Tire Testing - Strange, but interesting results...
- Thread starter calan_svc
- 261 comments
- 60,584 views
- 1,929
- Finland
- HI-tauer
Just now found this interestin test. Lets see if it gets some more conversation. Must say im a bit let down if that .6 g is only difference that changing tyres make.
dont know if these tests still are accurate for the current version of the game, but i did come up with a fact that makes this even more complicated.
"the size of the contact patch is not figured into GT5's tire equations."
if this is correct, trying different tires and measuring the final g achieved doesnt mean the car has the behaviour of the real life stock car. Given the fact that most of the faster cars, like the zr1, have very different tire size on front and rear axle. This lead to point, that the traction per axle/tyre might still be off even the cornering g matches, hense cars characteristics are different fom real one. If so - simulating the characteristics would require combining and testin different tyres on front/rear axles.
If you read this far, I can tell that I wont be bothered by such a minor imperfection in the game. These tests and conclusions just tell how hard a racing sim is to make without proper tire physics including the change of width/profile of the tyre.
dont know if these tests still are accurate for the current version of the game, but i did come up with a fact that makes this even more complicated.
"the size of the contact patch is not figured into GT5's tire equations."
if this is correct, trying different tires and measuring the final g achieved doesnt mean the car has the behaviour of the real life stock car. Given the fact that most of the faster cars, like the zr1, have very different tire size on front and rear axle. This lead to point, that the traction per axle/tyre might still be off even the cornering g matches, hense cars characteristics are different fom real one. If so - simulating the characteristics would require combining and testin different tyres on front/rear axles.
If you read this far, I can tell that I wont be bothered by such a minor imperfection in the game. These tests and conclusions just tell how hard a racing sim is to make without proper tire physics including the change of width/profile of the tyre.
- 377
- Enzo_Guy
- GTP EnzoGuy
Just now found this interestin test. Lets see if it gets some more conversation. Must say im a bit let down if that .6 g is only difference that changing tyres make.
dont know if these tests still are accurate for the current version of the game, but i did come up with a fact that makes this even more complicated.
"the size of the contact patch is not figured into GT5's tire equations."
if this is correct, trying different tires and measuring the final g achieved doesnt mean the car has the behaviour of the real life stock car. Given the fact that most of the faster cars, like the zr1, have very different tire size on front and rear axle. This lead to point, that the traction per axle/tyre might still be off even the cornering g matches, hense cars characteristics are different fom real one. If so - simulating the characteristics would require combining and testin different tyres on front/rear axles.
If you read this far, I can tell that I wont be bothered by such a minor imperfection in the game. These tests and conclusions just tell how hard a racing sim is to make without proper tire physics including the change of width/profile of the tyre.
I think I can offer some proof that there is at least some provision for contact patch between cars, whether it calculates for separate axles I don't know. If we look at the '74 Lamborghini Countach LP400 compared to the '88 Countach 25th Anniversary (essentially a Quattrovalvole underneath) the major difference, disregarding engine changes, in the characteristics of these cars lies in the tire sizes. The LP400 has a 205 or 215 front and rear (if my memory is correct) as opposed to the 25th wearing 225 in front and no less than 345 in the rear!
I haven't done any kind of testing but driving these cars in stock format back-to-back yields very noticeable differences in handling.
- 123
- Poland
- Szejok666
I know this is an old thread, but I did some testing recently, checking if the lateral G numbers changed with last updates.
Cars used were Aston Martin DB7 Vantage and Corvette ZR-1 (1990 - the old one).
First, I tested both on TGTT, and the results were identical for both cars (CH - 0.80, CM - 088), just like Calan's.
But then I switched track to SS Route 5, then to Tsukuba.
The results were different: on CH I get 0.88, and on comfort mediums 0.94/0.95. Can't tell the exact number, but It looks like the grip was higher by a value similar to fitting one-step better tires.
Conclusion: Grip is not only dependable on the tires, but also on track surface.
Cars used were Aston Martin DB7 Vantage and Corvette ZR-1 (1990 - the old one).
First, I tested both on TGTT, and the results were identical for both cars (CH - 0.80, CM - 088), just like Calan's.
But then I switched track to SS Route 5, then to Tsukuba.
The results were different: on CH I get 0.88, and on comfort mediums 0.94/0.95. Can't tell the exact number, but It looks like the grip was higher by a value similar to fitting one-step better tires.
Conclusion: Grip is not only dependable on the tires, but also on track surface.
- 7,765
- Yorkshire
- ScapeGoat4U
- WhippingBoy
Interesting results!
