Camber Theory

That applies to part a), yes. But if you want to dismiss parts b) and c) too it means that you also consider LFS to be just as bad and I've yet to hear somebody calling its physics model ludicrous. Or is it the usual double standards again - a "renowned simulator" doing thing X is perfectly acceptable but when PD does the same thing it's massively flawed?

To put things in perspective, there are cars in LFS that need their camber values to be set from 0.0° to +1.0° to actually run -2.0° or even more.
As I said, I wouldn't put anything past PD. It could be anything from broken to backwards to having hidden built in measurements to anything, even , "oops we forgot to program the camber, don't worry, we'll fix it in an update someday, carry on..". :lol:

I used to wonder why PD didn't release more detailed info on car tuning and adjustments and settled for vague innuendo and double meanings but the answer is obvious. They don't want to avoid as much as possible getting nailed down by putting into writing something that turns out not to be true. As it is the little info they do reveal on tuning is often incorrect at best or as in the case of camber, doesn't appear to work at all.
 
I'd like to note that I have actually improved a laptime using camber.

My Scirocco was consistently about a second faster around Tsukuba with 1.0 rear camber. The car had noticeably better turn in, due to the reduced grip in the rear. It's still modeled incorrectly, but I just found it funny that a slight grip loss can make a FWD car corner faster.

Somebody else should try this and tell me I'm not crazy.
 
It's still modeled incorrectly, but I just found it funny that a slight grip loss can make a FWD car corner faster.
These Volvos have multi-link rear suspension.
55696vol.jpg


volvo-race.jpg
 
Last edited:
Has anyone looked at the gallery view of their car and seen what is displayed when absolute max or absolute min camber values are applied?

Curious about that, I would try by my HT receiver is having a firmware update and has knocked out my video for about 30 minutes. I have tested this with ride height and that is working (at least being displayed) properly.

Finally, after the last "Hot (ninja edit) Fix" I was just able to get the Team Oreca Audi R8 GT3 car around Bathurst.
Trust me, I had to work spring rate and dampers until I was actually able to turn the car without the rear end trying to beat the front of the car to the corner. Camber was set at zero. I also had to add 40kg and slide the ballast all the way to the front of the car. It's still a pig, just not as fat. Although it weighs more.
 
You know, I DID notice something today.
Not the expected effect of camber, but the side effect, DOES work.

What am I talking about? Glad you asked, lol.
Now, take this into effect, I'm just one person, and it is possible my tests are biased seeing that I'm on controller, and I'm only a 'pretty good' driver, not an expert.
Anyway, think of GT5.
You adjusted camber according to two things, well, I did anyway, and that's what tire you were using, and your cornering performance. (Trial and error)
But no matter what numbers you had, if you ran your front camber higher then your rear, you got more response in oversteer, it was slight yes, but it was present. And vice versa, of you ran your rear higher, you became more planted.
This was true wether you ran .5 and 0, 2.0/1.0, or extremes of 3.3/1.2. Didn't matter the number, only mattered which was higher.
In my tests, this holds true.
I'm on 4 different cars now (two FR's and two MR's) and running 0/0 cam versus 0.5/0.0.
And unless I'm crazy, (high possibility) I see the slight add to exit oversteer with the second setting.

Do what you want with this info. If my tests were just my inconsistency in driving, then I was right in the first place, camber doesn't do enough to care about yet. But if it's not just me, this second theory stated above is something to think about. I'm using it. It's what I see, so I'm going with it.
Please test for yourself.
Don't even do full laps. It's not the point. If your worried about lap times, go change something else, lol. Just hit the turns, and watch to see if this can HELP the other settings.
I think everyone is putting to much into expectations about camber, wanting it to perform exactly like ether real life, or GT5, well, that's not happening, so get over it and move on, lol.
But like I said, I see it slightly helping my damper settings and LSD with the perfect amount of under/over steer.
 
Posted this on the other thread but it should go here aswell:


Bought a stock civic type r touring car.

Without looking at the settings I went to tsukuba and did 17 laps with a best of 1.00.222.


I then looked at the settings. 3.5 front and 2.0 rear.

Set to 0 all round and ran a 0.59.521 on my warm up lap. I have paused it to post this.


Loving the diagrams etc.


