I was struggling with this on my first outing last night, nothing to do with my FFB so it's either a Mod or Setup issue. Hopefully get some time tonight to try a few things out.
I personally prefer data to feelings, unless my RF2 instalation and Motec are working in a strange way, if I run 1.8 version with default setup at Sao Paolo, with my channels smoothed by 50 in Motec, in the second straight I read the following: - at 257 km/h Front around 9 mm, Rear 2,8 mm. - at 286 km/h Front around 6,7 mm, Rear 0 mm. - at 295 km/h Front around 4,1 mm, Rear at -2mm. Even with those settings the car is drivable, but understeers. The same was happening at Isi Portugal, that is why I thought that default setup could be improved for ride heights. If my findings do not correlate with what other see in their game, I would be very interested to know (allways desesperate to find an excuse to my poor driving!). Cheers.
This is misinformation. rFactor 2 is meant for both online and offline. It's not for nothing that is has the best AI of all the moddable platforms and things like resume from replay. Also Tim W is mainly an offline racer. Yes 1.8 is way closer to realistic and even still too forgiving. I don't think that these cars should be easier to drive and have more mechanical grip than the default Marussia. 1.7 had way too much grip and you had to deliberately drive like a madman to get the car to spin.
That's right, saying rF2 isn't meant for offline is just not true. Add a 'imo' to that statement, and maybe we can talk. But I guess the devs made an offline mode for a reason... I race mainly offline too. Best offline racing I've ever tried. So it is indeed as much meant for that as it is online.
Is not possible to make the 92 cars close to the Ferrari 91? The 91 cars doesn't have that rare behaviour at low speed that the 92 v1.8 have.
That picture is just the full win 1) Sparks 2) Tobacco Adverts 3) Standard Country Road (white center line) What a different time.
I think I had some success in getting more feeling and more mid corner front grip. There is much less FFB than in the V1.70. I imagine its due to the huge grip the V1.70 has and the resultant forces on the steering rack. In fact I had my FFB turned down for 1.70 to 88%. For this car I maxed caster (only 8.0 deg) and increased FFB to 124%. Also with suspension changes geared around increasing front grip: Low front slow bump, bit higher rebound. High rear damper on bump and low on rebound. ARBs and springs all set for more front grip. when camber was right the mid corner grip improved and I started to get a nice feeling. Diff lock way down too. My lap time at Sao Paulo is now 1:17.5 with 50 litres fuel. I think I was on about 1:14.5 with that fuel load in V1.70. So about 4 seconds but starting to enjoy the car. Still havent mastered tyre wear but its not horrific, the balance is what needs work. After a 16 lap run I think it was fronts @66% Rears say @80%. So will need to push grip forwards more and learn how to handle it.. Tyre change is 11 sec not 3!!!! Strange thing is that the tyre carcass isnt getting as hot as V1.70 where it would just keep getting hotter till say 150 deg at end of stint. This one stabilises at appr 100 deg. So it seems its wear not thermal deg thats eating the tyres. Oh and needed more packers as was grounding the rear at the end of the long straights.
Quick question.can l have both versions installed at the same time ? One small graphic glitch I've noticed is in the Tyrell the led. Gear indicator comes and goes. Not sure if its in game/ mod or my low end graphics card struggling
How's your FFB clipping at that setting? In 1.7 I mostly ran in the .8 to .85 range. I find it more pleasing right now but I did two runs on it. I found the tire degradation to have been improved overall and in 1.7 the thermal aspect didn't matter much to degradation ALTHOUGH it did build up way too much with anything. The tire could provide a similar result either completely new or with a block gone. After two blocks gone things really went up to sh*t and it got exponentially worse. I understand the setup alterations but I'd be a bit cautious of trying to "cure" everything with too many setup changes. Our range is rather limited and having to move things to a generally more different range isn't the right way to go. By that I mean, our range is from 1-10, if behavior is only improved at 7-10, we only have 4 values to select from instead of 11. Resting on the packers to help with grounding fits the same category, it helps but is a bit of a stop-gap cushion.
To GCCracer and surexxen. Thanks for the reply. I have installed your file surexxen but still have no real perf order in practice. I'll check on qualy and race tomorow For the rest, i've not learn how to add some stuff in mas files etc.. but i'm really intersted to learn that and why not in ai adjusments. I've found some good stuff on this forum already.
I just double checked all the new .veh files to make sure the class and driver names all match up and it all looks ok. Has anyone else had any issues? I've talked about various issues with AI cars holding each other up in other threads and recommended practice and qualifying sessions to be set to private for the most consistent results. Run a few qualifying sessions with the cars and let me know if you still have problems. I'll test more again later. Seems strange cause the 1.7 version worked fine and all I changed was the class names.
Ok so I did a few more tests and the results are rather interesting. Our example will be Malaysia with 24 cars .... 1.7 with F1ASR92V1 talent, AI laptime spread was about 3 seconds from 1st to 23rd (me being the 24th car sitting in the pits watching everything). Williams, McLarens, Ferrari, Benetton's all up the pointy end of the time sheet as you'd expect. 1.8 LE with ASROWC92 talent, AI laptime spread was about 1.7 seconds from 1st to 23rd. Williams, McLarens, Ferrari, Benetton's still in the top half of the sheet, but a fair few random mid fielders 'on a good day' breaking it up, a few of them further up than they would normally get due to the closer spread. With the baked in variance that you get from AI behaviour, the closer time-spread can cause the results to appear more varied than they should be. Overall, running the test several times and for several circuits the 1.8 LE is definitely reading the new talent files ... but something has definitely changed how the AI are handling with this mod. I did notice the AI struggling in certain corners (turn 6 Malaysia they seem to be having real issues mid-corner which might be related to the longitudinal slip problems we have mentioned ... I'm not sure). Since the talent files speed/qualifying speed is already 100/32 I don't think I can really 'spread them out' any further than they already have been without causing some other problem. AI limiter is set to 0, so it's got nothing to do with that.
I'm getting fairly realistic results with the talent file, surrexen. A few surprises here and there, most likely due to all the variables that have been mentioned already. But you've done a good job on this so far.
Yes it is a stop gap solution, but used a lot in racing cars in real life. As this car has no third spring to manage high aero loads it is the only way to avoid fitting an ultra stiff suspension or very high ride heights. Cheers.
Like Surrexen says, the closer the AI gets to each other in lap time, the more you'll see "mixed" qualifying results. It's weird though, given that this version is overall slower, the spread should have increased. One thing I did for myself - feeling bad to discard a bit of ASR work - is to remove the three lowest scoring teams from the mod. 100/32 = 3 points difference per driver, 100/26 is 4 points difference per driver. Brabham, Fondmetal and Andrea Moda rarely showed up in an actual race, and the grid was limited to 26 cars... dropping them gives a bit more space to the AI.