Another factor that might negate calan's findings, is whether he conducted his original test online or offline. Before the 2.09 physics update there was a considerable difference in grip levels between online & offline.
Anyway, at least this thread got me to thinking seriously about which tyres to put on a car, & not always to go with the stock options.
Another factor that might negate calan's findings, is whether he conducted his original test online or offline. Before the 2.09 physics update there was a considerable difference in grip levels between online & offline.
Anyway, at least this thread got me to thinking seriously about which tyres to put on a car, & not always to go with the stock options.
- 123
- Poland
- Szejok666
There is another interesting thing: lateral G does depend on the car. I just get an average 1.07 in Viper SRT-10 Coupe on Comfort Softs. According to Calan's research it should only get 0.94 on these tires. Couple hours ago i tested Toyota GT-86, it get 0.90-0.92 on Comfort Hards (very close to real). The results were achieved on flat surface, doing 65-80 km/h, maintaing a steady radius for about 4-5 seconds.
Edit: I may have been wrong with different grip on various tracks - it seems that they are comparable. But I'm (probably
) not wrong with lateral G results. Viper SRT-10 Coupe achieved average 1.06-1.07G on both SS route 8 and TGTT on Comfort Softs. Viper GTS '99 (Standard one) did about 0.96-097 on Comfort Mediums.
Also, I recently compared real laps vs virtual ones. First test was comparison between new Honda Civic Type R (JDM model) at Suzuka. There is onboard footage available on YouTube, laptime was 2.35 I think. I managed to do this on Comfort Soft tires. Comparing onboard footage from the real lap to my lap in GT5, there was almost exactly the same braking points, racing lines and corner speeds. I admit I'm a pad user (unfortunately) and I was not that perfectly consistent in the corners, but car was slightly faster in game, so I managed to keep up.
Second test was also at Suzuka, I grabbed 1997 NSX Type S Zero and compared to actual footage of NSX Type R (old one - 1992 I think, also on youtube). In reality these cars lapped Suzuka in almost exactly same time (2.34 I think). In game I fitted NSX with Comfort Mediums. This resulted in almost perfect similarity between real and virtual lap. Later I tested this NSX Type S Zero for lateral G. It scored 0.94-0.95 on CM tires.
It would be great and interesting, if someone with a wheel could do some testing. If actual G reading and lap times that can be achieved will match, we could find the stock tires equivalent for each car - something, that Polyphony Digital forgot to tell us.
Edit: I may have been wrong with different grip on various tracks - it seems that they are comparable. But I'm (probably
) not wrong with lateral G results. Viper SRT-10 Coupe achieved average 1.06-1.07G on both SS route 8 and TGTT on Comfort Softs. Viper GTS '99 (Standard one) did about 0.96-097 on Comfort Mediums.Also, I recently compared real laps vs virtual ones. First test was comparison between new Honda Civic Type R (JDM model) at Suzuka. There is onboard footage available on YouTube, laptime was 2.35 I think. I managed to do this on Comfort Soft tires. Comparing onboard footage from the real lap to my lap in GT5, there was almost exactly the same braking points, racing lines and corner speeds. I admit I'm a pad user (unfortunately) and I was not that perfectly consistent in the corners, but car was slightly faster in game, so I managed to keep up.
Second test was also at Suzuka, I grabbed 1997 NSX Type S Zero and compared to actual footage of NSX Type R (old one - 1992 I think, also on youtube). In reality these cars lapped Suzuka in almost exactly same time (2.34 I think). In game I fitted NSX with Comfort Mediums. This resulted in almost perfect similarity between real and virtual lap. Later I tested this NSX Type S Zero for lateral G. It scored 0.94-0.95 on CM tires.
It would be great and interesting, if someone with a wheel could do some testing. If actual G reading and lap times that can be achieved will match, we could find the stock tires equivalent for each car - something, that Polyphony Digital forgot to tell us.
Last edited:
- 269
- St.Petersburg
- Olegonic
I used real life car's lateral G before to equip GT5 car with tires similar to that G.
I did some time attacks on GT5 tracks which were used on same car and track in real life. Then I found one interesting thing:
For instance, I take Honda NSX-R '02 and do time attack on Tsukuba, I manage to make the same lap time like in real life on Comfort Soft tires. That was easy.
Then I took same Honda on Nordschlife and my best time was 7.56.xx but that was REALLY extreme, but not the fastest driving, I had to push my car hard which if compared to real life is way more aggressive. If you watch the video of HONDA NSX-R Nurburgring, the driver goes very smoothly, without sliding. To match his time in GT5 you have to be on the edge of sliding all way till the finish line. Yes, I know that on that video his time ends whin hi hits the last corener, in GT5 you need to drive 100 meters more. So his time would be 8.00, not 7.56.xx.
