Gran Turismo – A comparison across the series (part 2 added)

[Scaff]

👍 Great work again Scaff 👍 Just a shame the wife doesn't appreciate it as much as we do Maybe if you got Jordan to pay you for your endeavours? Send him an invoice, see what happens... 💡 ("You just keep on thinking, Butch, that's what you're good at... ")

Driving in GT4 with a DS2 is a much more delicate affair than it was. Would you say that this has altered your driving style for the older games, or did you revert to old habits?

I always had a problem in the early games with MR and RR cars: some of them (Lotus Europa especially) would oversteer if you merely showed them a corner at speed, and were almost undriveable as a result (Lotus Esprit and Ruf Porsches also bad). NSXs were always one of the better ones but the race versions were very tricky to drive fast (I usually preferred a Supra). This used to p*ss me off, as, power on, most mid- and rear-engined cars will grip till kingdom come (Beetles on wooden tyres excepted, of course). How do you feel about this?

Thanks again,

Ed

Thanks

As far as 'reverting to form' for the older games, it was a case of having to with GT, and a little less so for GT2 and GT3.

The problem with the MR and RR cars in the first two games seems to be down to the low levels of understeer in the games, it gives you little or no warning of when you have reached the limit. So the car just steps out of line very quickly. It also happends to the shorter, powerful FR cars, like the Griffith I used in the first test.

Regards

Scaff

 

[Wolfe]

Great write-up, Scaff, and an interesting read.

Basically, what you've proven is that, from an acceleration and braking standpoint, there's nothing unrealistic about GT4. Whether that is because of the tires, or simulated gear-ratios/power/"braking" of the cars in question, is unknown, but the bottom line is that it's close to reality. 👍

As for handling/understeer/oversteer, perhaps you should do slalom tests and skidpad tests (not very useful for comparing GT vs GT, since GT4 is the only one to provide these tests if I recall correctly, but useful for determining realism nonetheless)...

In any case, I still stand by my opinion that GT4's understeer comes off as a bit strong; more when it comes to escaping understeer, than being afflicted with it.
GT4's oversteer is too difficult to initiate (no such thing as power-oversteer, mid-turn, from an application of full-throttle in a powerful FR?? Gimme a break.) and nearly impossible to hold (thanks to the strange phenomenon that occurs during sharp countersteer, where the front wheels pull you -- very strongly -- out of the drift, as if they were being given power (on a RWD)), and is not a result of improper tuning, or a lack thereof; anyone who is unaware that stock cars can drift is blind.
GT3's cars felt hyper-smooth, floaty, artificial, and computerized. GT4 fixed that by making them feel ultra-stiff, nervous, and touchy, while still feeling artificial and computerized. I'm still not sure which is better.

Just tonight I discovered a copy of "Gran Turismo - Driving the Game" at my local EB, and read through it for a bit. One quote struck me as possibly explaining the whole physics problem...I can't remember what it said exactly, but it explained how ever since GT1 the physics system has been "evolved, tweaked, refreshed, improved," etc.
I suppose it was supposed to make you go, like "yeah, that's awesome, they've put so much work into it it must be realistic," whereas I just thought to myself, "What the ****? They've been dragging along that archaic, PS1-era physics system the whole time? That explains a lot..."
...if it's true, that is, and not just romantic banter.

This is usually the part where I promote two other specific racing sims...but, as per Scaff's request, I'll keep my mouth shut. Instead, I'll just say that I expect PD would be able to match the realism of said-competitors if they would just start over fresh for once...!! 💡

 

[niky]

Great update, Scaff. I agree, you ought to get paid for this.

That's a strange thing to note... while a lot of people complain about the understeer, they fail to notice that the oversteer has increased, too. A lot of the physics for weight transfer has been upgraded, and for cars like the NSX, it is quite noticeable... for RR cars andshort-wheelbase MR cars (like the Elise and the Clio V6), it gets downright intimidating at times.

 

[RenesisEvo]

Great update, Scaff. I agree, you ought to get paid for this.

That's a strange thing to note... while a lot of people complain about the understeer, they fail to notice that the oversteer has increased, too. A lot of the physics for weight transfer has been upgraded, and for cars like the NSX, it is quite noticeable... for RR cars andshort-wheelbase MR cars (like the Elise and the Clio V6), it gets downright intimidating at times.

I found this very much the case with my RX-7 last night... completely stock, a quick blast round the 'Ring on S2 tyres and it was oversteering far more than I have ever known an RX-7 do to in the GT series. It was even losing control at times - that would not have happened in GT3.

 

[Arwin]

I found this very much the case with my RX-7 last night... completely stock, a quick blast round the 'Ring on S2 tyres and it was oversteering far more than I have ever known an RX-7 do to in the GT series. It was even losing control at times - that would not have happened in GT3.

I agree, GT4 is just wonderful. I think there are two things holding it down: the routines that prevent the car from being able to flip over, and the simulation of tire pressure. But GT3 and GT4 made some serious improvements over the previous version each.

Now something more important. I have noticed, and I think you can find an old thread on this back even before GT4 was released, that there seems to be a quite elaborate form of Clutch emulation in GT4 for something that is not discussed in the manual or anything else. It has partly to do with the weird reverse system. More relevant to the discussion at hand, you can use it when making a standing start, which is important for donuts. You still can't do donuts like you expect to in reality (though I think there are a various number of things you have to take into consideration, including 1st gear ratio settings), but you do get closer.

Two important things about clutch emulation in GT4:

1. When you tap the reverse button, the car is not put back into gear before you press accellerate again. This is, contrary to what some claim, not just a side effect of the reverse button thing, as is easily proven if you put your car somewhere on the slope of a hill or if you drive at speed and disengage the clutch: when the clutch is disengaged, the car will no longer brake on its engine so it will slow down much less.

Now unfortunately, you cannot use this all that much while driving a lap, because whenever the gas pedal is pressed, the clutch automatically re-engages. But when making a standing start, you can do something to prevent that from happening.

2. When you disengage the clutch, and then put the brake down, the clutch will not engage until you release the brake! This means that you can freely rev the engine, which in itself is already very cool. But it also means that you can rev up the engine before engaging the clutch, which gives your start a lot more bite. This basically happens automatically already before a test/race because you have the countdown and during the countdown you can rev up the engine freely. But using the method I described, you can also do this when just fooling around on the parking-lot, and trying for donuts.

