So what's new Physics wise?

[CoolColJ]

Those cars are in motion before they donut.

I tried in the TT and I couldn't, you could in GT PSP but only with 4wd cars. If you can't in GT5 then the tire model is still flawed.

Maybe you weren't doing it right, but I had no problems doing it from a standing start in the Time Trial, once the car is declutched, rev'ed and then clutched. Though it starts to flatten out a bit after a while

 

[eight6er]

Not hard, but opposite lock, wider diameter donuts is a better test

I remember trying this and was pleasantly surprised the cars could do a donut, but (from memory) when you try proper opposite lock donuts, it didn't really work and felt unrealistic. I was dissapointed to discover cars would spin out, or understeer, causing the donut to end.

 

[JDMKING13]

Can you do donuts from a stationary start in the GT5 demo without the car edging down the track?

This is me having some fun



Yes Sir.

 

[Kryz11]

Wait, so all demos do not have Skid marks, do you think guys its even older build over the the demo form TGS that did have skid marks? Maybe physics change even more.

 

[Spagetti69]

Nice vid king, no skidmarks?

 

[Kryz11]

Also sorry for offtopic, but the videos with Skidmarks at TGS, cars sounded 10 times better over the BestBuy demo.

 

[JDMKING13]

Nice vid king, no skidmarks?

Yea this is the Demo without the skid marks, just wanted to show you can do donunts in GT5. I know it would be easier with a wheel 👍

 

[Kryz11]

Someone could show some burnout? We know the demo can do doughnuts


Also GT5 is maybe from half to a month away or more... so I think I might buy out one month of iRacing and play a bit.
Yeah there are no good racing games these days, I only see two choices to play with my wheel.

 

[cpp214]

I like how everything Mr Iracing knows about GT5's physics is either thirdhand or based on a vastly inferior GT5P...

And I like how everything Mr. GT5--YOU--knows about GT5's physics is either thirdhand or based on a vastly inferior GT5P

Putting it simply - GT5P is rubbish compared to GT5

How about not criticizing someone for doing the same thing you are doing? 💡

What exactly was so bad with the TT physics anyway? I don't get it, all you claim is the track wasn't bumpy enough?

Tell me what I said was bad about the TT demo's physics. Nothing.

I have no idea what cars are available in iracing, but even if they have a 370z on that same track, even the slightest change in setup will alter the way it behaves, so I'm curious to know what you found was so wrong with TT physics, even though gt5 will be an even further improvement on it?

Again, where did I say anything was wrong with the TT demo's physics?

For lfs, Ive tried out lots of cars setups and tracks I can conclusively say imo its too floaty and cloudy across the board to be realistic, there's no resolution to important part of the edginess and sudden spikes/ changes in physics behviour driving real cars have. TT captured this nicely imo.

The fact you keep saying lfs is just a notch below iracing and comparable to it, and then at the same time claim lfs has better physics that gt5 leads me to believe you might be mis-representing the realism of iracing as well.

That fact that you have that opinion about LFS' physics makes me question your taste in driving sims. I'm just telling you, and lots of people feel the same way, LFS has excellent physics. You are in the minority in your opinion of LFS. And where did I say it had BETTER physics than GT5? You keep putting words in my mouth with no evidence to back it up 👎 I DID say that GT5 now has physics COMPARABLE to LFS.

Btw laser scanned track is great to replicate track accuracy and autheticity, but speaks nothing for how realistic the physics engine is or in turn how realistically the game behaves

Are you kidding? Are you telling me that Gran Turismo, with it's non laser scanned tracks is more realistic than iRacing? Because by that type of reasoning, I think that's what you're trying to say.

I think you should separate the canned "micro-bump" wheel jiggle from the real FFB in iRacing. The two are separate things, because I an tell you right now a 60-150mb track in iRacing is not going to have that kind of polygon data to create those mm/cm sized resolution needed to create those bumps for real. If the track sizes were in the 1 and more GB range then you can start to think that

And to the people that are saying that the bumps in iRacing's tracks amount to a "canned" effect--that's really funny. Are you a game programmer? I think your grasping at straws with your statment that those bumps and nuances aren't real. And yes, I know the difference between a bump in the track and other FFB effects. iRacing's FFB is so nuanced and subtle, it makes other games FFB feel way too strong and clumsy. iRacing has the best FFB in any racing game I've ever played, and lots of people feel the same way. One of the main reason's for this is the nuances in steering feel provided by laser scanned tracks. I guess you guys are trying to say that GT5 and other games that have unrealistically smooth tracks are more accurate? That's one of the weakest and most ridiculous arguments I've seen.