Edit :


Lap 2: 0.58.788
Lap 5: 0.58.624


3 4 6 and 7 were scruffy but all still under a minute. I was on for a 58.5xx but ran wide at the Dunlop right hander.
 
Lap 2: 0.58.788
Lap 5: 0.58.624


3 4 6 and 7 were scruffy but all still under a minute. I was on for a 58.5xx but ran wide at the Dunlop right hander.

And I'll wager that if you put a touch of camber back on the car, you'll go a little faster. By a touch I mean less than 0.5...On the civic I'm guessing 0.4 on the front.

I'm finding 0.3-0.4 on the rear or the drive wheels is "better". I can't say the car is any faster in terms of corner speed, it's just better behaved.
 
Last edited:
And I'll wager that if you put a touch of camber back on the car, you'll go a little faster. By a touch I mean less than 0.5...On the civic I'm guessing 0.4 on the front.

I'm finding 0.3-0.4 on the rear or the drive wheels is "better". I can't say the car is any faster in terms of corner speed, it's just better behaved.
I tried 0.5 all round on the civic, did get in front of my ghost in places, but not a single lap after 10 did I go faster than the 0.0 test.

I will try small I increments on either end later. (the nordshliefe and a certain amuse s2000 is calling me)

I think the fact that PD put 3.5 on the front, and that 0.0 is clearly much faster, (along with all the other testing) is pretty conclusive evidence that camber isn't working as PD intended, and therefore I think we will see a patch at some point, rather than it kind of working, and them leaving it.
 
I think the fact that PD put 3.5 on the front, and that 0.0 is clearly much faster, (along with all the other testing) is pretty conclusive evidence that camber isn't working as PD intended, and therefore I think we will see a patch at some point, rather than it kind of working, and them leaving it.

Well, that's a complicated situation. Working "as intended" by the community and "as intended" by the coding team are two different animals. The cars are often set up to feel proper rather than being fast. Given that there is quite a few new elements to the physics this time around, the camber may be working "as intended" within the code, but the results are not palatable to the community. That may call for a hack to the system.

It could even be that the way it works now is a fix for some other problem, like frame rate or something.

It could totally be a straight up bug though.
 
My point about the car having 3.5 stock was that, if I was coding a game, and designed a touring car model of a FF performance car, I would probably set all the standard settings somewhere near what a real world version had.

I would expect to see 3.5 on the front of a real world touring car.

I would expect the optimum camber angle to be anywhere between 1.5 and 3.5 depending on tyres/track/other settings.

I would expect a serious drop in grip with 0.0 camber. Not off a cliff. But noticeable in high compression situations.

Right now with the civic. 0.0 gives me the grip I would expect with a nice camber setup. If the sweet spot was between 0.5 and 4.0, with 0.0 giving the grip that 3.0 currently gives, I would be more than happy.

I do agree with you though.


We know nothing.
 
I'm finding 0.3-0.4 on the rear or the drive wheels is "better". I can't say the car is any faster in terms of corner speed, it's just better behaved.

This is exactly what I was saying.
You guys looking for better lap times aren't going to find them.
Should you? Well, yeah, if camber worked like GT5 did, or real life, but it's not right now.
What it IS doing, is slightly complementing or disturbing (depending on the camber and situation) the job the LSD and dampers are doing with corner in/corner out.
It isn't strong enough to add time, but it does add slight stability, IF you find the number you need. If you add some when you shouldn't, it will do the opposite, and make it handle worse, making the LSD and dampers less effective. That WILL add to your lap times, as many of you have seen.

Your all over thinking this.
This isn't brain surgery, lol.
This is simply a part of the program that either wasn't finished being written but the deadline to produce the game came and they went with it thinking they would Finnish it later, or they tried something new and it is simply a complete fail.
I really really believe it is the first part of that theory.
 
I just found it funny that a slight grip loss can make a FWD car corner faster.

That's just how you set up a FWD car, I drive an SVT focus and the factory alignment specs are -1.8* front/-2.3* rear. If you want an FF to be fast through the corners you need some extra camber in the back. Don't think of it as reducing grip though, true there's less grip at a neutral stance and that helps you swing the back into the corner but once you're under side load that tire gets to do plenty of work. Have to be careful with liftoff oversteer however.
 
I have a new theory, based on the idea that the information is being misinterpreted by the code.

So, when you adjust the toe angle, there is a negative and a positive value. You can slide either way.

