A couple more small tweaks going in this week.
- Another very slight increase in spring strength.
- A slight tweak to bump damping.
- A ride height consideration for bottoming out.
- Caster revision to assist wheel users.
- A bump in front camber, a drop in rear.
Aside from some tuning notes for off-road vehicles I’d say I’m about wrapped up with anything but minor tweaks.
However, there are some things I’m still tracking.....
Lightweight cars, sub 2000lb.
Caterham, Abarth, etc. and Heavy weights, 4000lbs+, and if I need a new weight transfer spring tweak.
I spent some time tuning up a Mini Cooper S yesterday. I've always tried to set the cars stiff enough that the springs telemetry isn't hitting the top or bottom, but I think I'm giving up on that idea, as long as the car isn't thudding onto the ground or losing control suddenly. The car can feel good and not do anything weird even when the telemetry shows it running out of spring travel.
I haven't had a chance to run the f40 tune yet but so far the others have been pretty good.
The Silvia FE can set some really strong times on a bunch of tracks I think I'm somewhere around 11th in the world at COTA West. It's not the most populated leaderboard but at 11k players it's not too bad for this game.
Like the Silvia the Mk1 Golf also feels really locked down. Funny thing about it is that I think the camber numbers on my tune were essentially the opposite of the calc tune.
For me both cars felt a little too locked down? They both took a "set" in corners and weren't very nimbel mid corner. I'm not sure that the calc tune was ultimately faster. I think the Silvia calc tune might be a tenth or two faster than mine but it is probably more consistent. My Mk1 tune feels quite a bit faster (I'd have to look at times to get an accurate comparison) it's much nicer to change direction mid corner, a trait that I feel is important.
I think your notes about tweaking camber will probably go a long way towards fixing some of that feeling. I think my cars are generally running softer springs, more ride height, and higher front camber numbers.
Overall it feels like a big improvement over previous calcs.
Front camber I’m going to default up to -2.0 from -1.8
It CAN be effective with even more, but simplicity and trade offs.
Rear camber is going get dropped from -1.8 to -1.6,..... mostly for grip in higher powered classes, but it won’t affect the lower classes enough to matter.
Caster on the low end is bumped up to 5.6 minimum, again for wheel users mostly, and it won’t affect controller users in any noticeable way.
Mid-corner feel......agreed.
I’ve dropped the overall bump and ARB just a hair to avoid pitch based understeer, and as a result the rebound values will also slightly drop.
The end result should simply undamp the chassis and bring in a little more action without any loss of balance.
I spent a good hour playing around with the Caterham last night,.... using the above alterations. I’ve improved its feel somewhat,.... but it’s still simply overpowered.
A lost cause in my book.
I was doing some caster value testing for my calculator and kept running into the inconsistency of high dynamic camber between the left and right values in the games telemetry.
For example: Take any typical street car in Forza, set zero camber, zero toe, and max caster(7). Go into telemetry and observe the dynamic camber differences when steering at full lock. One side will be only a couple of degrees of camber, the other side can exceed as much as 10 degrees. These values should be much closer.
Why the high difference?
Part of the explanation is that Forza models an additional steering geometry known as Steering Axis Inclination, or SAI.
What is SAI?
Caster is the angle at which the steering axis leans from front-to-rear.
SAI is the angle at which the steering axis leans towards the centerline/inboard left and right.
Since these two angles are on the same axis these angles are technically a single compound angle, but the effects of them can be defined independently.
Especially if you use a spreadsheet calculation to do so.
What this calculator does is allow an understanding of how Forza determines the SAI for a given vehicle.
We know all the other angles and settings, so can then solve for the unknown variable.
How does SAI affect tuning?
The caster inclination angle works in opposition from left to right, adding dynamic camber to one side, and subtracting it from the other.
SAI angle works the same for either side, regardless of turn direction. Essentially adding positive dynamic camber to both sides.
Even simpler.....
Caster can add grip based on body roll.
SAI neutralizes that difference of dynamic camber.
What does it all mean?
Short answer. Well, not much. In a basic sense we still just test drive the car, and judge by feel and response how useful our caster setting will be,.... typically not far from 5~6.
Long answer......
Since SAI is an unknown variable, and does seem to vary between vehicles, the only way to measure its values is via an observation and ctesting of these values in a calculator as shown.
However, some initial testing seems to indicate that the more sporty a car is the lower the SAI value may be.
A typical street car even a Corvette, Mustang, or 911 will have a higher SAI value of 5~6.
A hyper car like a Veyron, FXX, or even a V8 Supercar might have an SAI range of 4~5.
Dedicated racing cars like an F1, P2, or Formula Ford could have lower SAI values of 2~3.
I’m still testing in this area.
In the end, it all comes down to optimizing potential feel.
Even more useless observations......
The interesting result of this practically useless exercise in attention to detail is that SAI turns out to be different for different vehicles.
This is realistic, but surprising for an otherwise generic suspension model.
These results may also reveal a possibility that while every vehicle is visually modeled having about 35-40 degrees of steering angle, not all vehicles necessarily meet that limit.
Another side effect of determining SAI reveals that the displayed dynamic camber values are the angular results of the left and right steering angle combined as one.
This is outright wrong,.... but we already know that the Forza telemetry data is questionable at best, so it’s not a surprise.
However this error does make some sense if the Forza physics model uses a bicycle style, two-point traction model.
Most, if not all, of the individual tire data parameters are measured individually for the visual and telemetry data, but for the core grip physics it would make computation sense to use a more efficient two-point traction model.
Tire temps, wear, pressure, and even camber values would all be local calculations, derived mostly from the core physical traction model.
