[REL] F1 2013 Single Player v1.61 (DX 11 compatible, visors, rain update)

Sorry mate. I won't do that. I only change physics and talent files. You may give Sompir, the orginal mod maker, a try.

 

I could not reach them either. So I first asked on RD for permission and it was up to me. I would make an update with only my changed files, if only the packaging tool let me do that. After two weeks of trying, I gave up and decided to make a whole mod. But I have to say, the mod is 2.8 GB, whereas my changed files could be zipped and transfered by floppy disk...

 

First, thank you very much for your feedback, really appreciate that. For all that have a red quoted mod in the mod Manager, saying "- damaged mod -". This is normal. In both versions I do not add showrooms, but I think the problem is due to not having assigned a single track to the mod. In game I have not a single problem. I've seen another formule 1 mod without tracks having the same "damaged mod" problem, but that one also works in game with zero problems.

I do not use Motec, but will use it maybe (sounds interesting, does it work with 64 bit build?). Maybe that difference between in- and output is also a reason for the relatively slow AI. Basically, I want the AI now at 120%, driving the same at 100%. ATM I have no clue why there is a change in values, but am really interested in someone who knows the answer.

The order of teams is: 1. Red Bull, 2/3. Mercedes and Ferrari (about +0.3secs at RB), 4. Lotus (+0.7secs at RB), 5. McLaren (+1sec at RB), 6. Group of FI, TR, Sauber, Williams (about 1.4secs at RB) 7. Caterham (+2.5secs at RB), 8. Marussia (+3 secs at RB). The differences are lower at low downforce tracks. Red Bull also carries a lot of drag, but comes with the highest downforce, especially from diffuser/undertray. Marussia have very low drag, so it may have a change against Caterham at Monza.

 

You're welcome and thank you for your feedback. The damaged mod problem is only an aesthetic problem, nothing to worry about. It is due to a mod not having assigned a single track. I think I cannot help you man, because I only do Notepad editing of physics and talents. You may ask Sompir, the original mod maker, for the graphical issues.

However, I'm interessted in the part "the mod has too much grip", because that are physics. Do you think it is not realistic? And so, in what kind/direction? Downforce? Front or rear? Or in general? Or the tire grip is too high? Or the center of gravity too low? It is my purpose to make this mod, from a physics point of view, as realistic as possible. I want this mod as the most realistic f1 mod ever available for us "normal" people (people who don't come in the Red Bull simulator or so).

I will be on holiday from tomorrow till the end of July. So don't worry if I do not answer that quickly.

 

Hello Frenky,

"For all that have a red quoted mod in the mod Manager, saying "- damaged mod -". This is normal. In both versions I do not add showrooms, but I think the problem is due to not having assigned a single track to the mod. In game I have not a single problem"

Sorry to disagree, the problem has nothing to do with not having tracks assigned (I have done this many times without causing any error, and for the RFRT2013 mod, where I had changed also the talent file some time ago).

I just noticed when comparing the original mod and yours that you didn't have any reference to a showroom, and I made a new package with your files and an existing showroom. I don't have any error in my mod manager for any installed mod.

Yes Motec works fine for me with 64 bits (Thanks Lazza), just some lag sometimes to create the log file after going back to garage.

I wonder how you were able to test your modificactions without using Motec (may be an alternative tool?), for me it's impossible to tune an F1 without data on dynamic ride height which is the key to have the diffusor working adequately (with your standard setup, after several tries when I roughly dialed my ride height my lap times went down 2 s).

Off line racing is much more enjoyable now.

Thanks for your work I'll go on testing, I did try myself to quicken AI but I didn't achieve such a result (I used to tune RF1 mods, and I have found a lot of good information for this on Nogripracing.com).

Bye.

 

Oke, strange problem then. I don't have that problem with the tracks version of the mod. And in both versions I do not use any showrooms. When it doesn't give other problems right now, I will include the showrooms in both versions in my next update, that will come after my holiday.