I think the clutch simulation already there is quite interesting. Experiment with it a little, you'll be amazed.

By the way, in GT4, worn and cold tires can also be allowed to be a factor (and ditto in GT3 incidentally).

 

[Greycap]

Arwin, isn't the same effect achieved with the handbrake?

 

[Arwin]

Arwin, isn't the same effect achieved with the handbrake?

Actually, if I remember correctly then no, it isn't ... the gear still engages when you use the handbrake.

 

[Greycap]

Strange. I've used handbrake starts and burnouts since GT3. But it doesn't work on FF cars, though.

 

[Arwin]

Strange. I've used handbrake starts and burnouts since GT3. But it doesn't work on FF cars, though.

Yes, but its a different thing, don't you see? You can do a burnout with a handbrake, because the gear engages and your wheel start spinning while you lock the car in place using the handbrake. But if you use the normal brake to prevent the gear from engaging, your wheels won't spin at all, but your engine will rev.

 

[Greycap]

I'm gonna make this simple, I hope this isn't insulting.

In FR, MR, RR and 4WD cars it works like this: 1. engage the handbrake. 2. rev the engine as much as you want. 3. release the handbrake and have a nice burnout.

I admit that it doesn't work in FF cars.

 

[Arwin]

I'm gonna make this simple, I hope this isn't insulting.

In FR, MR, RR and 4WD cars it works like this: 1. engage the handbrake. 2. rev the engine as much as you want. 3. release the handbrake and have a nice burnout.

I admit that it doesn't work in FF cars.

I'll make this even simpler. Why don't you try what I wrote above, and come back later to tell me you now understand what I meant .

Hint: look at the difference in how fast the rev counter comes up.

Also, do you know what might be the reason that it doesn't work in an FF car?

EDIT: have you actually ever driven a car with a clutch? I keep forgetting that's rare in the U.S.

 

[Greycap]

The reason that it doesn't work in FF cars is that the front wheels are the driving wheels but the handbrake locks the rear wheels.

Yes, your way works too but on my controller configuration it's quite difficult to use so I'll stick to my own system.

And another yes, I don't even have a driving license but have driven my dad's car on several occasions. Here in Finland almost all cars have ordinary manual gearboxes.

Whatever you use for it, keep on revving!

 

[Arwin]

The reason that it doesn't work in FF cars is that the front wheels are the driving wheels but the handbrake locks the rear wheels.

Yes, your way works too but on my controller configuration it's quite difficult to use so I'll stick to my own system.

And another yes, I don't even have a driving license but have driven my dad's car on several occasions. Here in Finland almost all cars have ordinary manual gearboxes.

Whatever you use for it, keep on revving!

Of course. But I still have a feeling you don't understand that using the handbrake WITH THE CLUTCH ON to rev the engine while standing still isn't at all the same thing as my trick that allows you to rev the engine WITH THE CLUTCH OFF .

What you do is make the engine turn the wheels, and then use the brake to hold the car from moving.

What I do is use the brake from preventing GT4 to engage the clutch, to prevent the engine from turning the wheels! This after I disengaged the clutch by tapping the reverse button.

You can even see in GT4 when the clutch is engaged and when it isn't - the gear indicator becomes dim when the clutch is disengaged. Depending on the car, you will also see this happen for a short while when up or downshifting while driving.

Anyway, Kitos and Hyväa Yvätäa (hadn't seen you lived in Finland, sorry)

Arwin

 

[Greycap]

No no no, the wheels don't turn in other than FF cars with the handbrake trick. The gear is engaged, but so is the clutch.

Kiitos ja hyvää yötä for you too (what the hell, it's half past 4 PM) I should say hyvää päivänjatkoa

Rene

 

[Wolfe]

In GT4, applying the handbrake and then applying throttle disengages the clutch on any car with power going to the rear wheels (FR, MR, RR, 4WD/AWD); thus allowing your engine to rev freely. I'm not sure if the same rule applies to FF cars; I'm at school and cannot currently confirm this.

Creative use of the handbrake and/or reverse button can also disengage the clutch while you're moving, as Arwin already noted. This can be particularily helpful in special situations, like coasting down hills while trying to get the Benz Patent Motor Wagen around the 'Ring.

 

[Alfaholic]

What you do is make the engine turn the wheels, and then use the brake to hold the car from moving.

Since Greycap's method works in FRs and MRs, I doubt that the engine would be turning the driven (rear) wheels while the handbrake (which in nearly all cars acts on the rear wheels) is stopping the car from moving. If the clutch was engaged, then either the engine would stall or the car would pull away relatively slowly. For GT4 to allow the car to stand still and rev the engine while the handbrake is engaged on the driven wheels, it would have to disengage the clutch. Ergo, it would be exactly the same as keeping your foot on the brake (the gear indicator should be grey). In FF cars there is no conflict of interest between a handbrake operating on the rear wheels and an engine feeding power to the front wheels, so GT4 does not bother to disengage the clutch. FF cars therefore can be expected to sit still while creating little wafts of tyre smoke.

 

[Scaff]

In any case, I still stand by my opinion that GT4's understeer comes off as a bit strong; more when it comes to escaping understeer, than being afflicted with it.
GT4's oversteer is too difficult to initiate (no such thing as power-oversteer, mid-turn, from an application of full-throttle in a powerful FR?? Gimme a break.) and nearly impossible to hold (thanks to the strange phenomenon that occurs during sharp countersteer, where the front wheels pull you -- very strongly -- out of the drift, as if they were being given power (on a RWD)), and is not a result of improper tuning, or a lack thereof; anyone who is unaware that stock cars can drift is blind.

Just tonight I discovered a copy of "Gran Turismo - Driving the Game" at my local EB, and read through it for a bit. One quote struck me as possibly explaining the whole physics problem...I can't remember what it said exactly, but it explained how ever since GT1 the physics system has been "evolved, tweaked, refreshed, improved," etc.
I suppose it was supposed to make you go, like "yeah, that's awesome, they've put so much work into it it must be realistic," whereas I just thought to myself, "What the ****? They've been dragging along that archaic, PS1-era physics system the whole time? That explains a lot..."
...if it's true, that is, and not just romantic banter.