Comments
265hp
2863lbs
235/35r19 front
255/30r19 rear
51% bias
MK1 golf
147hp
2076lbs
185/55/r15 f&R
60% bias
A couple more small tweaks going in this week.
- Another very slight increase in spring strength.
- A slight tweak to bump damping.
- A ride height consideration for bottoming out.
- Caster revision to assist wheel users.
- A bump in front camber, a drop in rear.
Aside from some tuning notes for off-road vehicles I’d say I’m about wrapped up with anything but minor tweaks.
However, there are some things I’m still tracking.....
Lightweight cars, sub 2000lb.
Caterham, Abarth, etc. and Heavy weights, 4000lbs+, and if I need a new weight transfer spring tweak.
The Silvia FE can set some really strong times on a bunch of tracks I think I'm somewhere around 11th in the world at COTA West. It's not the most populated leaderboard but at 11k players it's not too bad for this game.
Like the Silvia the Mk1 Golf also feels really locked down. Funny thing about it is that I think the camber numbers on my tune were essentially the opposite of the calc tune.
For me both cars felt a little too locked down? They both took a "set" in corners and weren't very nimbel mid corner. I'm not sure that the calc tune was ultimately faster. I think the Silvia calc tune might be a tenth or two faster than mine but it is probably more consistent. My Mk1 tune feels quite a bit faster (I'd have to look at times to get an accurate comparison) it's much nicer to change direction mid corner, a trait that I feel is important.
I think your notes about tweaking camber will probably go a long way towards fixing some of that feeling. I think my cars are generally running softer springs, more ride height, and higher front camber numbers.
Overall it feels like a big improvement over previous calcs.
Front camber I’m going to default up to -2.0 from -1.8
It CAN be effective with even more, but simplicity and trade offs.
Rear camber is going get dropped from -1.8 to -1.6,..... mostly for grip in higher powered classes, but it won’t affect the lower classes enough to matter.
Caster on the low end is bumped up to 5.6 minimum, again for wheel users mostly, and it won’t affect controller users in any noticeable way.
Mid-corner feel......agreed.
I’ve dropped the overall bump and ARB just a hair to avoid pitch based understeer, and as a result the rebound values will also slightly drop.
The end result should simply undamp the chassis and bring in a little more action without any loss of balance.
I spent a good hour playing around with the Caterham last night,.... using the above alterations. I’ve improved its feel somewhat,.... but it’s still simply overpowered.
A lost cause in my book.
Distribution? 45%
RWD
Tire widths? 320/320
Aero downforces? 524/750
Tuner tweaks and SAI -
I was doing some caster value testing for my calculator and kept running into the inconsistency of high dynamic camber between the left and right values in the games telemetry.
For example: Take any typical street car in Forza, set zero camber, zero toe, and max caster(7). Go into telemetry and observe the dynamic camber differences when steering at full lock. One side will be only a couple of degrees of camber, the other side can exceed as much as 10 degrees. These values should be much closer.
Why the high difference?
Part of the explanation is that Forza models an additional steering geometry known as Steering Axis Inclination, or SAI.
What is SAI?
Caster is the angle at which the steering axis leans from front-to-rear.
SAI is the angle at which the steering axis leans towards the centerline/inboard left and right.
Since these two angles are on the same axis these angles are technically a single compound angle, but the effects of them can be defined independently.
Especially if you use a spreadsheet calculation to do so.
What this calculator does is allow an understanding of how Forza determines the SAI for a given vehicle.
We know all the other angles and settings, so can then solve for the unknown variable.
How does SAI affect tuning?
The caster inclination angle works in opposition from left to right, adding dynamic camber to one side, and subtracting it from the other.
SAI angle works the same for either side, regardless of turn direction. Essentially adding positive dynamic camber to both sides.
Even simpler.....
Caster can add grip based on body roll.
SAI neutralizes that difference of dynamic camber.
What does it all mean?
Short answer. Well, not much. In a basic sense we still just test drive the car, and judge by feel and response how useful our caster setting will be,.... typically not far from 5~6.
Long answer......
Since SAI is an unknown variable, and does seem to vary between vehicles, the only way to measure its values is via an observation and ctesting of these values in a calculator as shown.
However, some initial testing seems to indicate that the more sporty a car is the lower the SAI value may be.
A typical street car even a Corvette, Mustang, or 911 will have a higher SAI value of 5~6.
A hyper car like a Veyron, FXX, or even a V8 Supercar might have an SAI range of 4~5.
Dedicated racing cars like an F1, P2, or Formula Ford could have lower SAI values of 2~3.
I’m still testing in this area.
In the end, it all comes down to optimizing potential feel.
Even more useless observations......
The interesting result of this practically useless exercise in attention to detail is that SAI turns out to be different for different vehicles.
This is realistic, but surprising for an otherwise generic suspension model.
These results may also reveal a possibility that while every vehicle is visually modeled having about 35-40 degrees of steering angle, not all vehicles necessarily meet that limit.
Another side effect of determining SAI reveals that the displayed dynamic camber values are the angular results of the left and right steering angle combined as one.
This is outright wrong,.... but we already know that the Forza telemetry data is questionable at best, so it’s not a surprise.
However this error does make some sense if the Forza physics model uses a bicycle style, two-point traction model.
Most, if not all, of the individual tire data parameters are measured individually for the visual and telemetry data, but for the core grip physics it would make computation sense to use a more efficient two-point traction model.
Tire temps, wear, pressure, and even camber values would all be local calculations, derived mostly from the core physical traction model.
TBC.....