GPL: automatic linear/non-linear adjustments?

Here is an easy solution for GPL steering difficulties for Papyrus to think about: When you are on a straight, the steering automatically goes to non-linear and when you brake and enter a corner, the steering automatically goes to linear. You get the best of both worlds a controlled ride on the straights and precision steering in the corners. But then again, maybe that is too complicated to do. Any opinions?

> Here is an easy solution for GPL steering difficulties for Papyrus > to think about: To any of the world's problems, there is clear, concise answer whose only fault is that it's wrong. :-) I think would be fairly frustrating to get used to. Bunch of end cases to deal with: what about when the car is on a short straight between two curves in opposite direction? Is this a time for linear or non-linear steering? Same for a short curve between two straights. I would expect that I could set the car up such that a constant steering angle (with the wheel) would cause the car to describe a constant arc through the turn. When it didn't, I'd be surprised, and when I couldn't figure out why, frustrated, and when I finally *did* figure out why, I'd be cursing! :-) What about a short "straight" between two curves in the same direction (thinking of a 1/3-1/2 mile oval here). The car would "feel" wrong as you set the car up, and exited turn 1 on the throttle with a fixed steering input. Overall, IMO this is a worse solution than a "reduce with speed", and if that's what you really want, the above proposal is: a. more difficult to engineer, balance and playtest. b. might be exactly what you asked for, but not what you want. ---Jim

Jim, Could this be implemented as a "reduce linearity with speed" option? I don't think it would feel as unreal that way... Any thoughts? A. Renault

Antoine Renault wrote in > Definitely, that way you could get the ease of going in a straight line without losing the maximum lock to get yourself out of trouble. Steve

Quoted text (11 lines)

> Here is an easy solution for GPL steering difficulties for Papyrus to think > about: > When you are on a straight, the steering automatically goes to non-linear > and when you brake and enter a corner, the steering automatically goes to > linear. > You get the best of both worlds a controlled ride on the straights and > precision steering in the corners. > But then again, maybe that is too complicated to do. Any opinions?

How about train tracks? Look Mum, no hands!

Quoted text (10 lines)

>Antoine Renault wrote in > >>Jim, >>Could this be implemented as a "reduce linearity with speed" option? >>I don't think it would feel as unreal that way... Any thoughts? >>A. Renault >Definitely, that way you could get the ease of going in a straight line >without losing the maximum lock to get yourself out of trouble.

I'm not sure I'm interpreting you correctly, but you're not losing maximum lock with non-linear steering. 100% steering wheel deflection still translates to maximum wheel lock, same as with linear steering. It's just around the center that the steering wheel behaves less sensitive. -- Wolfgang Preiss \ E-mail copies of replies to this posting are welcome. Uni des Saarlands \ and U.S. law. You have been warned.

Quoted text (13 lines)

> > Here is an easy solution for GPL steering difficulties for Papyrus to think > > about: > > When you are on a straight, the steering automatically goes to non-linear > > and when you brake and enter a corner, the steering automatically goes to > > linear. > > You get the best of both worlds a controlled ride on the straights and > > precision steering in the corners. > > But then again, maybe that is too complicated to do. Any opinions? > How about train tracks? Look Mum, no hands!

....Not to mention traction control , active suspension , ALB , Ground effects venturis , ect.....all standard fare on a 1967 GP car right !!!! ????? =-)But seriously.... Ummm. I think the solution lies with the DRIVER. Since I have been driving GPL I have had to develope " touch " and " smooth input ". From what I understand these cars were a handfull even in a straight line ( no wings ). I've also noticed that this newfound driving style has translated to much better lap times in ALL of my other sims as well. Steve at RXspeed7 racing Visit the RXspeed7 page at... http://www.geocities.com/MotorCity/Garage/7177/index.html

% I'm not sure I'm interpreting you correctly, but you're not losing % maximum lock with non-linear steering. 100% steering wheel deflection % still translates to maximum wheel lock, same as with linear steering. % It's just around the center that the steering wheel behaves less % sensitive. This is one of the areas where I noticed a marked difference between linear and non-linear. As the exiting of a corner is where one needs to get on the power, I find that the wheel takes alot longer to straighten out coming out of the turns (in the non-linear setting) and thus one can not accelerate as strongly with the wheels still turned. With the non-linear settings the wheel does not straighten as fast or as "smoothly" with a more linear setting. -- **************************** Michael E. Carver ************************* Upside out, or inside down...False alarm the only game in town. =-=-=-=-=-=-=-=-=-=-=-=-=-=-=<[ /./. [- < ]>=-=-=-=-=-=-=-=-=-=-=-=-=-=