With regard to the understeer side of things I have to say that I think that GT4 has got he balance right, particularly for the road cars, one thing that I have found with GT4 is that controlling/exiting understeer requires both the reduction of throttle and a decrease in steering angle. Now this is right for real-world applications, but not somthing found in previous versions of the game. Also if you are using a wheel/pedal set-up (unlike in my tests) you get the advantage of nice controllable left-foot braking.

In regard to oversteer in general, but also power oversteer. I have just been playing around with a stock M3 running on N2 tyres, I've had no problem at all getting it to power oversteer out of the hairpins 2nd gear corners at Grand Valley. You do however need to ensure the car is set right, enter the corner too fast and understeer will screw with the cars balance and you are never going to get it. The 2nd hairpin at GV is perfect for this. Nice slow entry and then as you clip the corner apex get on the power, an M3 will understeer slightly at first, but this will transfer to nice controlled oversteer.

With the devlopment of the Physics engine from game to game, as far as I am aware GT, GT2 and GT3 (and possiably GT:C) are all developmentsof the same basic engine. As you say tweaked and evolved with each new game. I am, howver, sure I can remember reading that the engine for GT4 Prologue and GT4 was built from the ground up. I can recall the piece making a major point of the fact that it was the first time in the series that a new engine had been created. The difference was clearly visable when GT4 Prologue first came out and so of the issue discused today about GT4 have roots in threads in the GT4 Prolouge area.

BTW, thanks for you're restraint regarding 'the other two games'; I know from other topics and threads on GTP that this is a subject that you are very passionate about. To hold back and keep it on topic means a lot to me. 👍

important things about clutch emulation in GT4:

Read about this and found it of great interest, and as I had GT4 set up with an M3 ready and waiting I gave it a quick test and found the following:

First off for anyoine who does not know, if the gear indicator is grey then the clutch is engaged, if its black then the clutch is release and the engine is under load.


Handbrake - its use always engages the clutch

Brakes - use of the brakes below approx 6mph will also engage the clutch, explains why you can't stall the car in GT.

Release the throttle - release the throttle while doing less that 4mph and the clutch will engage

Reverse - not permitted until your speed is below 12mph, use of reverse will engage the clutch when released.

With all of the above if you release the button and get back on the throttle the clutch will release (but not straightway). However, as Arwin says if you keep the brake button pressed the clutch stays engaged even when you use the throttle. It will stay engaged until you release the brake button.

So does this help with do-nuts, well in the case of the M3 the answer is no. Uven using the method above, when the clutch is released is done gradually, it doesn't 'dump' the clutch, which can be of great use when getting a do-nut going.

Regards

Scaff

 

[niky]

This is usually the part where I promote two other specific racing sims...but, as per Scaff's request, I'll keep my mouth shut. Instead, I'll just say that I expect PD would be able to match the realism of said-competitors if they would just start over fresh for once...!! 💡

...I've often ribbed you over that myself...

But IMHO, a lot of what you have to offer besides the rants against GT is worth posting... ...didn't notice your post the other day... seems we posted at the same time.

Like Scaff said, this one feels a lot different from previous GTs, and is a remarkable improvement, but as you say, given the competition now, it could be better. I think one of the major problems GT had in regard to this was the prolonged development time. Any video game nowadays that spends more than 2 years in development (and GT4 took over twice that) will be dated when it comes out, or it is often overshadowed by games which were developed more recently but which came out first. I think the other games' advantage lies in the lower amount of rote-work (encoding tons of cars and tracks), which let them focus more on engine development... but we'll see about GT5, since they probably have to write everything from the ground-up AGAIN, they MIGHT just change things.

I've done Skidpads myself, but having no steering wheel as of yet (last wheel I had was for the PS1... and my experiences with that ensured I wouldn't buy another for a looooong time), I can't vouch for the accuracy of 'feel' from the dinky double-sticks, so I didn't particularly take note of 'understeer or oversteer' characteristics. But the numbers were (to a point) realistic.

@Scaff: Alfaholic has it right about FF's. They don't automatically get "clutched" on handbrake, and you can drag the rear end for a pretty long time with the throttle on full and the handbrake on.

Also, it would be very nice if you could "dump" the damn clutch from anything but a standing start... even from a standing start, a lot of times, it feels like a slipping clutch, and it's hard to juggle. I'm hoping in the next GT, we get a "clutch" button where the nitrous is... or even a "hold both gear-shift buttons to clutch" option.

 

[Wolfe]

With regard to the understeer side of things I have to say that I think that GT4 has got he balance right, particularly for the road cars, one thing that I have found with GT4 is that controlling/exiting understeer requires both the reduction of throttle and a decrease in steering angle. Now this is right for real-world applications, but not somthing found in previous versions of the game. Also if you are using a wheel/pedal set-up (unlike in my tests) you get the advantage of nice controllable left-foot braking.

I'm always on the DS2, because I don't own a DFP; a friend of mine had it, but he moved away to college...not that we used the DFP for any PS2 games anyway -- too difficult to set up correctly in front of a TV.

Anyway, you're completely right that reducing the throttle and(/or) decreasing the steering angle is how you eliminate understeer in real life. However, since GT4 second-guesses your steering input to make the DS2 usable (direct analog-stick-to-steering-wheel control would be quite difficult), reducing the throttle, especially combined with a (light) application of the brakes, should eliminate understeer relatively quickly. For whatever reason, it doesn't. Tapping the brakes to slow your speed and shift weight forward, then slamming on the gas in an FR should kill the understeer, kick out the tail, and let you drift your way out of the corner. For whatever reason, it doesn't.

I'm not even talking about over-cooking a corner, which is pretty hard to fix if you're especially careless; I'm talking about entering the understeer mid-corner, such as the final corner of Tsukuba. As I said in my last post, I've pretty much retracted my argument that understeer is too easy to encounter. However, what I am continuing to argue is that understeer's behavior is wrong, and that it is too difficult to fix; albeit, this difficulty in correcting probably has more to do with the lack of proper oversteer...

In regard to oversteer in general, but also power oversteer. I have just been playing around with a stock M3 running on N2 tyres, I've had no problem at all getting it to power oversteer out of the hairpins 2nd gear corners at Grand Valley. You do however need to ensure the car is set right, enter the corner too fast and understeer will screw with the cars balance and you are never going to get it. The 2nd hairpin at GV is perfect for this. Nice slow entry and then as you clip the corner apex get on the power, an M3 will understeer slightly at first, but this will transfer to nice controlled oversteer.

