yea, if you don't like one of the 3-ish cars, you have to sell it. Stupid move!"

"It's stupid to say no, cause in a racing game you can never have all the good car, nor tracks."

"Hope that is just that stupid channel."

 

[Dynamite White]

The turbos seem very very unrealistic in prologue and in the videos I've seen. There is no lag at all and no spool up or venting? This is'nt a fast and furious wannabes opinion because all of the features I've mentioned would add to realism and car choice.

Lag is a massive influence when driving because it doesnt give as linear a powerband as N/A cars. Also the spool up and venting "noises" would enable us to drive without having to keep a constant eye on the boost gauge. I think gt is an AWESOME! game however there still seems to be a few issues that seem "unfinished"?

 

[Boris Lozac]

As a matter of fact it was, Stefano Casillo (nKar Pro's developer) who recently mentioned GT Academy TT in the same breath as iRacing and nKar Pro in a recent interview. In his word "I think there is a fundamental difference in how FF is calculated in the group netKar PRO, iRacing, Live For Speed, VirtualGP3, even Gran Turismo Academy felt "netKar-ish" to me." And the word so far is that GT5's physics is even better than the one in GT Academy. Go figure.

You can read the rest of the interview here: http://www.radicalsonline.com/conte...peaks-about-his-life-simracing-and-netKar-PRO

Cheers for that.. 👍

 

[phil-t]

Just though I'd pop in this thread and see if we can have a sensible look at the Physics now some of us have been driving for 20 hours or so..

MY experiences, I'll start with my initial overall experience, as I think it sets the tone and also puts any minor observations into some perspective.

This doesn't take into account any 'possible' hidden changes to the Physics engine, all assists where 'off' and I use Arcade mode to instantly try out a range of cars, I've also raced online a bit to experience yet more cars/tyres etc.

+ Initial impression was very good indeed, taking the Cicic EK9 as a first car, the grip levels and overall demeanour of the car indicates a pretty decent handling model indeed. I would say from a user enjoyment/experience that was 9.5/10 for me..

+ Under braking you instantly notice that PD have sorted this since GT5p, tyres on the limit of locking or actually locked correctly allow little to no steering effect, much more realisitic
+ Braking also has that nice 'squirming' feel when threshold braking into a corner (ABS on '1'), it feels 'right'.
+ Weight transfer seems to readily 'felt' in a good way, Much better then GT5p
+ Overheating tyres through excessive sliding/wheelspin seems to behave correctly with the tyres quickly losing grip.



+/- The handling model is definitely forgiving (again, all assists off), I can control and largely get around 3 laps of suzuka in a high power RWD supercar with tyres that don't allow traction in the first 3 gears. I'm not saying it's easy or that it's a terrible thing, but it's not brutally unforgiving as I find some hardcore 'sims', and in a way, I don't want that, I prefer having just about most situations controllable with some very quick reactions and a bit of 'vim and verve' applied to throttle/steering.
- I'm not convinced the brakes aren't slightly wrong. The issue is, if you turn off ABS, things get quite unrealisitic for me, even half brake pressure culminates in almost locked wheels, and any slight steering input has the car going into a slide, in fact you seem to be very disadvantaged by turning ABS off, I can't stop as quickly or with a semblence of control. What really highlights something is not quite perfect is that just putting ABS to '1' has a massive effect, the car rarely slides when trail braking to an apex, and you can stamp on the pedal and you get reasonably controlled braking (still a 'nice' touch of squirming and weight transfer affecting the car's balance). The final indicator that the brakes need a little more tweaking is that it's the only assist that can't be forced off in MP. Saying all that, with ABS set to '1' I find braking quite good, so it's not like I can't get it how I want it!
- Drivetrain effects and FFB - When setting off under immense wheel spin from standing start there is zero effect on the car, and no FFB effects either. It'd be nice to have some slighty contrived tramping/torque steer and a little 'fight' in the steering, on in the case of RWD, the rear end just sliding out a little.
- Collision physics - although the cars tend to react better the GT5p, it still feels a bit numb, which on one hand shouldn't be an issue if you are a clean racer, but when PD have events where you inevitably have to get amongst other cars (TG Lotus race) it would be nice to feel (even if more contrived, I don't know) slightly better car interaction.
- Tyre grip levels - The TG lotus race, or take an McLaren MP24-12C around suzuka on the default Sports (Hard?), That's massively unrealisitically reduced grip levels, you couldn't buy tyres in the real world that behave like that.. It's entertaining on one level of course, but when 'sports' tyres on the McLaren result in almost zero grip in the first 3 gears, you realise what is going on, that to make people 'happy' with the physics, they have reduced grip levels quite far to give the impression in low performance cars of more realism, but being harsh, I'd say despite me liking it, it's not right.
Conversely, Race->Soft tyres on the McLaren resulted in not being able to break traction in any meaningful way for 3 laps of Suzuka, the grip level was insane (And again quite entertaining in a god like way tbh).

I don't know if people can discuss the physics by trying to quantify things (I've made an attempt to highlight simple conditions that show any observations that I noted). I'm not some idiot trying to discredit the game, I'm just a very keen racer who really tends to get quite into these games.

My summary would be
Intially the handling model/grip levels give the best experience I've had to date on a console, I've had a lot of fun!
I would say that my experiences so far indicate for me that it's a nice step up from GT5p, but the brakes/grip levels and some missing drive train/car interaction stuff do show where it's a little mixed up at times..

With all this talk of things improving at higher levels and no sign of tyre/fuel depletion yet in my event racing etc, I need to revisit the physics when I'm at a higher level 'if' things really do change etc.

I don't even want to mention PC sims, I'm sure with the careful choice of tyres on cars GT5 can approach the natural feel that PC sims tend to offer, but I don't actually think it's got as complex physics as the more established sims all round, maybe in a couple of areas, time may change my mind on that.

 

[boxox]

Great writeup, thanks.

 

[Anes]

Just tell me how it feels compared to TT physics?To me TT physics felt best.

For some reason I can't get the word 'arcadey' out of my head. There's just something not quite right about GT5 physics.

TT was the best imo.

 

[CAMAROBOY69]

The physics are excellent. This the one area that impresses me the most with GT5. Once you finally get inside a "premium" car you can really appreciate the physics. This is not talking at all about damage physics, because those are just terrible.

 

[Fusion]

Well I rolled my RX-7 onto its side (hit the opposite barrier before rolling onto roof) cutting the last corner in Trial Mountain way too much lol

I remember people who had the game early seemed to have problems getting the wheels to leave the ground even. I'm finding it damn easy and to a point somewhat realistic - a definite plus in that regard.