GT S vs Real Life Accuracy

  • Thread starter Yassi_99
  • 27 comments
  • 5,412 views
234
Germany
Germany





I guess this video shows just how realistic even a sim-cade game can look compared to its real life counterpart. Almost everything is spot-on: Cornering speed, brake and shifting points, the acceleration...
The only major fault is the top end speed which is around 30-35km/h too high. I had to lift the throttle to not completely pull away at Döttinger Höhe but I was still faster by quite a big gap. Perhaps the massive aerodynamic parts of the ZL1 in-game do not produce the same amount of air resistance like the real one.
But I'm more than impressed by the accuracy of both car and track. Awesome work, PD!
 
The only major fault is the top end speed which is around 30-35km/h too high. I had to lift the throttle to not completely pull away at Döttinger Höhe but I was still faster by quite a big gap. Perhaps the massive aerodynamic parts of the ZL1 in-game do not produce the same amount of air resistance like the real one.
Temperature and pressure can account for some of that, but nowhere near all of it. I just did a few calculations, and for a fairly extreme comparison, a high air density day at 10C and 1030hPa would only make the car around 9kph slower than a low air density day at 25C and 990hPa (I also corrected Crr for the different temperatures).

It's a bit surprising they don't either get the correct CdA from each manufacturer, or at least estimate it to make the top speed match reality. Or maybe the game assumes a constant CdA with speed but the real CdA varies with speed. A quick search found a paper where the summary suggests vehicles do have a dependency on Reynolds number, but the full text isn't free to read:
https://www.sae.org/publications/technical-papers/content/2015-01-2859/
 
Temperature and pressure can account for some of that, but nowhere near all of it. I just did a few calculations, and for a fairly extreme comparison, a high air density day at 10C and 1030hPa would only make the car around 9kph slower than a low air density day at 25C and 990hPa (I also corrected Crr for the different temperatures).

It's a bit surprising they don't either get the correct CdA from each manufacturer, or at least estimate it to make the top speed match reality. Or maybe the game assumes a constant CdA with speed but the real CdA varies with speed. A quick search found a paper where the summary suggests vehicles do have a dependency on Reynolds number, but the full text isn't free to read:
https://www.sae.org/publications/technical-papers/content/2015-01-2859/
It really depends on the vehicle. The ZL1 with quite high downforce seems way too fast even calculating in all the factors like temperature and so on. The McLaren F1 on the other hand feels too slow in my eyes. I know the vehicle which was driven on Ehra Lessien was a bit customized (higher rev-range...) so that it could reach 390+km/h but the car feels so damn powerless over 330km/h in the game. I don't know if PD is calculating these stats themselves or if they just use the manufacturers info but it sometimes seems quite off.
 
It really depends on the vehicle. The ZL1 with quite high downforce seems way too fast even calculating in all the factors like temperature and so on. The McLaren F1 on the other hand feels too slow in my eyes. I know the vehicle which was driven on Ehra Lessien was a bit customized (higher rev-range...) so that it could reach 390+km/h but the car feels so damn powerless over 330km/h in the game. I don't know if PD is calculating these stats themselves or if they just use the manufacturers info but it sometimes seems quite off.
You obviously don't know about the ground effect issue... PD don't simulate the aerodynamics underneath the car very well, meaning that some cars have too little downforce or too much. This ruins the aero balance of the car. The McLaren F1 struggles to go past 330KM/H because PD has given it too much downforce on top. If PD redistributed the aero balance so that ground effect sucks it down, the car would manage to hit it's claimed top speed because there'll be less aero resistance. The thing about ground effect is that it provides extra grip without increasing drag or creating a large aero wash... Hence why F1 will place more emphasis on it in 2021. But that's another topic.
 
There’s also issues where PD’s physics engine starts to break down at extreme speeds.

Remember doing 300mph wheelies in the Escudo?

In GT5 and 6, the Veryon could reach ridiculous speeds if you raised front ride hight and lowered the rear, which would cause the car to actually start lifting off the road.

In GTS, I can get my Supra to do 400km/h, with no draft, on Tokyo East, with “only” 900 hp

There’s also always been a really big discrepancy between top speeds that can be achieved on a regular circuit, vs top speeds that can be achieved on Route X. Route X has always seemed to have less aero drag than the other circuits, despite it being right at sea level (same could be said for SSR7, but there was always a roughly 30kph difference in the top speed you could achieve on Route X vs SSR7, even when on the downhill run on SSR7).
 
Thats because car does not produce as much as downforce like in real life. 0/50 downforce is a joke. Those splitters and front canards do something in real life.
While we don't know what the value figures are, the entire aero kit is said to produce 300 lbs (136 kg for the world outside the USA) of downforce. That's not a lot - and nobody says how much lift the car generates in its base form. Most of the aero parts' benefit may even go into canceling the lift.

https://www.motortrend.com/cars/che...8-chevrolet-camaro-zl1-1le-first-test-review/

To put that into perspective, F1 and LMP cars produce upwards of 2000 kg at those speeds.

https://www.f1technical.net/forum/viewtopic.php?t=17209&start=15
 
While we don't know what the value figures are,

We actually have some clue about those values.