Ok, after reading this I took an M3 with N2's to Grand Valley myself. Other than mind-boggling understeer at any speed above approximately 60km/h (I'm not talking about the hairpins or sharp corners, either, even though my 20-year-old 318i can take a sharp (25mph recommended), un-banked onramp near my house at 40mph, the equivalent of 65km/h), and a complete lack of ability for the car to kick its tail out in larger corners, even with fierce attempts at using weight-transfer to initiate a drift (instead bogging down the engine, if not just pushing towards the outside wall upon throttle application)...yes, you can briefly power-over on the exits of the 2nd-gear hairpins.

However, the reason why these power-overs ended up being so short is something I'd like to hear your opinion on, since you haven't talked about it yet -- applying countersteer imediately pulled the M3 out of the drift, pulling me too far and causing fishtailing if I wasn't careful. Now, sloppy countersteer can cause you to pull too far out of a drift and fishtail in real life. What GT4 does wrong is that it seems to "assume" that the front wheels have the power to "pull" a car out of a drift, whereas in real life, the front wheels of a drifting car are very much playing a simple "following" role, much like the rear wheels of a FWD car. Have you ever noticed this countersteer phenomenon?

With the devlopment of the Physics engine from game to game, as far as I am aware GT, GT2 and GT3 (and possiably GT:C) are all developmentsof the same basic engine. As you say tweaked and evolved with each new game. I am, howver, sure I can remember reading that the engine for GT4 Prologue and GT4 was built from the ground up. I can recall the piece making a major point of the fact that it was the first time in the series that a new engine had been created. The difference was clearly visable when GT4 Prologue first came out and so of the issue discused today about GT4 have roots in threads in the GT4 Prolouge area.

That's certainly plausible, but whichever case is true, I personally believe GT4 isn't really any more realistic than GT3.

BTW, thanks for you're restraint regarding 'the other two games'; I know from other topics and threads on GTP that this is a subject that you are very passionate about. To hold back and keep it on topic means a lot to me. 👍

Well, I did "restrain myself" in a more suggestive, humorous way than I could have, but I certainly wanted to respect your wishes. You're one of the most intelligent, literate, rational-thinking people I've ever argued about GT4 with. 👍

 

[FastEddie12]

I've found that there's plenty of off-throttle oversteer to be had, usually, of course, on the way into the corners. For some reason, '80s mid engined cars such as the original Lotus Esprit and the Renault 5 Turbo are especially good at this. But I've rarely been able to transfer that into power oversteer.

Except on the loose. It seems pretty easy to get classic power oversteer on the loose with any RWD configuration, and it's remarkably easy to control. So there must be something right with the physics.

In fact my overall impression is that the less grip you have, the more accurate the physics. Anyone agree?

Also I think it's worth pointing out that PD have made GT4 very much with steering wheel controllers in mind. I often get the impression that my level of control is very ham-fisted with a DS2, and I would get more from the game with a DFP, especially when golding licences.

 

[Scaff]

Ok, after reading this I took an M3 with N2's to Grand Valley myself. Other than mind-boggling understeer at any speed above approximately 60km/h (I'm not talking about the hairpins or sharp corners, either, even though my 20-year-old 318i can take a sharp (25mph recommended), un-banked onramp near my house at 40mph, the equivalent of 65km/h), and a complete lack of ability for the car to kick its tail out in larger corners, even with fierce attempts at using weight-transfer to initiate a drift (instead bogging down the engine, if not just pushing towards the outside wall upon throttle application)...yes, you can briefly power-over on the exits of the 2nd-gear hairpins.

All I can say is that we must have very different driving styles, because I have had none of the problems you have just described with the M3 on Grand Valley, so I put on Gt4 and ran a quick series of laps and the following is a description of the quickest lap. All the speed information is taken from the Data Logger and in the interest of good comparison I used a DS2.

BMW M3 (N2 Tyres) @ Grand Valley

Cross the start/finish line at 130 mph

The first corner is always tricky as you need to start braking while still in the slight left hand 'kink', brake down from 139mph to 50mph and turn into the corner, keep a contranst speed of approx 50mph. The rear of the car will creep out as you accelerate out of the corner.

Flat out towards the chicane section before the next hairpin, this can be tricky to take, I know that my line is critical here, get it wrong and its big corrections with lots of understeer and loss of speed, or worse off the track. The first left hand part is taken with a slight lift of the throttle to check the speed down to 100mph. Keeping the car stable as I brake down to approx 80 mph for the next right hand kink. Then accelerate through the last left hand kink at 97mph and on to the second hairpin.

Braking hard and straight at the second hairpin, its a very tight corner and the speed loss in huge, down from 110mph to 35mph for the turn. Its a common mistake to take this one too fast, as its tighter than it looks (as are a lot of corners at Grand Valley). Showing off a little for the cameras I get the back out of line a little under power as I accelerate out of the corner

Hard on the throttle and into the complex, like the chicane line is critical here, the corners can be taken at a good speed, despite the way they try and unsettle the car. It starts with a deceving left hard sweeper, which I brake down from 86mph to 60mph to take.

The second part of the complex is a sharp right hand turn than needs me to drop my speed from 73mph to 40mph, accelerating away from here upto the final part of the complex.

The key to the complex for me is alway thinking about it as a series of corners and that the optimum line through it is not made up of the standard line for each corner, but a line for the whole complex. I have always found if I try and treat each corner seperatly then average speed drops and understeer kills the car.

One of the corners that every GT player knows, the sharp left hander into the first tunnel, run wide here and its into the wall. So I come down from 62mph to 50mph and take the apex at a constant 50mph to take the corner.

Accelerate towards the next tunnel corner, which is again a tricky corner, it tightens just past the apex and is quickly followed by the next right hander. I have to drop my speed from 93mph down to 50mph to take this one.