 

[Kryz11]

I know laser scanned tracks add realism. But, we really don't know how the physics of iRacing stock up. You could Laser scan a track and put it in Need For Speed game, does that mean realistic game? Only for the track, not the physics. So far it seems iRacing is best sim out there.

We are not saying that unrealistic smooth tracks are more accurate, we are talking about car physics not track physics.

By the way anyone try to compare iracing to real life about Time wise? Like how much NASCAR in iRacing time around the track vs real car?

Also was it Dale Jr talking about NASCAR in game is bit harder then driving one in real life? I remeber watching some interview with him he mentioned its best sim, but on some turns and moves it was harder for him to do in game then in real life.

 

[JDMKING13]

I'm telling you the GT5 demo car physics are way better than GT5P and a little better than the TT demo. With N3 I felt good pushing the cars. Cant wait to feel the game with a wheel!

 

[Whistle Snap]

This is me having some fun



Yes Sir.

So the tire model is still flawed. 👎

 

[JDMKING13]

So the tire model is still flawed. 👎

Please explain.

 

[Kryz11]

So the tire model is still flawed. 👎

... there are no skid-marks like in GDC show.. could mean different physics also.

 

[JDMKING13]

... there are no skid-marks like in GDC show.. could mean different physics also.

Is that why he said the tire model are flawed? Thats pretty silly if thats the reason skid marks has been confirmed.

 

[Kryz11]

Yeah I Agree.

 

[amar212]

So the tire model is still flawed. 👎

Flawed? I see almost perfect low-speed physics there for a car with MR and 650HP on the rear axle.

But, you probably know better.

 

[CoolColJ]

And to the people that are saying that the bumps in iRacing's tracks amount to a "canned" effect--that's really funny. Are you a game programmer? I think your grasping at straws with your statment that those bumps and nuances aren't real. And yes, I know the difference between a bump in the track and other FFB effects. iRacing's FFB is so nuanced and subtle, it makes other games FFB feel way too strong and clumsy. iRacing has the best FFB in any racing game I've ever played, and lots of people feel the same way. One of the main reason's for this is the nuances in steering feel provided by laser scanned tracks. I guess you guys are trying to say that GT5 and other games that have unrealistically smooth tracks are more accurate? That's one of the weakest and most ridiculous arguments I've seen.

If these micro-bumps are real then what is providing the data for them? If it's actual track geometry in terms of polygon data, then the 50-150 meg sizes of iRacing tracks (which includes textures) is far too small to contain this data.

I dabble in 3D art/rendering, and laser scanned model of a small object with high detail level takes up a good 100+ megs as is, now multiply that over a whole track of KMs of polygon points, and it'll be way over 1GB of data!

The game runs on 2GB systems, so it certainly can't be that detailed in polygon data, and Pentium 4 systems would not be able to run it, which my last PC system was, and it could run it just fine.

One thing about iRacing, that has always made me question it's tyre model and physics, is how hard it is for most of the cars in the game to break traction from a dead stop, clutch dump....
Try doing the figure 8s as shown in the Time Trial video and stuff in iRacing and see where you go, not very well

So the tire model is still flawed. 👎

Skidmarks and smoke in the Gamescom build, very easy to brake traction.
Much easier to do this in FR cars off course. There is another video of a Zonda R doing donuts on the same build, but I can't find it right now.
Even in real life, you need to car to move a little before the rear end swings out.

 

[JosVerstappen]

nice physics at the end of the vid

http://www.youtube.com/watch?v=LOIUF6ykvTY&feature=player_embedded

 

[Nish]

That's called a glitch not physics.

 

[DeolTheBeast]

nice physics at the end of the vid

http://www.youtube.com/watch?v=LOIUF6ykvTY&feature=player_embedded

The reaction from the player sums it up .
On the other hand it's insanely funny 👍

 

[CoolColJ]

Ok I found the Zonda R, Gamescom build, video - the fun starts at 12:04

This is me having some fun



Yes Sir.