Besides that fact that the steering is twitchy with a full linear setting on the straights, I find that alot of this comes from not being smooth with the accelerator. Due to the power of the Lotus, one really needs to pay attention to easing on or off the accelerator. The other factor that comes into play is you must pay extra attention to the car's balance when coming out of a turn and onto the straight. If I have the car properly balanced I don't find it too difficult to keep the car straight. However, if I have the car just slightly off balance as I come onto the straights, it is a handful to keep the car in a straight line. This is one of the reasons I enjoy the physics of GPL, it isn't just a point and squirt simulation, it's a total package. -- **************************** Michael E. Carver ************************* Upside out, or inside down...False alarm the only game in town. =-=-=-=-=-=-=-=-=-=-=-=-=-=-=<[ /./. [- < ]>=-=-=-=-=-=-=-=-=-=-=-=-=-=

You're right, you're not interpreting me correctly :) I meant that having a reduce maximum wheel lock with speed would be bad as you couldn't get yourself out of trouble if it became necessary, but having reduce linearity with speed option would help a lot of the people struggling to get going in a straight line, without reducing the total lock for if it was needed. (Personally I have no trouble going in a straight line, but maybe if I was using a joystick I would?) I'm not sure if that came out any better than before, but I tried :-) Steve

>I'm not sure I'm interpreting you correctly, but you're not losing >maximum lock with non-linear steering. 100% steering wheel deflection >still translates to maximum wheel lock, same as with linear steering. >It's just around the center that the steering wheel behaves less >sensitive. What I think is that some people think of a "reduce lock with speed" while others think (myself included) of a "recude linearity with speed". There is some confusion because I and others only use "reduce with speed" when we're talking about it. Now a "reduce lock with speed" option would really limit wheel lock while you're at full speed, and that's not acceptable IMHO. A. Renault

I agree. I think too much emphasis is placed on modifying control setups. Moving from sim to sim inveitability requires some degree of learning the feel of the controls, and you need to invest the time to get to grips with it. For me the GPL demo has reinforced how much lap times come down with practice rather than constantly changing setups or controller settings. With just one slider to adjust I quickly settled to one I felt provided adequate control for corners and straights and left it there. After practice the times have steadily come down (still 2 secs off Mr Mansell though). I know I couldn't have resisted changing the setups or controller settings if they had been available, because I wanted to blame the ill-handling car - not my driving ability. Now I'm sure with changes to the setup I could further reduce my times, but I'm convinced I wouldn't have invested enough time exploring driving technique compared to fixing it in the garage. The GPL demo has changed my approach to car setups to put a lot more time in to driving than tweaking. Cheers Tony

> I think too much emphasis is placed on modifying control setups. Moving > from sim to sim inveitability requires some degree of learning the feel of > the controls, and you need to invest the time to get to grips with it. Exactly! I look at different sims as being the same as jumping into a different car in real life. Ditto! This is exactly what some others need to do. It's hard enough to get good at GPL without constantly changing the linearity and thus the feel of the whole thing. Yes! The good thing about hotlapping with the GPL demo is that you know everyone is on the same playing field (cheating aside but that seems non existant at this point) and you can compare yourself very well to other drivers rather than wondering if your/their setup is inferior/superior. Great to see a post I agree with 100% :))

Which is I think what GP2 does and it makes places like the tunnel at monaco trickier than they need to be due to the lack of lock. Reduce linearity with speed would help out anyone not drivng with a very high quality wheel, as not everyone cen afford these. Steve

Maybe I'm misunderstanding your use of the word "fast". Could you clarify exactly what you mean here? You seem to be saying that the steering angle of the wheels is somehow disconnected from the position that the steering wheel controller is at, and that the wheels don't come back to center as quickly, even though the steering wheel is centered. Are you saying that there is a lag between the control input and the steering movement of the front wheels on a non-linear setting? Or, are you just saying that it takes longer to physically move the wheel (because of the less sensitive steering on non-linear)? Thanks, -- Pat Dotson IMPACT Motorsports