"A push-rod type double wishbone suspension is mounted in the front and rear, giving it a structure like that of a pure racing car. The vehicle bottom is flat with a large diffuser at the rear end which produces 200 kg of downforce at a speed of 200 km/h. These aerodynamics allow the car to pull 1.5 horizontal Gs while cornering."

This writes on the Ktm Xbow description. It clearly says rear diffuser produces 200kg of downforce while going at 200 km/h. In game stats Xbow has 300 downforce at rear. Since Xbow also has a rear spoiler which probably produces 100 downforce so the total rear downforce is 300.

With that logic if game says 50 that means Camaro produce 50 kg of downforce at rear while going at 200 km/h. (Probably)
 
We actually have some clue about those values. *snip* With that logic if game says 50 that means Camaro produce 50 kg of downforce at rear while going at 200 km/h. (Probably)
Not a bad one! If I've got my physics anywhere near right, 50 kg at 200 km/h would mean around 80 kg at around 250 km/h - the entire aero kit is quoted at 136 kg at 155 mph. With a perhaps 40/60 split between front and rear it would mean around 55 kg at the front and 80 kg at the rear... believe it or not but I estimated the split first and then calculated those numbers.

Now, if the non-ZL1 Camaro actually produced around 50 kg of lift at the front at 200 km/h - not entirely unreasonable seeing how most cars indeed do produce lift - and none at the rear, it would mean that the 0/50 downforce is actually pretty close to the truth. It's definitely right in the ballpark and as little as 50/100 would already be way too much since the quoted 136 kg is produced at a noticably higher speed.
 
Not a bad one! If I've got my physics anywhere near right, 50 kg at 200 km/h would mean around 80 kg at around 250 km/h - the entire aero kit is quoted at 136 kg at 155 mph. With a perhaps 40/60 split between front and rear it would mean around 55 kg at the front and 80 kg at the rear... believe it or not but I estimated the split first and then calculated those numbers.

Now, if the non-ZL1 Camaro actually produced around 50 kg of lift at the front at 200 km/h - not entirely unreasonable seeing how most cars indeed do produce lift - and none at the rear, it would mean that the 0/50 downforce is actually pretty close to the truth. It's definitely right in the ballpark and as little as 50/100 would already be way too much since the quoted 136 kg is produced at a noticably higher speed.
If downforce is 50 kmh while going at 200kmh then the downforce is 62.5 kg while going at 250 kmh. 62.5 is nowhere close to 136 kg. Its not even half.

Camaro deserves some front downforce imho. I dont understand why Pd gives 0 downforce to front of almost every N class cars.
 
So there's definitely something off with GTs simulation of aerodynamics and the ground-effect (atleast on road cars). The system seems to work quite ok on slower to midspeed sections but it completely fails at higher speeds. Both front and rear downforce are almost non existent at some vehicles (especially the Aston Martin Vulcan and the Ferrari LaFerrari) even though they should sometimes produce as much as some race cars.

Another thing, which has to be fixed IMO is the redline sound/bump. Some cars have such a nice sounding rev-limiter but in GT, the car loses a lot of noise and bass during this act (while hitting the limiter with a way too high frequency). Other games are far more progressed in this regard.
 
How 'bout turbo lag? Is it a million times the same as it is in real life? 💡:D
Oh daaaamn. Almost forgot this. Cars like the F40 would be a gazillion times more fun to tame if the turbos really had to build up their boost. I think this was one of the aspects which were better in GT5 and GT6 (if I remember correctly). As you can see, the next GT has multiple possibilities to improve the quality even further :)
 
Oh daaaamn. Almost forgot this. Cars like the F40 would be a gazillion times more fun to tame if the turbos really had to build up their boost. I think this was one of the aspects which were better in GT5 and GT6 (if I remember correctly). As you can see, the next GT has multiple possibilities to improve the quality even further :)
Oh right. Glad to know you are accurate with that one. :lol:

Also, automated transmissions in the game need to be more realistic, too.
 
The turbo boost has something going on with it. Some cars, at least BMW M6 GT3 and Honda NSX GT3 off the top of my head, act differently to what the boost gauge says. While the gauge indicates that full boost is available as soon as the throttle is floored the actual engine behaviour seems to suggest that there indeed is turbo lag, picking up quite sluggishly for a short while before the full power comes on even in mid to high rpms where boost threshold isn't an issue. Very different from NA engines that give instant power as they should.
 
Oh daaaamn. Almost forgot this. Cars like the F40 would be a gazillion times more fun to tame if the turbos really had to build up their boost. I think this was one of the aspects which were better in GT5 and GT6 (if I remember correctly). As you can see, the next GT has multiple possibilities to improve the quality even further :)

Might explain why the F40 is so much faster in GT Sport, compared to GT5 or GT6.
 