The real trick part starts with the next corner, it starts in the tunnel straight after you clear the last corner. I'm doing 69mph here and need to brake down to 55mph, its a minor drop in speed but the car is still unsettled from the last corner. I keep the car stable and drop the speed, keeping to 55mph until I can accelerate out of the tunnel and onto the bridge. These two corners are a real test of the car and its ability to remain stable while braking and changing direction rapidly, get it right and you can keep the car in check. Get it wrong and you will either take too much speed into the corner and understeer into the wall, or unsettle the car and end up facing the wrong way.

Hard onthe throttle over the bridge and onto another of Grand Valley's tradmark corners, the long right hand sweeper through the final tunnel. Approaching speed here is 108mph and its on the brakes and down to 70mph, I start the car well to the left and ride the corner at a steady 70mph. The back of the car starts to step out as I apex the corner and start to accelerate, I keep it in check to ensure I'm set up for the next two corners.

Tricky pair these, the first left is very tight requiring a drop from 84mph to 45mph and I keep the line controlled and the throttle steady at 45mph ready for...

the even tighter right hand corner. Its common to take to much speed into this one and suffer understeer as a result. So I come down from 47mph to 38mph, its not a big drop in speed, but it is needed to make the corner without running wide.

These two out of the way and its onto the most important corner of the circuit, the right hand open sweeper onto the long start/finish straight. Coming out of the last corner my speed is 100mph, I check my speed down to 95mph, with a dab of the brakes. This also helps settle the car and at a constant speed I turn in early and clip the right hand apex at the curb. I resist the urge to get back on the throttle too early and only accelerate when I know the car will run no wider than the left hand curb on the straight.

I have attached the data logger screen from the right hand sweeper in the final tunnel section here, as its a classic location for understeer at Grand Valley.



As you can see from the data logger, I tend to use a very rapid blipping motion on the DS2 to keep the speed constant, I find it impossiable myself to use the buttons just by pressure and can't get on with using the right hand stick.

Total lap time of 2.22:891, in a BMW M3 on N2 tyres, I'm sure people on here are quicker (in fact I have no doubt). However the only corners taken at less that 40mph were the second hairpin and the second corner of the last chicane, both of these corners are very tight and slow.

Well thats my lap of Grand Valley, please don't take the above as a flame on your comments, its not. I just find it hard to believe that we are talking about the same car, given my experiences with it.

However, the reason why these power-overs ended up being so short is something I'd like to hear your opinion on, since you haven't talked about it yet -- applying countersteer imediately pulled the M3 out of the drift, pulling me too far and causing fishtailing if I wasn't careful. Now, sloppy countersteer can cause you to pull too far out of a drift and fishtail in real life. What GT4 does wrong is that it seems to "assume" that the front wheels have the power to "pull" a car out of a drift, whereas in real life, the front wheels of a drifting car are very much playing a simple "following" role, much like the rear wheels of a FWD car. Have you ever noticed this countersteer phenomenon?

This one I need some time to play around with and have a damn good think about whats going on.

Give me some time with GT4 and a few good books and I will get back.

Well, I did "restrain myself" in a more suggestive, humorous way than I could have, but I certainly wanted to respect your wishes. You're one of the most intelligent, literate, rational-thinking people I've ever argued about GT4 with. 👍

Once again thanks, It's one of the things I love about GTP, is the fact that we can have these threads running and no one gets stupid about it. GTP is the most intelligent place to discuss this kind of thing I have found. Its members (and you are a wonderful example of this) are pasionate about what they discuss, but open and intelligent enough to do it in a constructive manner. I can't think of another site in which this topic could have been discussed in this way.

👍 GT Planet 👍

Regards

Scaff

 

[Wolfe]

All I can say is that we must have very different driving styles, because I have had none of the problems you have just described with the M3 on Grand Valley, so I put on Gt4 and ran a quick series of laps and the following is a description of the quickest lap. All the speed information is taken from the Data Logger and in the interest of good comparison I used a DS2.

BMW M3 (N2 Tyres) @ Grand Valley

*Long, impressively detailed description of the lap*

Total lap time of 2.22:891, in a BMW M3 on N2 tyres, I'm sure people on here are quicker (in fact I have no doubt). However the only corners taken at less that 40mph were the second hairpin and the second corner of the last chicane, both of these corners are very tight and slow.

Well thats my lap of Grand Valley, please don't take the above as a flame on your comments, its not. I just find it hard to believe that we are talking about the same car, given my experiences with it.

Ok, maybe I should have been more descriptive, and more literal, in explaining how the M3 felt to me; when I said "mind-boggling understeer," I didn't mean that the car was incapable of taking the corners. I meant that the tires were squealing and beginning to lose grip much earlier than they should have been, that the car wasn't nearly as balanced as it should have been, and that this frustrated and confused me; thus, "mind-boggling." In making the comparison to my car in real life, my 1985 BMW 318i (with its 195/60R14 Pirelli's and 20-year-old suspension), I was noting that there is an onramp near my house that is quite similar in sharpness and banked-ness to the second-gear hairpin at the southern-most point of the course, and that my car is capable of carving up that particular onramp at about 65km/h before the tires even start complaining. My point is, the much younger, M-tuned E46 should be able to do significantly better than my car...

After reading your latest post here, I went and attempted another lap, this time aiming for a low time, rather than experimentation with the car and tires, as I did last time. I apologize if you are offended by my pointing this out, but I managed a 2'20"509 on the first lap. I made a couple of mistakes, too, including nearly understeering into a wall at one of the corners (I completely let off of the throttle as soon as the understeer began, but the car kept going, and going, and going...<-- the main reason why I find GT4's understeer to be excessive). This picture here is proof of this lap, and this proof is not really intended for you -- it is simply to protect myself from potential nay-sayers who would be more than glad to yell at a GT4-basher like myself, "you're lying!! You're just slower than him, and aren't willing to admit it, so you blame the game!!"

This one I need some time to play around with and have a damn good think about whats going on.

Give me some time with GT4 and a few good books and I will get back.

While you're pondering that one, there's another strange GT4 phenomenon my friends and I have noticed that I forgot to mention -- it seems as though GT4 completely ignores the cornering benefits of a banked turn. I haven't done research comparing cornering speeds between GT4 and real life yet, but the Karussell at the Nürburgring seems to be a glaring example of this.

Once again thanks, It's one of the things I love about GTP, is the fact that we can have these threads running and no one gets stupid about it. GTP is the most intelligent place to discuss this kind of thing I have found. Its members (and you are a wonderful example of this) are pasionate about what they discuss, but open and intelligent enough to do it in a constructive manner. I can't think of another site in which this topic could have been discussed in this way.