With camber, the values are all positive. The negative is only in the UI. It reads "Camber Angle (-)". The visual is converting these values to a negative value, as evidenced by the visual change. However, I think it is quite likely that the values are not converted to negative values in the code.

When I drive a car with a high camber value, it definitely wants to turn more (twitchy), but not stick more. Isn't that consistent with positive camber? Isn't everything that we have found consistent with positive camber...which would be exactly what the UI is telling us?
 
That's not exactly a new theory but you gave the best description of it that I've seen so far.

I'm not really sure about the cornering effects of positive camber in real life, because the only places I've heard of it being used are in drifting and drag racing (because they want the rear wheels to lean to 0 as the suspension compresses) and normal cornering forces don't really apply for either of those sports.
 
I have a new theory, based on the idea that the information is being misinterpreted by the code.

So, when you adjust the toe angle, there is a negative and a positive value. You can slide either way.

With camber, the values are all positive. The negative is only in the UI. It reads "Camber Angle (-)". The visual is converting these values to a negative value, as evidenced by the visual change. However, I think it is quite likely that the values are not converted to negative values in the code.

When I drive a car with a high camber value, it definitely wants to turn more (twitchy), but not stick more. Isn't that consistent with positive camber? Isn't everything that we have found consistent with positive camber...which would be exactly what the UI is telling us?
It's entirely possible but we'll never know for sure. One day there will be a patch if we're lucky and all it will say is something like, "Fixed some issues with camber on some cars" and we'll have to retune everything in our garage!!:banghead::banghead::ouch:
 
By "inside" I mean the inside of the turn. So, if you're turning right, the inside (right) wheel will tilt away from the car, achieving more positive camber.

This car is technically turning left but the wheels are turning right and that's what I'm referring to.
4151313210_a1ce5e42ea_o.jpg
I read your first post and this one and I'd like to clarify something about camber usage in drift cars. I think you're touching on it here but not stating it explicitly.

Obviously we see a lot of drift cars, amateur or professional, running a high degree of negative camber on the front wheels. What you can't see is that they're also running higher than usage negative caster - the steering axis of the wheels is tilted further backward, creating more straight-line stability and causing the steering to return to center more quickly. Obviously this helps during quick transitions where we witness the driver let go of the steering wheel as it spin furiously in the opposite direction.

The high negative camber angle of some drift cars actually compliments in the high caster angle and improves cornering grip in that application. You've stated correctly that the front wheels of a drift car mid-corner are technically steering the opposite direction. Because of this, the wheel on the outside of the corner (inside of the steering direction) will camber in a positive direction by default, and that's not good. Here's an example of a stock car's suspension doing this:

atl3_1b3I_1.jpg


To counteract this, they run a high negative camber angle so that as it becomes more positive it actually flattens the contact patch and never becomes truly positive, unlike the photo you've posted. Combined with this effect, the outside wheel is loaded so its camber angle will change as a result of that as well, but this change depends on the type of suspension geometry.

The car in the picture you posted is running too little negative camber to flatten the contact patch at high drift angles. It doesn't necessarily need to as it's a driver preference thing. But for the sake of the geometry argument, here's an example of a car running high camber (even on the rear for dem style points) which flattens the front contact patch during a drift:

atl1_KYBs_1.jpg


The car you posted in Matt Powers's who runs in FD. Here's a picture of his newest car with its wheels straight:

227942_10150588891485543_602145542_18500550_5563470_n.jpg


Based on my own car which has -2 degrees of camber front and rear, I'd guess his car has more than that, around -4 degrees, possibly more. In fact, here's a picture of a car which the owner stated has -4 degrees front and -1.5 degrees rear:



Now, with that said, this alignment of the front wheels is useful for drifting but not for circuit driving because of the way the suspension geometry moves in this style of driving. I think you realize this based on what you've already said, but I wanted to clarify because some GT fans seem to think that higher camber settings should always be better. In reality, and with my own experimentation on my own car (1991 RX7, coilovers, 235/40-17 NT05, -2 camber all around, very slight toe-out front, very slight toe-in rear) -2 degrees seems to work well for most street-legal performance tires. My experience is based on small circuit, autocross, and street driving among myself and friends with various other performance tires on differing suspension types. Testing the contact patch during cornering is fairly simple in real life but obviously the game doesn't give us that luxury.