This is a real life Enzo FXX doing the same, you can see the real car doesn't spin on a spot as well.
If only GT5 had smoke that stayed a bit longer

 

[boxox]

If these micro-bumps are real then what is providing the data for them? If it's actual track geometry in terms of polygon data, then the 50-150 meg sizes of iRacing tracks (which includes textures) is far too small to contain this data.

I dabble in 3D art/rendering, and laser scanned model of a small object with high detail level takes up a good 100+ megs as is, now multiply that over a whole track of KMs of polygon points, and it'll be way over 1GB of data!

The game runs on 2GB systems, so it certainly can't be that detailed in polygon data, and Pentium 4 systems would not be able to run it, which my last PC system was, and it could run it just fine.

Erm, that's not typically how it works. Creating a surface model of a track that relies on individual polygon placements is about as efficient as creating a model of a brick wall where you model it one brick at a time. The laser scanning gets them a giant point cloud, which can then be assigned to a 3d model showing areas of the track which gave a uniform response, which can in turn then be given a surface type assignment.

The bumps happen in the surface model, which has nothing in particular to do with 'the polygons' (not that any of the physics primitives will be rendered in any traditional sense anyhow), other than acting as an index for location assignments for surface types. The surface type itself will contain a set of physics parameters for that part of the track surface, which influence how the tyres/car body/ffb interacts with it.

I don't think anyone outside of First has seen exactly how it works, but that's certainly how they did it in NR2003, which is the codebase they started from, and every racing game I'm aware of does the same thing. It's just obscenely wasteful not to do it like that.

This is a real life Enzo FXX doing the same, you can see the real car doesn't spin on a spot as well.

I don't think the inputs on this example are quite the same

The reason people throw games through "the donut test" is that it's very good at picking up on steering assists (though some games will defeat this by just disabling them at lower speeds). It's not particularly indicative of what the tyre model does under the surface, but it's VERY good at picking up on whether there is a limit on how much steering lock you can apply when the tyres are slipping out.

 

[cpp214]

If these micro-bumps are real then what is providing the data for them? If it's actual track geometry in terms of polygon data, then the 50-150 meg sizes of iRacing tracks (which includes textures) is far too small to contain this data.

I dabble in 3D art/rendering, and laser scanned model of a small object with high detail level takes up a good 100+ megs as is, now multiply that over a whole track of KMs of polygon points, and it'll be way over 1GB of data!

The game runs on 2GB systems, so it certainly can't be that detailed in polygon data, and Pentium 4 systems would not be able to run it, which my last PC system was, and it could run it just fine.

So, the bumps that everyone feels on tracks in iRacing--in spots where they're supposed to be in real life--are just made up? Riiiiigggghhhht

So you dabble in 3D stuff, but you're not actually a programmer? Do you know what kind of software iRacing is using to produce the tracks in the game? What you are doing is basically saying iRacing is lying aren't you?
Why would they lie? Lots of intelligent people play the game, don't you think there would be someone on their forums or somewhere on the internet that would have called them out on this if they were really lying about the accuracy of their tracks? If they are lying, you should be able to produce a link or something to someone who has already proven this, if not, you're on your own and the burden is on you to prove that what they are saying is not accurate, which you haven't done and can't do unless I'm badly mistaken. But, if you could, that would be pretty big news I think.

One thing about iRacing, that has always made me question it's tyre model and physics, is how hard it is for most of the cars in the game to break traction from a dead stop, clutch dump....
Try doing the figure 8s as shown in the Time Trial video and stuff in iRacing and see where you go, not very well

Have you played iRacing recently? I have no problem dumping the clutch and doing burnouts in just about any car, except maybe underpowered ones like the Skip Barber and Solstice.

Here's a burnout in iRacing with skid marks. He seems to do pretty good with figure 8's and skid marks. Must not be too hard.



EDIT: and I am officially done defending iRacing in this thread. I can't believe it needs defending in the first place, but I'm done with it. If people want to believe that GT5's physics will equal iRacing that's fine.
You're entitled to your opinion.