👍 GT Planet 👍

Regards

Scaff

Thank you for the compliment! I disagree with niky that promoting another game while discussing GT4's faults is, by default, a bad thing to do -- and if I'm reading your final statement here correctly, I believe you would agree with me on this -- but when a member such as yourself kindly requests that these outside comparisons be put aside, I'm more than willing to comply. I know I can get heated up sometimes, and express things in a more blunt, slightly exaggerated manner, but I typically post in such a way in topics that are more subjective than objective, and you've certainly set up an objective discussion environment here.

Cheers to the rational thinkers of the GTPlanet community! 👍

 

[FastEddie12]

While you're pondering that one, there's another strange GT4 phenomenon my friends and I have noticed that I forgot to mention -- it seems as though GT4 completely ignores the cornering benefits of a banked turn. I haven't done research comparing cornering speeds between GT4 and real life yet, but the Karussell at the Nürburgring seems to be a glaring example of this.

Chatting to people at the 'Ring, general opinion is that you can't get more than about 45mph round the Karussell in a road car. Same for bikes. And once you're in it, that's it, you stay in it, and if you're on 2 wheels for God's sake don't try to come out early, it'll pitch you over the barrier!! It's the maddest place: it looks as though you could stand at the bottom of the banking and the top of your head would be level with the top of the banking. You kind of drop into it with a thunk, and of course it's concrete so, even on a bike, you get that Hitler Autobahn road noise... Petrolhead's paradise. You want 75% of the cars there (Ed rambles on ad infinitum about his trip...)

 

[Wolfe]

Chatting to people at the 'Ring, general opinion is that you can't get more than about 45mph round the Karussell in a road car. Same for bikes. And once you're in it, that's it, you stay in it, and if you're on 2 wheels for God's sake don't try to come out early, it'll pitch you over the barrier!! It's the maddest place: it looks as though you could stand at the bottom of the banking and the top of your head would be level with the top of the banking. You kind of drop into it with a thunk, and of course it's concrete so, even on a bike, you get that Hitler Autobahn road noise... Petrolhead's paradise. You want 75% of the cars there (Ed rambles on ad infinitum about his trip...)

I rarely look at my speedometer when racing, so as I said, I haven't really looked into comparing GT4's speed with real life. As for 45mph, I can believe it. I drove a couple of laps around the 'ring on my trip to Germany a couple of months ago!

"OH MY GOD THIS IS STEEP!!" - Me, upon entering the Karussell on my first lap

 

[Scaff]

Ok, maybe I should have been more descriptive, and more literal, in explaining how the M3 felt to me; when I said "mind-boggling understeer," I didn't mean that the car was incapable of taking the corners. I meant that the tires were squealing and beginning to lose grip much earlier than they should have been, that the car wasn't nearly as balanced as it should have been, and that this frustrated and confused me; thus, "mind-boggling." In making the comparison to my car in real life, my 1985 BMW 318i (with its 195/60R14 Pirelli's and 20-year-old suspension), I was noting that there is an onramp near my house that is quite similar in sharpness and banked-ness to the second-gear hairpin at the southern-most point of the course, and that my car is capable of carving up that particular onramp at about 65km/h before the tires even start complaining. My point is, the much younger, M-tuned E46 should be able to do significantly better than my car...

After reading your latest post here, I went and attempted another lap, this time aiming for a low time, rather than experimentation with the car and tires, as I did last time. I apologize if you are offended by my pointing this out, but I managed a 2'20"509 on the first lap. I made a couple of mistakes, too, including nearly understeering into a wall at one of the corners (I completely let off of the throttle as soon as the understeer began, but the car kept going, and going, and going...<-- the main reason why I find GT4's understeer to be excessive). This picture here is proof of this lap, and this proof is not really intended for you -- it is simply to protect myself from potential nay-sayers who would be more than glad to yell at a GT4-basher like myself, "you're lying!! You're just slower than him, and aren't willing to admit it, so you blame the game!!"

OK so my original lap time was not the quickest, I admit I should have tried harder.

So I did


Anyway, I've been having a closer look at the 2nd hairpin and using replay data and the data logger have worked out the following.

Length of turn - approx 60 meters
BMW M3 length - 4.492 meters
Length of turn in car lengths - 13.36

Average speed during turn - 43 mph

Time in turn - 5.52 seconds
Time in turn above 1g - 4.68 seconds

The hairpin itself is 180 degrees.

Obviously I have not seen the onramp you describe, but I must confess that if it matches the above its damn tight; because if we trust the GT4 g-meter to be even resonably acurate, then 4.68seconds above 1g in a corner makes it very nasty.

I have to say that I still think that the cornering speed for the 2nd hairpin is not that far out, its a deceptively tight corner with a downhill approach and an uphill exit.

While you're pondering that one, there's another strange GT4 phenomenon my friends and I have noticed that I forgot to mention -- it seems as though GT4 completely ignores the cornering benefits of a banked turn. I haven't done research comparing cornering speeds between GT4 and real life yet, but the Karussell at the Nürburgring seems to be a glaring example of this.

I would agree that the true effects on banking are a weak area of the GT4 physics engine, how far out it is I have no idea, but its an interesting one to think about.

Thank you for the compliment! I disagree with niky that promoting another game while discussing GT4's faults is, by default, a bad thing to do -- and if I'm reading your final statement here correctly, I believe you would agree with me on this -- but when a member such as yourself kindly requests that these outside comparisons be put aside, I'm more than willing to comply. I know I can get heated up sometimes, and express things in a more blunt, slightly exaggerated manner, but I typically post in such a way in topics that are more subjective than objective, and you've certainly set up an objective discussion environment here.

Cheers to the rational thinkers of the GTPlanet community! 👍

I do agree with you that looking at the weak points in the GT4 physics engine (because to say that it was perfect would just be daft) in comparison to other games is a very valid thing to do. With the provision that the game being used ofr comparison is also subject to the same critical analysis.

The reason I created this thread and asked that it be pure GT series was to try an look at these areas without it turning into another GT4 vs the rest. I wanted this because first, a lot of threads already exist looking at 'GT4 vs the rest' and secondly because I wanted to look at how the GT series has evolved over the years, sort of GT vs itself.