On to your testing question. Well, thank you very much for introducing Motec. I will give that a try. What I did right now is "trial and error" (which cost more than a month...) based on information I gather from the web. Assuming that rFactor2 uses the right physical equations, the car should react like you would suppose. And yes, in the version 1.17 I've completely redesigned the diffuser equations, based on information f.e. http://www.f1technical.net/forum/viewtopic.php?f=6&t=15480. The total amount of downforce by the undertray doesn't differ that much from v1.15, but especially when the rear ride height is low (fast corners on flat circuits, Silverstone/Copse f.e.), the diffuser should give less downforce now. I may, however, even radicalize this even more in a next version. Bottoming the car, now loses 40% of the undertray downforce (was only 11%). So, Copse should still be flat with low fuel load, allthough it might be more tricky now. Of course, I use Excel for all the wings and diffuser equations, so I can graphically see and compare with the older equations. And last, but not least, I compare the physics of the physics of other rFactor1 and 2 F1 mods from 2005 onwards. Another example is that the radiator and brake duct settings should influence the car much more now both in terms of drag and negative downforce (lift), thanks to the CTDP mods and by watching onboard movies and timelists of the Malaysia GP.

Thanks for pointing to Nogripracing. I know that site. I also use a site that has all the equations written out. It's for rFactor1, but most holds also for rF2. Right now, I gave the AI about 5% extra tire grip. I can give them even more, but do not like this, because I only give them fake/non real grip. Any suggestions about improving the AI are very welcome.

Have fun racing!

 

Well that's difficult to answer and really who am I to say.. Let me clarify my statement. On a green track on med/hard tyres pushing qualy laps yeah there can be some tricky moments. Pacing at 95% is rather easy. Please don't take me the wrong way. I have allot of fun driving this mod and the work is greatly appreciated.

Here is a small bit of data. Entering Maggotts the Lotus F1 caries 36% more rear down force the the ISI Marussia F1 and roughly the same for the front. Its not comparing apples with apples (year & car) and who's to say ISI got it right. Who's to say they didn't. Does that seem excessive to you? I don't know. If I did I'd be working at the German GP this weekend

I also logged a couple of 6G moments. That seems excessive. The Marussia never went over 4.7G. Not that it was a great lap.

Ill do some slip angle analysis tomorrow.

 

Wow, interesting! And this is all due to Motec, I guess? I wish, I had known that before. You can, however, compare apples to apples (to a certain extent) by driving the Marussia in my mod (iso Lotus) and drive both cars with full wings. I'm then very interested how the figures really are, assuming that the 2012 car is in reality about half a second to 1 second slower. The Marussia is by far lacking the most downforce, although their drag levels are not that bad.
You are talking about 36% more rear downforce, but at the moment I do not know if that is only the rear wing, or also (part of) the diffuser downforce. I do know that the Lotus car has a pretty good "diffuser downforce", but lacks some front wing downforce compared to Ferrari. In my honest opinion, I would roughly say that the Lotus, driving full wings at about 300 kph, can carry about a couple of 100's kilo's extra TOTAL downforce, compared to Marussia, which would be in the range 20-30%. It's the best guess I can make, having heard once that about 150 kg of downforce at full wings @300 kph (roughly Maggotts) could give you about a second.

 

Thanks Frenkky for your response.

Up to the the moment I didn't really try to setup F1s in RF2 because the tyre model was not definitive, and a lot of Motec channels were not reported (as damper velocity, or rear tire temperature).

I liked very much the RFT mod which reassured me of what was possible to do graphically on RF2 (to the moment not rivaled in my view, I don't know if they use the standard shaders or their owns) but the fact that all cars were identical was detracting a lot from off line racing.

Due to the investment it takes to master the tuning of an F1 car I was waiting that well known moders produce a sound mode before playing with them (I prefer GT and multi class racing like Endurance series in RF1).

I am delighted you made a serious attempt at physics, I made worksheets with diffuser data and ride height to find optimal dynamic ride height for the cars I wanted to tune in RF1.

I have an overdose of Silverstone (last time I tried but did not succeed in tuning an Isi Master, the rear suspension is very different to all other suspensions I had practised and after a while I realized I was totally erring - softening rear anti-roll and I was on three wheels on slow curves...) and I'll wait sometime before returning to this track.

I'll join a Motec sheet with the ride height displayed with your standard setup (same wings, just adjusting ride height and packers, and dampers). It's just a guesstimate of optimal ride height (13,8 FL, 16,5 RL at 222 km/h in last curve before main straight). Never went much above 4G (30-62 Wing).