I'm new to this argument that GT6's camber system is broken. From what I've read, it seems effective, but at lower angles than what it should be. What I haven't seen is whether or not the game factors in differing suspension geometries. Camber needs will cary from car to car - McPherson struts tend to go negative under slight compression, but then go positive past a certain point; double-wishbone tend to increase negative throughout the suspension stroke; multi-link suspension tend to act similarly to double-wishbone; live-axles don't have camber adjustment at all (thus the interesting NASCAR alignment terms which don't exist in GT); torsion-beam axles move in different ways as well. All the other various alignment angles change through suspension stroke also - most angles are not adjustable in real life and aren't mentioned in the game at all.

I'm just wondering if GT players take into account the various changes which take place. I'm also wondering if the game takes into account the various changes. Frankly, I don't think it accurately models individual suspension geometries because when you slam a Miata the front and rear double-wishbone suspension causes extreme negative camber. I've done it. This doesn't happen in the game as far as I can tell. In fact, on a Miata with stock suspension and performance tires you shouldn't need to adjust camber at all because the geometry creates its own negative camber as it compresses in a corner.
 
not going to give a list of details as it will just add confusion , but i tested 0.0 to -1.0 in 10 increases (-0.1 each time) and after 20 laps of each increase up ive come to my own conclusion that anything over -0.3 is killing times and my fastest laps was at -0.2
thats my findings after 200 laps =)
great post PsuPepperoni
 
I got a car that runs faster with camber than without, so at least some of the time some cars will benefit from camber.
 
Evidence? Specific vehicles? Settings? Lap times? [/anecdote]
Nissan Fairlady Z 280z Tune in my garage (link in my sig). Faster with specified camber values than with lower values.

GT40 tune in my garage. Insignificant difference between specified camber values than with 0 / 0. If 0 camber was clearly better, I would have used it. No perceivable negative effect from specified camber values, in fact, I have a preference for them, that's why I used them.
 
testing results and details ? curious to what you got going on , as ive just tested the same car and had the same results as all the other cars ive tested
0.0 being my 10 fastest laps and time lost over every increase of camber :S
 
Really? Would you care to provide this setup, with your camber? I would be very interested to see this.
 
Nissan Fairlady Z 280z Tune in my garage (link in my sig). Faster with specified camber values than with lower values.

GT40 tune in my garage. Insignificant difference between specified camber values than with 0 / 0. If 0 camber was clearly better, I would have used it. No perceivable negative effect from specified camber values, in fact, I have a preference for them, that's why I used them.
Let me guess, you put a little bit of camber on the front and now oversteer isn't as bad?
 
testing results and details ? curious to what you got going on , as ive just tested the same car and had the same results as all the other cars ive tested
0.0 being my 10 fastest laps and time lost over every increase of camber :S
What same car? GT40 or 280z?
Really? Would you care to provide this setup, with your camber? I would be very interested to see this.
Provided. Tunes in my garage. Link in my sig.
Let me guess, you put a little bit of camber on the front and now oversteer isn't as bad?
No. Tunes provided for you to actually try, rather than making useless comments if you like.
 
In GT5, running 2.0/0.0 would add some entry rotation, or turn-in, but it seemed to hinder exit rotation.
Opposite, 0.0/2.0 would add exit rotation, but hinder turn-in/entry rotation.

In GT6 I started out using this method, and it seemed to work. Except that 0.0 simply has more grip. So now, my take on GT6, is that it's a mix, it works as per GT5, except mid-corner rotation is much more significantly affected, and therefore it's largely masked and unusable.
 
Could you say "on brakes/on power" instead of "entry/exit"? It's a bit less confusing, especially since the latter is basically a location, and the exit of one corner can be the entry to another corner.
 
Could you say "on brakes/on power" instead of "entry/exit"? It's a bit less confusing, especially since the latter is basically a location, and the exit of one corner can be the entry to another corner.

I have always heard different terminology. Every driving book and tuning book that I have read refers to the corner segments as locations; corner entry, mid-corner (or apex) and corner exit. Getting back to the throttle can happen at any one of those points in the corner. I'm not sure how to teach a new driver where to get back on throttle without some sort of "location" reference. For example, "start applying throttle just before the apex" is very different than "start applying throttle just past the braking zone at the throttle application point."
 
Back