Plus, my Comfort Soft tires were squiling loud almost all the way, when his tires just did it few times. You can hear that on video as well.
Conclusion, for times attacks in real life they use cars which could have been possibly tweaked (suspension), and of course they install better tires than commercial version, because that hot lap will represent a car's performance on famous track like Nurb.
And not always the very same car did hot lap on different tracks which means that performance may vary from track to another. Thus, for modelling GT5 conditions of time attack in Real Life it is not enough sometimes just to equip car with appropriate tires according to real G. It is important to see video and hear how Real car behaves on track. Moreover, real drivers would never drive a car on a track like we do in game. they have fear, reasonable limit, human factor and the don't have restart button, so my opinion better tires for HONDA NSX-R '02 would be Sport Hard if not Sport Medium to make car corner and behave similar to real life. Good racer in GT5 must always show the better time than in real life due to factors I have mentioned above.
Thank you.
I did some time attacks on GT5 tracks which were used on same car and track in real life. Then I found one interesting thing:
For instance, I take Honda NSX-R '02 and do time attack on Tsukuba, I manage to make the same lap time like in real life on Comfort Soft tires. That was easy.
Then I took same Honda on Nordschlife and my best time was 7.56.xx but that was REALLY extreme, but not the fastest driving, I had to push my car hard which if compared to real life is way more aggressive. If you watch the video of HONDA NSX-R Nurburgring, the driver goes very smoothly, without sliding. To match his time in GT5 you have to be on the edge of sliding all way till the finish line. Yes, I know that on that video his time ends whin hi hits the last corener, in GT5 you need to drive 100 meters more. So his time would be 8.00, not 7.56.xx.
Plus, my Comfort Soft tires were squiling loud almost all the way, when his tires just did it few times. You can hear that on video as well.
Conclusion, for times attacks in real life they use cars which could have been possibly tweaked (suspension), and of course they install better tires than commercial version, because that hot lap will represent a car's performance on famous track like Nurb.
And not always the very same car did hot lap on different tracks which means that performance may vary from track to another. Thus, for modelling GT5 conditions of time attack in Real Life it is not enough sometimes just to equip car with appropriate tires according to real G. It is important to see video and hear how Real car behaves on track. Moreover, real drivers would never drive a car on a track like we do in game. they have fear, reasonable limit, human factor and the don't have restart button, so my opinion better tires for HONDA NSX-R '02 would be Sport Hard if not Sport Medium to make car corner and behave similar to real life. Good racer in GT5 must always show the better time than in real life due to factors I have mentioned above.
Thank you.
Last edited:
- 123
- Poland
- Szejok666
You're right I think
I have also tested Honda NSX-R on various tracks, and I've seen the video you've mentioned. Actually, most of real life Nordshleife laps are measured from "bridge to gantry", which is from the start line to the line just after last corner. During my lap in GT5 I've managed to achieve lap time of 7.53 on Comfort Soft, but it was very, very aggresive and almost constantly at the limit. There is also one factor that is very significant in GT5, particularly on fast tracks like Nurburgring. Most cars (apart from Toyota GT-86 I think) are considerably faster than in real life. It can be seen for instance comparing real and GT5 lap of Ferrari Enzo on the Nurb. While IRL on the last straight it achieved speed close to 300 km/h, in game I was doing about 340.
It would be great if PD give us players for instance a list of stock equivalent tires for each car, but I don't really think that they know, which is most true to life after all those updates, especially for standard cars.
I have also tested Honda NSX-R on various tracks, and I've seen the video you've mentioned. Actually, most of real life Nordshleife laps are measured from "bridge to gantry", which is from the start line to the line just after last corner. During my lap in GT5 I've managed to achieve lap time of 7.53 on Comfort Soft, but it was very, very aggresive and almost constantly at the limit. There is also one factor that is very significant in GT5, particularly on fast tracks like Nurburgring. Most cars (apart from Toyota GT-86 I think) are considerably faster than in real life. It can be seen for instance comparing real and GT5 lap of Ferrari Enzo on the Nurb. While IRL on the last straight it achieved speed close to 300 km/h, in game I was doing about 340.
It would be great if PD give us players for instance a list of stock equivalent tires for each car, but I don't really think that they know, which is most true to life after all those updates, especially for standard cars.