Have a look at the motec thread by Lazza. I just added a worksheet which combines all ride heights on the same graph (helps to see how both suspension react to increased downforce).

https://www.dropbox.com/s/03u79hku303hbyl/Motec Portugal Frenkky.doc

At 120 % In Portugal GP Vettel runs in low 1mn 09s which is way beyond my reach as a simracer. With default setups for AI, and the talent file I had made (when I packaged again the mod I took mine and not yours by error).

Excellent drivers who are still to fast for the AI, can choose a Marussia and race against the RB and Ferraris.

The grip is very subjective I know a simracer who regularly won championships playing with keyboard against people using G25 and Fanatecs, I am sure he would find your car to easy to drive (not an urban legend).

People who want very difficult cars to drive can choose low downforce setups, or drive a Marussia to train their skills.

I would wait and obtain a lot of opinions based on facts you can check in person before changing your physics (you can't expect people to test seriously an ever changing set).

Congratulations, your physics are a very clear improvement for off line racing, and this way the RFT mod is very enjoyable to many kind of simracers adjusting the difficulty level.

It is now my preferred open wheel mod by far and is as pleasant to race offline (AI of version 770 is very good in my opinion) as the best GT mods available.

Bye.

P.S. The error of the packaging of the mod is maybe unrelated to showroom (the one included with RFT is nice).
It happened to me with other mods, and all I had to do was to repackage the mod as is, and the error was gone.
Are you using the last version of the packaging tool?

 

Well, I'm really delighted to hear that it is your preferred open wheel mod and want to thank you for your feedback and your data collecting. I've just downloaded the MoTeC stuff and have to admit that I feel like, ehmm, yeah kind of reborn on another planet. A whole new world this. It makes me feel stupid for about a month, for not using this before.
On to your reply. I've downloaded and seen your MoTeC data and think it looks quite oke to me. Guessing that it is a Red Bull, the optimum rear ride is higher than you guessed. It's somewhere between 20 and 25 mm. Below those numbers, the downforce coëfficiënt drops (total downforce still increases then, because you are still accelerating the car /increasing speed if the rear ride height becomes lower) till about 15 mm. It is then about 85-90% of the maximum downforce coëfficiënt. What happens between 0 and 15 mm of rear ride hight, I'm not quite sure. At this height, stalling is starting to take place or may take place. But I don't know exactly how that is calculated in rF2. You may suddenly loose downforce, or not. From my diffuser equations the downforce coëfficiënt decreases even further in that region. Below 0 mm, the car is bottoming and you loose 40% of the available downforce by the diffuser/undertray.

Yes, I only make changes based on facts, like timetables and raw data and will do that also in the future. Good to hear that the AI is driving circles around you, but I want to give nearly everybody a good challenge. And the slider doesn't go higher than 120% (for some, that is still too slow ). I myself only drive the 19 tracks of 2013, so not Portugal and do the testing mainly on Silverstone (high downforce) and Monza (low downforce). I'm also driving on keyboard (yes unbelieveble, because I also own a Trust Force Feedback GT racing wheel), which cost a couple of kilograms, due to the "Aidpenalties".

Have fun driving!

BTW: I use build 772 of rFactor2 (64 bit exe) and the packaging tool version is "2.9.2.5". Really strange that something went wrong then, as there are no problems ingame. I also once had the F1L2013 v1.3 mod (well actually it is a multi-component) installed, which gave me the same red "damaged package" error and according to their forum it was due to not assigning a track/tracks. Anyway, I also never had problems with that mod. But yeah, you know, I didn't quite like their mod. The sounds were great, but graphics and especially phyisics not really to my liking.

 

If you want to increase Ai speed, if my memory is good you can increase grip bias in tbc files as you probably did and there was an ai downforce bias variable in hdv (in Rf1). I don't think it is artificial as in any case Ai doesn't use the same physics as the player's car, at least in Rf1.
Bye.

 

Yes I did that in the TBC files. I also changed the "AIDownforceZArm=0.050" value (offset from CofG where the AI has their aero pressure point) to bigger numbers in a former version of my mod, without resulting in a faster AI. I didn't change the "AIDownforceBias=0.0" value, as I don't understand its effect ingame, as I don't want to change things I do not fully understand and as I think it is better to keep it as close as their setup should be. Do you know what the downforcebias means and what it does ingame? So what is the result if I put that value towards "1.0"?
Have fun racing!