 

[xNEVER-ONEx]

Look man, once again, your missing the point. I don't think many people would say GT5 has better racing than Iracing, but GT5 is just so much more than Iracing thats its not even a fair comparison.

 

[cpp214]

Look man, once again, your missing the point. I don't think many people would say GT5 has better racing than Iracing, but GT5 is just so much more than Iracing thats its not even a fair comparison.

Missing the point? The thread topic is ostensibly about GT5 physics. Who said anything about racing? Although you are right, GT5 won't have better racing than iRacing. The reason I started posting in this thread is because people started stating a bunch of misinformation as fact, and they're still doing it, trying to make GT5 look as good or better physics-wise than games like iRacing. All I'm doing is trying to bring a little reality check to people who think GT5 will have physics that are going to rival the best PC sims. They are going to have 1,000 cars in the game. There is no way they can implement completely accurate physics that are individually detailed with tender love and care on each of those cars, it ain't gonna happen. The physics may be good, but they will not be anywhere near as accurate or detailed as a game like iRacing. People just can't accept that the GT series is not the best at everything it does.

 

[xNEVER-ONEx]

Missing the point? The thread topic is ostensibly about GT5 physics. The reason I started posting in this thread is because people started stating a bunch of misinformation as fact, and they're still doing it, trying to make GT5 look as good or better physics-wise than games like iRacing. All I'm doing is trying to bring a little reality check to people who think GT5 will have physics that are going to rival the best PC sims. They are going to have 1,000 cars in the game. There is no way they can implement completely accurate physics that are individually detailed with tender love and care on each of those cars, it ain't gonna happen. The physics may be good, but they will not be anywhere near as accurate or detailed as a game like iRacing. People just can't accept that GT5 is not the best at everything it does.

Well I'm ok with that, cause it's still got 1000 cars

 

[Kryz11]

Missing the point? The thread topic is ostensibly about GT5 physics. Who said anything about racing? Although you are right, GT5 won't have better racing than iRacing. The reason I started posting in this thread is because people started stating a bunch of misinformation as fact, and they're still doing it, trying to make GT5 look as good or better physics-wise than games like iRacing. All I'm doing is trying to bring a little reality check to people who think GT5 will have physics that are going to rival the best PC sims. They are going to have 1,000 cars in the game. There is no way they can implement completely accurate physics that are individually detailed with tender love and care on each of those cars, it ain't gonna happen. The physics may be good, but they will not be anywhere near as accurate or detailed as a game like iRacing. People just can't accept that the GT series is not the best at everything it does.

So where is your proof again that GT5 wont come close to iRacing. Sure iRacing is king in Physics so far, but who knows if GT5 will even be close to it?

You are stating something that no one play FULL GT5 game, and no one ever did a good non biased comparison between the two.

So when GT5 comes out and you and or someone else does in depth comparison, then you can state GT5 is not as good as iRacing or as good as iRacing in Physics.

 

[cpp214]

Well I'm ok with that, cause it's still got 1000 cars

And you know what, I am too actually 👍 I'm looking forward to GT5 as much as anyone. I think it's going to be great in so many areas.

But after playing iRacing pretty extensively this summer, I was blown away by the physics. It can be a bit stale and boring at times, and it occasionally feels more like a job than a game, but it has spoiled me in terms of the quality of how the cars feel in the game. It doesn't have many cars, but they are all so detailed and realistic, you can tell a lot of time and effort has gone into them. The track lineup is great in iRacing--especially on the oval side--and the road course lineup is outstanding as well and only getting better with Spa and Suzuka on the way.

GT5 will be great. iRacing is great. I'm going to quit comparing the two and trying to defend iRacing. It's a great game and definitely worth a look. And obviously everyone reading this thread is going to buy GT5. One game is designed to appeal to millions, while the other currently has a niche market of ~ 22,000 subscribers. One will have 1,000 cars, while the other only has 20 something. They are both quality games and can coexist nicely with each other, as they really don't compete against each other right now.

 

[Kryz11]

I was blown away by the physics. It can be a bit stale and boring at times, and it occasionally feels more like a job than a game

Haha, so its sounds like racing in real life!

By the way sorry for offtopic, whats your favorite car in the game?

I was getting use to that lil Pontiac, once I went to that Ford race car, man the feedback was a beast, I had to turn down the force feedback a bit.