% > This is one of the areas where I noticed a marked difference between % > linear and non-linear. As the exiting of a corner is where one needs to % > get on the power, I find that the wheel takes alot longer to straighten % > out coming out of the turns (in the non-linear setting) and thus one can % > not accelerate as strongly with the wheels still turned. With the % > non-linear settings the wheel does not straighten as fast or as % > "smoothly" with a more linear setting. % Maybe I'm misunderstanding your use of the word "fast". Could you % clarify % exactly what you mean here? It seems that there is a lag between the "angles" of the tire deflection in comparison to the wheel movement in the non-linear settings. One doesn't have the fine degree of movement. I notice this most coming out of the long turn. With a full non-linear setting I can't make the smooth straightening I exiting the turn. As I straighten the steering, there are points where the wheels are not responding. There fore I can't get on the throttle as soon and hard as I am with linear steering. Try just moving slowly down the track and turn the wheels. With full non-linear settings, there seems to be points while turning the steering wheel that the wheels don't respond. There seem to be numerous dead-zones in the turning radius. This is were the "not straightening as fast or as smoothly" comes in. Try this with full non-linear and then full linear to see the difference. % You seem to be saying that the steering angle of the wheels is somehow % disconnected from the position that the steering wheel controller is at, % and that the wheels don't come back to center as quickly, even though % the steering wheel is centered. Are you saying that there is a lag % between the control input and the steering movement of the front wheels % on a non-linear setting? Or, are you just saying that it takes longer % to physically move the wheel (because of the less sensitive steering on % non-linear)? I think I mean the former (lag between controller input and steering movement). The usual approach to cornering is to achieve maximum acceleration on the exits of the turns. I've found that a non-linear setting slows me down (both in the entrance and exit, but mostly in the exit). -- **************************** Michael E. Carver ************************* Upside out, or inside down...False alarm the only game in town. =-=-=-=-=-=-=-=-=-=-=-=-=-=-=<[ /./. [- < ]>=-=-=-=-=-=-=-=-=-=-=-=-=-=

Quoted text (14 lines)

> % You seem to be saying that the steering angle of the wheels is somehow > % disconnected from the position that the steering wheel controller is at, > % and that the wheels don't come back to center as quickly, even though > % the steering wheel is centered. Are you saying that there is a lag > % between the control input and the steering movement of the front wheels > % on a non-linear setting? Or, are you just saying that it takes longer > % to physically move the wheel (because of the less sensitive steering on > % non-linear)? > I think I mean the former (lag between controller input and steering > movement). The usual approach to cornering is to achieve maximum > acceleration on the exits of the turns. I've found that a non-linear > setting slows me down (both in the entrance and exit, but mostly in > the exit).

Papyrus, Could someone who really knows please comment on what effect the non- linear steering setting actually has? Does it really introduce a lag in the steering response? Thanks, -- Pat Dotson IMPACT Motorsports

> Could someone who really knows please comment on what effect the non- > linear steering setting actually has? Does it really introduce a lag > in the steering response? I can assure you that the old (ICR1->N2) controller code did not introduce a time lag (low-pass filter really) between input and physics sim when used with analog inputs (the keyboard of course did have low pass filtering :-) ). I can presume that the new controller code is similar in effect. The guiding principle at Papyrus is to let the human drive the damn car, not concoct fancy-pants computer-assistance mechanisms to make it overly EASY, but to put the human in control... This, coupled with the knowledge that the old code didn't implies that new code probably doesn't. ---Jim

Quoted text (20 lines)

> > % You seem to be saying that the steering angle of the wheels is somehow > > % disconnected from the position that the steering wheel controller is at, > > % and that the wheels don't come back to center as quickly, even though > > % the steering wheel is centered. Are you saying that there is a lag > > % between the control input and the steering movement of the front wheels > > % on a non-linear setting? Or, are you just saying that it takes longer > > % to physically move the wheel (because of the less sensitive steering on > > % non-linear)? > > I think I mean the former (lag between controller input and steering > > movement). The usual approach to cornering is to achieve maximum > > acceleration on the exits of the turns. I've found that a non-linear > > setting slows me down (both in the entrance and exit, but mostly in > > the exit). > Papyrus, > Could someone who really knows please comment on what effect the non- > linear steering setting actually has? Does it really introduce a lag > in the steering response?

Not in my experiences. I use full non linear and a time I did within first 20 hours (Gee, I still don't think I've done 20 hours yet) still is good enough for top 10 at the Apex - http://gpl.gamestats.com/

Quoted text (11 lines)

> I can assure you that the old (ICR1->N2) controller code did not > introduce a time lag (low-pass filter really) between input and > physics sim when used with analog inputs (the keyboard of course did > have low pass filtering :-) ). > I can presume that the new controller code is similar in effect. The > guiding principle at Papyrus is to let the human drive the damn car, > not concoct fancy-pants computer-assistance mechanisms to make it > overly EASY, but to put the human in control... This, coupled with the > knowledge that the old code didn't implies that new code probably > doesn't.