 

That is the explanation I found on Nogripracing, it was in a GTR2 discussion (I think GTR2 uses ISI model), I have never changed it in the RF1 mods I adjusted (I played with AI grip in tbc and the variable AIPerfUsage):

"This is a percentage between the AI setup downforce (0.0) and scripted downforce calculated by the game (1.0). If you're not using AI specific Setups, you should use 1.0. In certain sort of tracks as Le Mans with very long straights, to have a good AI Setup might be the way to go (more details further). But if you're lazy to make a setup for each car, you'll rather set this value = 1.0 and let the computer calculates the correct downforce. But it will not give proper gear ratios to AI. Some mods as GRID Prototypes has absurdly wrong default gears ratios and cars are unable to go over 200 km/h. In those cases AIDownforceBias will not cure the issue. I'll cover how to fix gear ratios later."

In RF1, I used to comment this variable, to allow the values in PLR for brake power usage, brake grip usage, corner grip usage and the global difficulty setting to be the only places where AI speed was defined:
//AIPerfUsage=(1, 0.989, 0.981) // PLR file overrides for (brake power usage, brake grip usage, corner grip usage) used by AI to estimate performance - only positive values will override, see PLR for defaults

In your RB hdv I see you changed these values to lower (slower) values than the original values of the RFT mod I have - (1, 0.989, 0.981) - to
AIPerfUsage=(0.995, 0.96, 0.98) // PLR file overrides for (brake power usage, brake grip usage, corner grip usage) used by AI to estimate performance - only positive values will override, see PLR for defaults.

When I tried to improve the AI speed of this mod, this were the two changes I made to the standard hdv

AIDownforceBias=1.0 // was 0.0 Bias between setup and hard-coded value (0.0-1.0)
//AIPerfUsage=(1, 0.989, 0.981) // PLR file overrides for (brake power usage, brake grip usage, corner grip usage) used by AI to estimate performance - only positive values will override, see PLR for defaults

It worked but AI was no so fast as yours because I didn't touch the tbc file, and I didn't do the step of having a specific hdv for each car. It was a project I had planned for later. After many thoughts I even considered trying an other alternative widely used in motor racing, just play with weight handicaping (different cars having same physics but different weights).

Where did you find the formula to interpret the diffuser of the RFTcars, DiffuserBasePlus=(-0.985, -0.50,-0.80, 62) // Base lift and Half/1st/2nd order with rear ride height. What is the meaning of Half?

I had formulas for simpler expressions like DiffuserBase=(-1.0246, 2.12, 83.74) // Base lift and 1st/2nd order with rear ride height

Cheers and thanks.

 

Hello again and many thanks for your reply.
On to the diffuser equations, the formula -0.985-0.50x^1/2-0.80x+62x2 (where x is the rear ride height in meters and you should still substract a linear factor times your front ride height from it) I did not found it anywhere on the internet. It is this formula that I came up with, after a day of testing and using the information on f1technical. Especially the half order allows to (if you wish drastically) lower downforce values at low rear ride height setting. I stated in the changelog, that I may "radicalize" this even more in a future update. What I mean is that I will test, after my holiday, the diffuser equation: -0.92-0.88x^1/2-0.52x+72x2, thus "radicalizing" mainly the half order factor. I'm still not quite happy about the downforce levels at very high Speed (>270 kph), but didn't want to make an update with a complete different feeling. Just put those equations in Excel and see the difference and let me know what you think. You know, I'm a racing driver, but not a formula one Driver, so I really want to know what other people think of this aspect.
Regarding the difference between "DiffuserBase" and "DiffuserBasePlus", I think it is this half order equation. I guess, you cannot use DiffuserBasePlus in rF1 (???). As you inmediately see with your formula -1.0246+2.12x+83.74x2, the downforce coefficient is the highest when the rear ride height is 0 mm (apart from the "possible" stall effect). In my honest opinion, that is just not realistic if you read the info on f1technical.