Regards

Scaff

 

[Tornado]

I agree that the banking aspect is rather poorly implemented. There is a good example that shows this: High Speed Ring. While I don't pretend to be an expert on aerodynamics or ground effects or any of the ideas implemented into the Group C cars of the past, or the physics of oval courses, I do know from my now-disheveled NASCAR roots that high banking allows for higher cornering speeds. I also know that aerodynamics do as well. My question is, why is it that on the High Speed Ring it is possible to take the banked corners just as fast as the unbanked corners in both cars such as Group C monsters and normal mid engine cars like an NSX? And why is it so impossible to do it in both cases at high speeds in almost any FR car (but for this example I will use the Corvette ZR-1)? It seems like I'm getting ahead of myself, but I do wonder these 2 questions. I believe the former GT games and GT2 in particular implemented oval physics fine. I point out this because of the mid engine cars that would slide around on oval tracks due to rotational mass, such as the Lotuses. That doesn't happen in GT4 and I'm not sure as to wheter it should or not.

 

[Wolfe]

OK so my original lap time was not the quickest, I admit I should have tried harder.

So I did

Cool; as I said, I didn't want to make you look slow, I just wanted to help validate my arguments to non-believers. 👍

Anyway, I've been having a closer look at the 2nd hairpin and using replay data and the data logger have worked out the following.

Length of turn - approx 60 meters
BMW M3 length - 4.492 meters
Length of turn in car lengths - 13.36

Average speed during turn - 43 mph

Time in turn - 5.52 seconds
Time in turn above 1g - 4.68 seconds

The hairpin itself is 180 degrees.

Obviously I have not seen the onramp you describe, but I must confess that if it matches the above its damn tight; because if we trust the GT4 g-meter to be even resonably acurate, then 4.68seconds above 1g in a corner makes it very nasty.

I have to say that I still think that the cornering speed for the 2nd hairpin is not that far out, its a deceptively tight corner with a downhill approach and an uphill exit.

I could have sworn that the onramp was a longer, 270-degree version of the "line" one should take through the 2nd gear hairpin...is the hairpin really that small?

(The one on the top)


In any case, I don't think we'll get anywhere comparing apples and oranges like this, so I'll admit defeat on the 2nd-gear-hairpin-cornering-speed issue. We're straying away from the point I was trying to make anyway.

The point I was trying to make is that GT4's cars are too biased towards understeer; that when given a choice, they would rather plow into a wall than spin out. This imbalance might not be as much of a problem with the understeer simulation as a problem with the oversteer simulation, but the problem is there.

Another issue I have with GT4, which hurts its realism greatly (and is a major contributing factor to my complaints about understeer, oversteer, and overall car behavior), is that the cars simply do not feel like cars. The game is much too mechanical and artificial. I was discussing this in the "Disappointed in GT4?" thread (edited for "other game" content):

GT3, though certainly no more realistic, was just more fun. GT4, as I already stated in the longer post above, took GT3's physics and tweaked them, adding more understeer, and playing with the effects of countersteer (in GT3 countersteer was too weak, and in GT4 it is too strong; both lead to easy spins, but for opposite reasons). I was also noting that I think GT3 did a better job with limited-slip differentials.

How about if I explained that GT4's physics model, just like GT3's, seems to be based around car versus pavement, as opposed to car versus running gear versus pavement, which is what [racing sims are now beginning to] use?

In GT3/4, X = move forward, Square = halt forward motion, Left/right = rotate car left/right. This is what you get when you simplify it all down to the core mechanics. Tack on some "tire grip" values, some calculated bodyroll movements, differential simulation, and several other complications, and the physics become more realistic. But they're still limited by the core functions, which don't follow the physics behind an actual car.

In [some racing sims], "X" = engine rpm's, which transfer power through the transmission to the appropriate drive wheels, as determined by the drivetrain type and differential being used. "Square" = the deceleration of the wheels' rotations. Left/right = rotate front wheels. In these two games, the inputs given to the car by your controller or wheel follow the same logic and basic mechanics behind a real car. The bodyroll, nosedive, tailsquat, and overall weight transfer of these two sims is relatively unregulated and uncalculated, being the result of a mass suspended on four points, each providing a certain amount of "upward" force. Tire grip becomes more important as the power being applied to a certain wheel, braking force being applied to a certain wheel, and the amount of weight being carried by a certain corner of the car all play a role.

To put it very, very simply, GT3/4 is kinda like playing with an RC car with your hands, moving it and turning it as you see fit. [The other racing sims] are like using the remote controller to control the car's power, braking, and steering, letting gravity and physics figure out the rest.

^ Along with all of GT4's physics peculiarities, this also explains why GT4 manages to fail at doing...well, pretty much anything realistically at slow speeds. Donuts, burnouts, even the way cars come to a stop (on one of the M3 at Grand Valley sessions, while experimenting with the car, I slammed on the brakes with the analog stick full-lock to the left. Horridly massive understeer as I slowed down, but right before I came to a stop, the front end suddenly rotated to the right for no reason! ), etc.

I would agree that the true effects on banking are a weak area of the GT4 physics engine, how far out it is I have no idea, but its an interesting one to think about.

Cool, so we're not just crazy...

I do agree with you that looking at the weak points in the GT4 physics engine (because to say that it was perfect would just be daft) in comparison to other games is a very valid thing to do. With the provision that the game being used ofr comparison is also subject to the same critical analysis.

The reason I created this thread and asked that it be pure GT series was to try an look at these areas without it turning into another GT4 vs the rest. I wanted this because first, a lot of threads already exist looking at 'GT4 vs the rest' and secondly because I wanted to look at how the GT series has evolved over the years, sort of GT vs itself.

Regards

Scaff

I understand completely that you wanted to keep this thread GT-only. However, I'm noticing that instead of GT vs itself, it has sort of turned into GT vs Real life...I hope that isn't a problem...

 

[niky]

Hmmm... must check Karussell... I always find that the banking there just sends an FF car to oversteer hell if you try to drive it much above 60-70kph (N-tires) and take the wrong line... which would be around 40mph for you guys? It's a perfectly normal reaction for an FF on uneven ground, the moment the drive wheels are slightly downhill of the rear tires and you're pulling high-G's, it's spin city. (I can attest to that) Banking feels accurate for low-powered FF as compared to real-life... for everything else... well, there's the problem.

