Real Life x SIM - Miata at Interlagos.

Guys,

I've been using Assetto Corsa as a training tool for a Rally championship that takes place at Interlagos (Brasil).

I drive a 1995 Miata (1.8L) and I can say I got quite good at it. I won both championships (2017 and 2018).

Interlagos racetrack has been the same for these past two years and the Assetto Corsa version is a real representation of what we have nowadays.

Yesterday we had the 4th and last round of the 2018 championship. I set my time at the 2m:58s mark and repeated that mark for 13 laps - give or take 1 or 2 seconds.

So, I went back to Assetto Corsa and did exactly the same I had done in the track - same final speeds (controled by the tach instead of the speedometer) same shifting points (also controled by the tach). I can say I was able to replicate what happend in the track.

I even took care to use the same tire pressure, street tires, same gas level and same suspension setting I use in my car.

BUT ... the Assetto Corsa times were 10 seconds faster than real life. It seems too much of a difference.

On the track I was doing 02:58 while on Assetto Corsa, 02:48.

I would love to hear from you guys, that are much more experienced than me on Assetto Corsa what could be causing these huge difference?

I race with a co-driver and some stuff in the trunk. All togheter - including my weight - the Miata must be carrying about 200 kg. I don't know if AC already takes into account the driver's weight and how I cad add this extra weight to the car - anyway ... it wouldn't be 10 seconds slower.

Any comments on the above?

Thanks

by the way ... the championshio site is: http://www.torneiointerlagos.com.br/

 

Well its not a laser scanned track for a start, so even tho the track may look close it could of been slightly off, even corners being slightly off could make a decent time difference.

 

I can't comment on the difference in time, but, if you like, i'd hear you comment on the general feel of the car in AC as well. Does it feel and handle like the real thing?

 

Well ... as it is a regularity rally we are far from driving at the limit - although I take the curves quite fast and use the straights to fine tune the lap time. The virtual Miata feels as precise as the real one specialy at the Mergulho Curve, an off camber curve that pushes the car to the outside at the exit.

The only modification my Miata has is a set of much stronger sway bars and it makes the car just a bit more precise than the virtual one - I can notice that at the Curva do Lago (by the end of the straight).

Also, I was not able to replicate the front camber I'm using, which is -1 degree. The rear is set up at -1,5 degrees in the rear - I'm still a little confused on how AC handles camber and convergence.

I was quite disppointed to find a 10 seconds difference between reality and AC.

Anyway ... despite that ... AC is a great tool to train for the Rally. It sure helped.

 

Thanks.

Interlagos is one of my favorite tracks BTW. Just lovely, that thing.

BTW, you can check out the same car/track combo in iRacing, and you even have a laser scanned track then. If you don't dread the subscription. I think the Miata is one of the free cars included in the initial content for the subscription. Interlagos won't be free though.

 

ChrisR, would the computer power make any difference on lap timing?

I use AC in a undepowered notebook and it is only good when I turn off almost all the bells and whistles and use a sinlge car.

I use only for training - computer doesn't have enough power to handle onther cars simultaneously.

 

chakko - I did use rFactor 1 - which is also faster for the same imputs. But the Miata I available at the rFactor is rather different than the stock - tires, suspension, brakes ... etc ... it is basicaly a 1.6 race prepared.

If you have the opportunity, try the old (almost 8 km lap) Interlagos available at rFactor. That was a real race track. The current one, although good to drive, is like a slot car race compared to the old one - you can still see parts of the old one on GoogleEarth.

 

Hate to be the guy but Wrong Forum Section..
and as for Interlagos it is mod track so it is not true 1:1 representation of the track.. it could be track problem. if you matched your IRL car settup with game one.

But cool to hear how car handles in game and IRL.

 

as far as im aware, not any more.. there used to be an issue in the tech demo where the more fps you had the faster your laptimes... but this was fixed almost instantly.

 

Tyres would be the obvious difference, Kunos don't make tyres (but if they did they'd be the best tyres in the world, (spot the brew)). Tyres can easily gain or lose seconds.

I think one major overlooked factor between sim and real life is, fear. In the real world I'm scared of being hideously disfigured, I could take dying over being horribly disfigured. I'm scared of that, in a sim I'm not afraid at all, in the slightest, it doesn't even enter the equation.

Doing 100 kph average speed to the next town from me would scare the living bejebus out any normal person.

Sim has no personal risk like even driving the car to the shops does, I honestly think that has to play a part in faster times in sims over real life.