Onto the downforce bias. I may also test this. I think, I improved already the default setup of the AI, mainly by lengthening the gearing. I will not use a value of 1.0, but may try 0.5 as a maximum. Thanks for pointing me to that and the info. If you have the possibility to test it, please let me know your results.

I'm quite sure your values of AIPerfUsage will make the AI a little bit quicker. But, on the other hand, I found those values for braking too high. The AI has already better braking power than humans and by making the brake grip usage so high, will make the AI crash more into your back. However, I will slightly increase the corner grip usage value to 0.99 in my next update, although I doubt if that will bring more than 0.2secs. It still seems to me that the AI is losing too much time in high speed corners, slow corners seems to be oke now. So mainly a "downforce problem" for the AI (?), although the default setup already gives good levels of downforce.

I will not give certain cars extra weight penalties, unless someone tells me that those cars had those penalties in real life too .

Have fun racing!

 

Thanks for your explanation of DiffuserBasePlus.

"My formula" is from ISI BMW formula 2007 or 2008 I don't remember, complete it looks like

DiffuserBase=(-1.0246, 2.12, 83.74) // Base lift and 1st/2nd order with rear ride height
DiffuserFrontHeight=(1.450) // 1st order with front ride height
DiffuserRake=(-0.002, -20, 440.0) // Optimum rake (rear - front ride height), 1st order with current difference from opt, 2nd order
DiffuserLimits=(0.013, 0.105, 0.044) // Min ride height before stalling begins (0.0=disabled), max rear ride height for computations, max rake difference for computations
DiffuserStall=(0.1, 0.56) // Function to compute stall ride height (0.0=minimum, 1.0=average), downforce lost when bottoming out (0.0=none, 1.0=complete stall)
DiffuserDraftLiftMult=1.40 // Effect of draft on diffuser's lift response
DiffuserSideways=(0.432) // Dropoff with yaw (0.0 = none, 1.0 = max)
DiffuserPeakYaw=(1.5, 1.03) // Angle of peak, multiplier at peak
DiffuserCenter=(0.0, 0.01, -1.284) //

So complete downforce is affected by FrontHeight, and optimum rake and stalls when to low from ground. It is the formulas I chose to standardize my F1 mod (2006-2013 in RF1). I wanted to use the same setups for all modern mods (purists will cry, but some of the mods I had recollected had very weird physics).

I agree with you for the default values for braking, but in my case I had them already changed in the PLR. I have no problems of AI crashes with your mod.

I agree with you that for F1 a weight penalty is not very realistic (except in the case of heavy drivers like Webber or Hulkenberg), but it is widely used in GT and touring class. But as you are a driver, an amateur like me will not try to teach you.

For AI slowing in fast curves I fear it is track related (AIW) or due to the way AI estimates it speed. I remember that in the Suzuka track of the pack the cars slowed in the middle of the main straight, and were pathetic in the R120 curve that is why I erased the tracks.

In my experience with RF1 the best way to alter AI speed is to play with the AI grip bias of the tires in the TBC, but I am not a moder, just someone able to tune existing mods, maybe someone more able can help you here.

I did alter engine.ini files when I plotted the torque curve of some cars I wanted to add to an existing mod and saw they were extravagant. This is another way to pursue your goal of having the more realistic F1 possible, tuning the engines. I wrote a VB program to export those values in csv that I imported in a spreadsheet to plot the torque curves.

I don't know for sure for RF2 but I assure you that AI doesn't use the same physics than the car you drive at least in RF1.

Next week will mean no RF2 for me until fall.

Bye and thanks.

P.S. The Motec was from a Lotus, with default wing values. For the record I tried a higher ride height in the last curve before the main straight (following your advice for the diffuser), I had more grip in this curve (may be not really needed there) but I lost time on the slower curves of the Portugal GP and had a slower lap time.

 

As promised a small comparison test:

Track: Silverstone (Green)
Laps: 5
Vehicle: 2013 RFT RBR, 2013 RFT Marussia & 2012 ISI Marussia
Tyres: Medium compound
DRS: Disabled
Centre of Gravity: RFT 15.8cm & ISI 21.4cm
Setup: Base RFT setup which essentially is a copy of the ISI Marussia setup with a few minor modifications. 20L fuel, break bias & brake pressure changed with front and rear wings set at max. This setup applied too ISI Marussia F1 except for the 3rd spring stiffness as this is not possible however I used a similar ratio of stiffness between front and rear.