That's good to hear. Non-linear steering is a good thing to use as long as the steering response is consistent, and I don't think it impacts the realism of the experience. I wouldn't really even call it a driving aid. Reduce with speed is more of a driver aid, and I've never liked it much because it makes the steering response inconsistent over the range of speed. That is, steering response is different at 90 mph than it is at 100 mph, and there is no good way of sensing that small of a speed difference in simulations. It's probably not a bad thing to include for people who want it though. -- Pat Dotson IMPACT Motorsports

% > Could someone who really knows please comment on what effect the non- % > linear steering setting actually has? Does it really introduce a lag % > in the steering response? Lag probably is not the right word. Lack of precision is more accurate (at least from my prespective). With linear, each movement of the wheel (or input device) translates to a similar movement of the steering mechanism on the car. From what I have observed with full non-linear, there are "stages" that the steering mechanism turns in response to the "angle" of the input device. This induces a "lag" when compared to a full linear setting. Of course this is based on my observations of the car's behavior on the track at slow speeds and not by observing the actual code of the program. % I can assure you that the old (ICR1->N2) controller code did not % introduce a time lag (low-pass filter really) between input and % physics sim when used with analog inputs (the keyboard of course did % have low pass filtering :-) ). % I can presume that the new controller code is similar in effect. The % guiding principle at Papyrus is to let the human drive the damn car, % not concoct fancy-pants computer-assistance mechanisms to make it % overly EASY, but to put the human in control... This, coupled with the % knowledge that the old code didn't implies that new code probably % doesn't. % ---Jim -- **************************** Michael E. Carver ************************* Upside out, or inside down...False alarm the only game in town. =-=-=-=-=-=-=-=-=-=-=-=-=-=-=<[ /./. [- < ]>=-=-=-=-=-=-=-=-=-=-=-=-=-=

> That's good to hear. Non-linear steering is a good thing to use as > long as the steering response is consistent, and I don't think it > impacts the realism of the experience. I wouldn't really even call > it a driving aid. I agree [that it is not a driving aid], and I would imagine that most people who gave it a moment's thought would also agree. ---Jim

Quoted text (17 lines)

> > Could someone who really knows please comment on what effect the non- > > linear steering setting actually has? Does it really introduce a lag > > in the steering response? > I can assure you that the old (ICR1->N2) controller code did not > introduce a time lag (low-pass filter really) between input and > physics sim when used with analog inputs (the keyboard of course did > have low pass filtering :-) ). > I can presume that the new controller code is similar in effect. The > guiding principle at Papyrus is to let the human drive the damn car, > not concoct fancy-pants computer-assistance mechanisms to make it > overly EASY, but to put the human in control... This, coupled with the > knowledge that the old code didn't implies that new code probably > doesn't. > ---Jim

I believe the difference between linear and nonlinear steering can be described as follows: With linear steering, for every X degrees of steering you turn the wheel (around 16), you get 1 degree of angle at the wheel - pretty much a straight line graph. Whereas with non-linear steering, around zero the conversion starts at 16 degree/degree but changes monotonically as your steering angle change so that at around 600 degree of steering angle, the conversion might be 10 degreee/degree. At least this is how it works for a passenger car steering system. -- Delphi Automotive Systems GM Proving Grounds, Milford MI

Quoted text (11 lines)

> I believe the difference between linear and nonlinear steering can be > described as follows: With linear steering, for every X degrees of > steering you turn the wheel (around 16), you get 1 degree of angle at > the wheel - pretty much a straight line graph. Whereas with non-linear > steering, around zero the conversion starts at 16 degree/degree but > changes > monotonically as your steering angle change so that at around 600 degree > of > steering angle, the conversion might be 10 degreee/degree. At least this > is > how it works for a passenger car steering system.

I'd say it is simply that linear is just that ie linear - meaning that steering wheel deflection is directly proportional to front wheel angle. Non linear is an exponential, logarithmic or whatever type of thing. The end result is that NL is less sensitive/more accurate about centre and more sensitive/less accurate at the extremes.

I'd go along with that Byron. Tony

> I'd say it is simply that linear is just that ie linear - meaning > that steering wheel deflection is directly proportional to front wheel > angle. Non linear is an exponential, logarithmic or whatever type of > thing. The end result is that NL is less sensitive/more accurate about > centre and more sensitive/less accurate at the extremes. Playing the part of an engineer pedant, replace "accurate" with "precise" and I concur completely... ---Jim

Quoted text (8 lines)

> > I'd say it is simply that linear is just that ie linear - meaning > > that steering wheel deflection is directly proportional to front wheel > > angle. Non linear is an exponential, logarithmic or whatever type of > > thing. The end result is that NL is less sensitive/more accurate about > > centre and more sensitive/less accurate at the extremes. > Playing the part of an engineer pedant, replace "accurate" with > "precise" and I concur completely...

Well, I'd say they are one and the same in this case! How "accurate" it is depends on the driver! As far as "precise" goes, a bad driver will hit the same part of the same wall with monotonous regularity lap after lap with NL steering :))