Last edited:
- 1,233
- Patra
- Rotorist
There IS grip difference between tracks or road courses in GT5. For example, Cote d' Azur vs Motegi or Tsukyba or Fuji feel very different in the same vehicle and similar turns. Anyone remembers Fuji 80's, 90's and 2005 from GT4 having different pavement and grip levels?
- 123
- Poland
- Szejok666
After watching many Best Motoring episodes I can tell that lap times on Tsukuba track for same car may differ 2-3 seconds because of temperature. For example best lap for Honda NSX Type S Zero is 1:05 seconds and it was done in winter with about 0°C. In summer and temp about 30°C same car lapped this same track in 1:07.30. I don't think it's possible in GT5 to lap this track with comfort medium tires in 1:05. On the other hand, on Suzuka it's easy to match 2:34 on comfort mediums - which is the record for this car on that track.
- 7,765
- Yorkshire
- ScapeGoat4U
- WhippingBoy
@2_FAST_4_U @Tim Voak @racecaraz @ShphrdOfFire
If you guys haven't tried this already, I highly recommend it. It really brings the cars to life! See the OP & post 81 for more details.
https://www.gtplanet.net/forum/thre...nteresting-results.160821/page-3#post-4638146
👍
- 131
- Canada
- Kodiak77773
- 7,765
- Yorkshire
- ScapeGoat4U
- WhippingBoy
It's a rough guide for choosing tyres for cars in GT5 that have the same or similar performance as their real life counterparts. Many of the default tyres in GT5 are totally inappropriate for various cars, which just serves to unnecessarily dull the physics model. This will only work well on factory stock/untuned cars. Just find the car's real life lateral g ratings from a skid pad test (Google it), & see what the Tyre Calculator recommends.
👍
- 918
- U.S.A.
True Dat VBR 👍 ... I mean are comfort tires really needed default on a race car LOL ! 
- 7,765
- Yorkshire
- ScapeGoat4U
- WhippingBoy
I mean are comfort tires really needed default on a race car LOL !
It all depends on the race car, & yes sometimes they would need Comfort tyres (although I don't know of any in GT5 that come with C tyres as default). If a race car pulled between 0.87 & 0.94g on the skid pad, then CS tyres would be the right choice for that car. According to my skid pad numbers for the Elise 111R, it should be on SM tyres (one level up from default). And according to numbers that Calan found for the Mclaren F1, it should be on CM tyres!
Some cars default tyres are near their real world counter parts, but many are not, & some are way off.When the game first came out I was bitterly disappointed with the numb & lifeless physics model. But the more I read this thread & tried Calan's tyre choice method, the more I realised the physics model isn't all that bad; it was just being smothered by the unrealistic levels of grip inflicted by PD's inappropriate tyre assignments. It's not perfect, & there are some potential holes in the theory, but it certainly breathes new life into the handling of many cars & allows the physics to shine through.
Personally I don't like the race cars in GT games, mainly because of the awful tranny whine, but also because of their lack of character (I love race cars in Project CARS, they got them right). So, in GT games I tend to stick to untuned road cars. And if married with appropriate tyres that reflect the cars real world performance, things can start to get really interesting!
EDIT: I used to run a club on here called Saturday Stock Car Series back in 2011. We used untuned road cars only, & chose tyres base of this method. It was a lot of fun & I have a ton of replays from those events.
Last edited:
- 918
- U.S.A.
Project Cars-What's that ? Must be one of those new games about doing projects trying to fix cars to work correctly-lol ! To much of a project for me... besides, I hate just renting cars that are really somewhere in a company owned internet garage somewhere up in the cloud only later to be taken down and lose all those beauties back into cyberspace-lol !
- 7,765
- Yorkshire
- ScapeGoat4U
- WhippingBoy
P CARS doesn't do cloud saves, it's saved on the PS4, & it's not possible to lose cars anyway because they're all available from day one (nothing is locked). There are good points & bad points to all games I've played, GT & P CARS included. Don't knock it till you've tried it.
Two of my favourite stock drives in GT5;
1) Dodge Challenger SRT8 2008 on Comfort Mediums at Daytona Road, with all driver assists off except ABS set to 1.
2) Chevrolet Camaro SS 2010 on Comfort Soft tyres at Indy Road, with all driver assists off except ABS set to 1.
Cars must be factory stock, no settings or tuning tweaks at all, apart from the change to specified tyres & turning assists off. These were some of my favourite car/track combos to race online, & I have several clean race replays of battles in public lobbies with these.
- 918
- U.S.A.
Ok I concede ! 
Similar threads