Having never been to the ring', I can't vouch for the accuracy of the game throwing cars off the top of the banking when you take it to hot... I'll leave that up to you guys...

As to Toronado's question, I have seen/experienced spinning in MR's due to rotational inertia. Maybe you haven't seen this merely because of another of GT's quirks... super-sticky racing tires. On road tires or worn tires, MRs of all types will spin if you push the rear end too far in a turn, and I've seen/done it on the F1, too.

I frankly quite like the evolving dynamic of GT. After each installment, you get the feeling that the last one was too "arcade-y". That happened between GT1 and GT2, GT2 and GT3, and with GT4, it was a "holy crap, how do you drive this thing!" moment. Once you get used to each one, it feels more natural... of course, some of the same old problems are there, but at least they're changing something.

@Wolfe, as for the understeer bias, like I've said, I think it's mostly due to some juggling with front-end physics to dial in more "realistic" understeer. maybe they tested it and found too little for their taste. And whatever I disagree with you on in other areas ( ), I still agree with your explanation of GT4 physics modelling. Hopefully, the next GT will have a more cohesive physics system rather than one that's "patched together".

A lot of the development time of the previous GTs was given over to rote-work and bulk data capture and programming (with so many cars, this was a given). If they really have improved the encoding method, maybe they can spend a lot more time tweaking the physics.

 

[Scaff]

The point I was trying to make is that GT4's cars are too biased towards understeer; that when given a choice, they would rather plow into a wall than spin out. This imbalance might not be as much of a problem with the understeer simulation as a problem with the oversteer simulation, but the problem is there.

Another issue I have with GT4, which hurts its realism greatly (and is a major contributing factor to my complaints about understeer, oversteer, and overall car behavior), is that the cars simply do not feel like cars. The game is much too mechanical and artificial. I was discussing this in the "Disappointed in GT4?" thread (edited for "other game" content):



^ Along with all of GT4's physics peculiarities, this also explains why GT4 manages to fail at doing...well, pretty much anything realistically at slow speeds. Donuts, burnouts, even the way cars come to a stop (on one of the M3 at Grand Valley sessions, while experimenting with the car, I slammed on the brakes with the analog stick full-lock to the left. Horridly massive understeer as I slowed down, but right before I came to a stop, the front end suddenly rotated to the right for no reason! ), etc.

In regard to the onramp, and just to close this one down. The 2nd hairpin at Grand Valley is far tighter that the onramp in the photo, which would explain your issue with the cornering speeds.

I still think that the understeer physics for GT4 are well developed and resonably accurate, part of my problem with the early GT games (and even GT3) was that the level of understeer was just far, far to low.

The type of understeer most commonly causing a problem in GT4 is that caused by overusing the front tyres. Now while it is relatively straightforward to correct this with a lift of the throttle if you have only overused the fronts by a small degree (true in GT4 and the real world). However, once you overload the fronts by a larger degree (60%+ approx), you get understeer of a much greater magnitune and a lift is not going to solve the problem as it will not reduce the workload on the tyres enough.

Now one thing that years of work on proving grounds has shown is that the difference between small and large levels of over use is a matter of a few mph or degrees of steering angle.

One thing that does make a difference is which 'control method' you use, I find that the DS2 makes it quite difficult to accuratly modulate the throttle, steering and brakes, which makes the possiability of over using tyres even easier.

The DFP helps this to a very large degree as you can be much more subtle with your inputs.

In regard to the oversteer side of things, what GT4 has got right is that for the road cars you do experience a degree of understeer, before oversteer kicks in. Which for a road car (or the vast majority) is correct, its also very much depends on the car and track conditions. With only a few exceptions the tracks in GT4 are all totally dry and give the impression of a high track temp, in these conditions a car such as an M3 will be reluctant to let the back stepout of its own accord, and grip will then tend to recover in a very sharp manner.

One that is worth having a play around with (if you have not already) if the Holden Monaro, much more friendly in terms of power oversteer.

Now, don't get me wrong I still think that GT4 has a long way to go with regard to its physics engine, particularly with regard to very low speed manouvers such as do-nuts and handbrake turns.

I would however dispute that GT3 was closer to reality in terms of the overall engine, and personally I like the 'feel' that GT4 gives you of the car working, particularly in regard to weight transfer and the work that the tyres are being asked to do.

Cool, so we're not just crazy...

Nope, as said its not perfect, but its also not none-existant.

I understand completely that you wanted to keep this thread GT-only. However, I'm noticing that instead of GT vs itself, it has sort of turned into GT vs Real life...I hope that isn't a problem...

No problem with looking at GT vs real life, one of my comments in the original post was that I would bring my own real world experiences into the picture, so I'm fine with that.

Regards

Scaff

 

[Arwin]

@Scaff: Alfaholic has it right about FF's. They don't automatically get "clutched" on handbrake, and you can drag the rear end for a pretty long time with the throttle on full and the handbrake on.

I'm starting to think we just have some confusion on how indicate what a clutch does. I am not sure if I have been using the term correctly. The way I understand it, pressing the clutch pedal disengages the clutch from the flywheel that connects the gear. So when I say the clutch is disengaged, I mean that the clutch pedal is pressed and the clutch and flywheel no longer connect.

If you can drag the rear end with the handbrake on, the clutch is ENGAGED to the flywheel, and the engine moves the wheels. Since the handbrake is applied to the rear wheels, the front wheels are not affected, and they can spin without the car moving.

Whereas if you rev the engine with the clutch disengaged, or in other words, the clutch pedal pressed, the front wheels won't move even in an FF car, as the engine does not bring force to the front wheels at all and can rev freely.

I'll verify tonight if I remembered correctly that you can use the handbrake in the same way as the brake to prevent the clutch (re)engaging so that I can be sure I understand correctly what everyone is saying, and what everyone means with what they're saying.

@Scaff: you are right, unfortunately the clutch is 'properly' and conservatively released, so you can't do as much with it as you'd have liked. But it does seem to help a bit with doing donuts in some cars, even if only slightly, just as a race start from the grid or a timed test in GT can sometimes be done faster if you rev up to 6-7000rpms (as was the case in GT3).