I guess this video shows just how realistic even a sim-cade game can look compared to its real life counterpart. Almost everything is spot-on: Cornering speed, brake and shifting points, the acceleration...
The only major fault is the top end speed which is around 30-35km/h too high. I had to lift the throttle to not completely pull away at Döttinger Höhe but I was still faster by quite a big gap. Perhaps the massive aerodynamic parts of the ZL1 in-game do not produce the same amount of air resistance like the real one.
But I'm more than impressed by the accuracy of both car and track. Awesome work, PD!

I did a quick test with the camaro with same stock power and weight on sport tyres. I can say i'm very impressed by the result. Exactly the same T-pass and sped reached at crucial point of the track.
Here it is.
 
We didn't compare the track though. I don't know in which year PD scanned the Nurb but the result is really outstanding. Visually, it has to be the best representation ever. I can't say anything about the physical aspects as I don't have a FFB wheel. I guess it won't be on par with true sims like Assetto or R3E but it has to be realistic enough to achieve such accuracy on track.
 
Should I do more videos like this? Of course it has to be vehicles which lapped the Ring IRL and have an onboard video on YouTube. Any suggestions?

I'm currently having those in mind:

Pagani Zonda R
Nissan GT-R NISMO
Honda Civic Type-R
Honda NSX
 





I guess this video shows just how realistic even a sim-cade game can look compared to its real life counterpart. Almost everything is spot-on: Cornering speed, brake and shifting points, the acceleration...
The only major fault is the top end speed which is around 30-35km/h too high. I had to lift the throttle to not completely pull away at Döttinger Höhe but I was still faster by quite a big gap. Perhaps the massive aerodynamic parts of the ZL1 in-game do not produce the same amount of air resistance like the real one.
But I'm more than impressed by the accuracy of both car and track. Awesome work, PD!

Sport tires are the most realistic. Race tires are way too fast though, just look at this:
https://www.blancpain-gt-series.com...id=9&filter_meeting_id=129&filter_race_id=903. They're doing 44s in soft tyres while in GT Sport the aliens can do that on hards. And yes I know that there's BOP but usually the cars are slower under BOP, I can do 44s in the Merc on hards without BOP. Hopefully the tyre partnership with Michelin will change this.
 
Sport tires are the most realistic. Race tires are way too fast though, just look at this:
https://www.blancpain-gt-series.com...id=9&filter_meeting_id=129&filter_race_id=903. They're doing 44s in soft tyres while in GT Sport the aliens can do that on hards. And yes I know that there's BOP but usually the cars are slower under BOP, I can do 44s in the Merc on hards without BOP. Hopefully the tyre partnership with Michelin will change this.
PD may not be trying to simulate the Pirelli tire, which is slower than Michelin on most occasions. On top of that, the drivers cut the last chicane a lot on GTS.

Super GT qualifying and race times from last year's Suzuka race are really close to what top drivers do in GT Sport. And, again, if the track limits were respect it would've been even closer.

https://supergt.net/results/index/2018/Round3/3/gt300
 
Real life GT3 cars are running very low power figures even when compared to the the GTS BoP, let alone the stock power of some cars. From this year's Nürburgring 24h:

Audi R8 LMS
Mindestgewicht: 1.310 Kilogramm
Luftmengenbegrenzer: 2x38,0 Millimeter
Max. Leistung Motor-/Rollenprüfstand: 468/487 PS
Maximales Kraftstoff-Volumen/-Füllmenge: 115/112 Liter

BMW M6 GT3
Mindestgewicht: 1.335 Kilogramm
Luftmengenbegrenzer: 2x32,7 Millimeter
Max. Leistung Motor-/Rollenprüfstand: 493/513 PS
Maximal zulässiger Ladedruck absolut: 1.715 Millibar
Maximales Kraftstoff-Volumen/-Füllmenge: 119/116 Liter

Lamborghini Huracan GT3
Mindestgewicht: 1.300 Kilogramm
Luftmengenbegrenzer: 2x38,0 Millimeter
Max. Leistung Motor-/Rollenprüfstand: 473/492 PS
Maximales Kraftstoff-Volumen/-Füllmenge: 118/115 Liter

Lexus RC-F GT3
Mindestgewicht: 1.325 Kilogramm
Luftmengenbegrenzer: 1x35,0 oder 1x36,0 Millimeter
Max. Leistung Motor-/Rollenprüfstand: 506/526 PS
Maximales Kraftstoff-Volumen/-Füllmenge: 117/114 Liter

Mercedes-AMG GT3
Mindestgewicht: 1.355 Kilogramm
Luftmengenbegrenzer: 2x33,5 Millimeter
Max. Leistung Motor-/Rollenprüfstand: 496/516 PS
Maximales Kraftstoff-Volumen/-Füllmenge: 121/118 Liter

Nissan GT-R Nismo GT3
Mindestgewicht: 1.310 Kilogramm
Luftmengenbegrenzer: 2x32,9 Millimeter
Max. Leistung Motor-/Rollenprüfstand: 522/542 PS
Maximal zulässiger Ladedruck absolut: 1.740 Millibar
Maximales Kraftstoff-Volumen/-Füllmenge: 119/116 Liter
 
Back