So fear and tyres can easily sway the time by seconds, even if you've removed all fear factors tyres can still make a huge difference and unless your working off a proper lazer scanned tracks it's basically impossible to compare times.

 

I think the first check you can do is measuring the length of the Interlagos version you drive in AC and compare it with the real life data, if available. I remember a version of Laguna Seca available for netkar Pro (like 7 or 8 years ago) that was around 250mts shorter than the real counterpart. I think you can have the length of a lap of yours in AC using a software like AiM Race Studio or similar to open the files you find inside Documents\Assetto Corsa\aim.
There can be other reasons of course.. Anyways: real lap is 2':58", as to say 178" and AC lap is 168", the different is more than 5%... Knowing how the guys from Kunos work, I would say it is too big of a difference for them, provided that the car is reasonably the same you drive. And, last but not least: Grats!!! for winning two years in a row, super job!

 

What track grip setting you use while hotlapping? If it's "optimal" which is default setting, it's definitely too grippy. Use at least "green" or even slower ones depending on a track.

 

Maybe the Interlagos mod is not accurate or they set the grip too high. If you are running at 100% grip in-game you would probably lose a couple of seconds by lowering it to 97 or 98%. If you record your real-life laps you could save a replay in AC and compare them side by side to get a better idea what the difference is.

 

Ok ... this is interesting. I found out why the 10 seconds difference!

It happens that I use the tach to control the car speed at the straights ( 4th gear at 4750 rpm ) and I down shift to 3rd for the fast curves and take then at 80 km/h. The slower curves I down shift to 2nd gear and take then at 60 km/h. Then I try to keep a constant speed through the curves and at the curves exit I shift up to 3rd at 6.000 rpm and to 4th at 5000 rpm and keep 4750 rpm on 4th until next curve. Well ... I just gave up my secret!

ANYWAY ... it happens that I forgot to consider the speedometer error!!! So ... I checked the curves speeds in an old Harry's Lap Timer data I had and ... when the speedometer shows 80 km/h ... it is more like 7oish km/h ... and when it shows 60 km/h ... in reality it is more like 50ish km/h.

I went back to AC and used those data ... BINGO ... lap time> 02:58:775

AC is A awesome simulator.

Even the shifting up points are much closer to real world using the correct curve's speed.

BUT ... it feels awfully slower than the real thing ... well ... I use AC for the real thing ... so ... another score for AC.

 

Well done for finding the issue Marcelo! You can always increase FOV for more sense of speed but I wouldn't recommend going too high, you will never have the same sense of speed as real life but atleast you have a way to practise and fine tune your technique from home now. No excuse not to win in 2019!

 

Real race cars use just a tacho
I'll try the Miata again when I drive in AC next time.

 

Sense of Speed will seem much lower in Sims, unless you have a wrap-around projection setup covering your full view. VR is better than a single monitor imo but, it's still a very limited FOV.

 

You can always setup a small wind turbine and fly cannon to shoot bugs in your face for added sense of speed but I find I get a bit of a dodgy stomach if I consume more than 2 or 3 kilos of flies a day, no pain no gain, sim racing is all about sacrifice. I'm sure the local insect population would agree.

 

Happy that you found the reason!
Infact tachometers have a margin of 5 - 10% over real speed. I think it is to limit the effective speed in everyday driving, but who knows

Me being a curious person, I wanted to check the length of some of the Interlagos tracks available for AC, so I tried some laps and exported the telemetry dump to Race Studio to check lengths.
No idea if the mods I tried are legit or not. I used the same setup, but nothing fancy, street tyres with even a wrong pressure, but whatever.
I listed the distance I ran and the laptime. Official track length is 4309 mts, but usually they measure it keeping the center of the road, so a typical race line should be shorter by quite some meters.

old interlagos version of 2014 (from racedepartment.com)
length of my lines in a lap: 4229 mts - lap time:2:13.4

acu interlagos (from assettocorsa.club)
length of my lines in a lap: 4231 mts - lap time: 2:14.2

beta interlagos (from mega.co.nz)
length of my lines in a lap: 4254 mts - lap time: 2:12.7


I tried to be consistent, but I don't think I succeded in that
Strange enough: the fastest lap has the longest distance, maybe I was faster, or the grip of that version is a bit higher than the other 2.

 

Probably they want it so over a variety of configurations (winter tires, higher tire pressure, tiny spare tire, whatever) it never shows less than the real speed, cause you want to drive at the upper speed limit without breaking the law more often than the lower speed limit.

iirc AC speedos generally show gps speed rather than wheelspeed too so slip doesn't come into play.