Notes:-

2013 RFT RBR: On rails. Point and shoot. I can be super aggressive without extreme slip angles. Can engage the throttle alot earlier than both Marussia's. IMO this would be great if I dialled in the the best setup in the world and was on super soft tyres on a rubbered in track however I wasn't.
2013 RFT Marussia: Good all round performance. I had to be on guard a little just in case.
2012 ISI Marussia: Massive amounts of understeer. Lacked top end speed despite carrying less drag. Average rear traction.

Lap Time Differential:-

2013 RFT RBR: +0
2013 RFT Marussia: +2.4 seconds
2012 ISI Marussia: +7 seconds

Max Front Downforce:-

2013 RFT RBR: 6807 Nm
2013 RFT Marussia: 5703 Nm
2012 ISI Marussia: 6176 Nm

Max Rear Downforce:-

2013 RFT RBR: 9643 Nm
2013 RFT Marussia: 8063 Nm
2012 ISI Marussia: 7286 Nm

Max Drag:-

2013 RFT RBR: 6158 Nm
2013 RFT Marussia: 6034 Nm
2012 ISI Marussia: 5893 Nm

Max Speed Km/h:-

2013 RFT RBR: 293 Km/h
2013 RFT Marussia: 294 Km/h
2012 ISI Marussia: 289 Km/h

Max Slip Angle at Abbey:-

2013 RFT RBR: 1.84 deg
2013 RFT Marussia: 3.9 deg
2012 ISI Marussia: 3.25 deg

Max Slip Angle at Copse:-

2013 RFT RBR: 2.28 deg
2013 RFT Marussia: 3.07 deg
2012 ISI Marussia: 3.3 deg

Max Slip Angle at Stowe:-

2013 RFT RBR: 4.1 deg
2013 RFT Marussia: 4.6 deg
2012 ISI Marussia: 7.6 deg

I haven't played with the setup yet. Can anyone get close to last years pole time? 1:29.6.


If you want my modified Motec template & additional maths channels p.m. me.

 

Thank you Lgel and Coanda, much appreciated!

@ Lgel: Yes, the AI can improve on some tracks, especially at Suzuka for example. But ATM I have no experience with track editing. Maybe in the future. Your last sentence is exactly how ride height should be. By making the rear ride height higher, you improved in that corner (because of a better downforce coëfficiënt), however, the car is then too high in the slow corners. And as you can win/lose more time in slow corners, your laptime increases.
@ Coanda: very interesting data you come up with man! What time did you drive in the RB car, somewhere low 1:30? I think I will slightly change the center of gravity of all cars. The rest of the data, yeah, what can I say. Sounds really oke to me, at least from a aerodynamic point of view. The Marussia 2012 must be a lot slower by less traction or so. Maybe slightly less rear wing downforce, although the difference car-to-car (yeah I know, different year) is less than 11%, I'm perfectly fine with that. Maybe slightly more front wing downforce per click (not base). But for all this, there may be more tests needed, for example on a low downforce track.

You guys take care, as I will go on holiday now. When I'm back at the end of the month, I will read everything and post the conclusions!

Have fun driving!

 

really interressant this study by Coanda; for myself, say, sound seem gretest than poor rft (horrible); just try on red bull racing ...thanks, the mod is really fine

 

The Lotus feels awesome. When driven on the edge this car doesn't feel glued to the track. It feels great!

I am however getting some really crazy frame rate spikes and dips when racing 21 AI at Silverstone; going from 60 to 30, up to 60 down to 25 and so on.
This is the only mod I'm currently having this problem with. I can usually race against 30-35 AI with no FPS problems at all.

 

Lap times of AI at 120 % in Silverstone between ISI Formula Master and the new RFT phyisics are coherent (slightly faster for RFT 2013 against ISI FM 2012).

Wouldn't that go against thinking the new RFT cars have to much downforce?

Unless you think that ISI Formula Master cars have to much downforce.

Even less if you factor that the grip bias for the AI in the new physics (.tbc) is 5% higher than in ISI.

Do you agree with my reasoning?

Thanks.