Gek54 said:
You may want to test it again. The game will increase the steering angle initially in an effort to counter steer for you and then it will reduce the steering angle after having corrected the slide as well as having scrubbed some speed, the opposite of what you are suggesting it does. Point the wheel in the general counter steering direction and the game does the rest of the work for you.
OK, so I've tested it again, and again, and again..
I am convinced that there is no 'automatic' counter-steering.. there are a couple of issues that could be looked at, but none look like deliberate counter-steering
Here's some of the many myriad of tests I've done.
The conditions are (set in My Profile->Controller)
1. Set the Wheel to have inside/outside deadzones to their endstops
2. Reduce FFB/Effects to zero
3. telemetry up in game showing wheel position
Test 1 - Shows weird wheel wobble effect
1. With the car stationary, put a small amount of lock on, and set off very slowly, invariably you can get the right steering angle and hold a speed that show 'wheel wobble' where it oscillates between two positions, this is with no sliding/power oversteer and can be held indefinitely. You can get it to wobble at a slow speed and at a fast speed. Note that it actually wobbles 'about' the steering angle you are holding, i.e. it flips above/below where it should, averaging out at the correct point.
Conclusion:-
There is some 'resolution'/'aliasing' issue in one of their algorithms that causes the steering output to flip quickly between two values, and it's not a function of the car's rotational velocity.
In the R class cars it shows up even worse, and if you just hold a gentle lock through a corner it can also be made to do this.
This is shown in the telemetry output, and visually affects the wheels (in a replay), but th interior wheel (and your own wheel) show no sign of movement.
Test 2 - Shows steering range reduction
In a large space, set off and hold full lock, you will see that as the car goes between 0 and 70kph, the max lock reduces, you need to slowly build speed to see it smoothly, then back off and see the lock increase again.
Test 3 - Small glitch as car rotates 180 degrees
In a large space, and a powerful RWD card, get upto some speed (100-200Kph), turn full lock and get it to power oversteer right around, you will see your lock is initially at it's reduced state (as you are at speed), but when the car goes through 180 degrees, it suddenly jumps to full lock.
Conclusion:-
There is a small glitch in the logic somewhere, crossing over 180 degrees does reset the steering range to max, but the reality is it's not important, spinning 180 degrees isn't something you are that concerned with.. be nice to be fixed though.
Test 4 - the side to side sliding wobble
As per the video shown in this thread, pick a RWD car that can easily slide side to side to show the effect.
- You need to do this under power, and at at least 70kph
The result is as per the video, as you cross centre on each slide to start counter-steering, the wheel oscillates between 2 positions before finally giving you a smooth steering angle.
This isn't ideal, granted, but crucially this is the source of the complaint.
The thing I do note is that it is oscillating between two points quite quickly which is not actually that helpful, if it where some clever assist it would hold more lock for a period of time, not oscillate between slightly more lock and slightly less lock then your steering input is asking for.
Conclusion:-
It's hard to be conclusive, but crucially, the fact it oscillates between more/less lock then you are asking does not seem that helpful as an assist. And oddly it seems related to the slow speed (non-sliding) wheel wobble, where the steering osciallates between two values..
I'm going with glitch, and some further tests that isolate the speed and sliding factor out show it's not always present..
Test 5 - Wheel wobble not there on a single slide
To do this, I went at speed in a straight line, then used the handbrake to induce a good slide to the left (reasonable lock used to start the slide, handbrake released, then countersteer with half lock to the opposition direction, the slide is sufficient that it can't be recovered). While sliding, you can 'saw' the wheel to turn into/out of the slide and no wobble is shown.
Conclusion:-
Whatever the wobble effect is, it does not exhibit itself if you do a single slide, and once sliding does not exhibit itself if subsequent turning into/outof the slide to re-initiate countersteering is performed.
Test 6 - Slow speed sliding
I simply did a donut on the spot in a powerful RWD car, using no more then about half lock in each direction, and only switching between the two lock positions.
Once the car was sliding nicely in one direction, I'd counter steer in the opposite direction until the car slowed and started spinning in the other direction where I'd then counter steer in the opposite direction again.
Conclusion - At slow speed, no wobble occurs (or very very infreqeuently, I never saw it), the turning through centre into the counter steer position was smooth and nothing untoward occurs.
So after a load more messing around and seeing what happens, trying some different class cars etc I can see some issues, that would be nice to be fixed, but actually observing the reaction of the car to it all, it seems to not have any noticeable effect.
The bottom line is, the steering does 'oscillate' between two positions at times, but this can be at a walking pace when not sliding.. I believe the same mechanism that causes it at walking speed is also causing it when sliding side to side in a drift style (the time it also shows itself the most).
You can through the use of the handbrake induce a slide, and counter-steer with no wobble shown at all, and also when sliding in one direction you can switch between steering into the skid, and counter-steering with no wobble.
You can also spin from one direction to another at a slow speed, and again initial counter-steering does not invoke any wobble..
I'd like a comment from Che/T10 on this, In my experience it looks like a side effect of calculation that is oscillating between two values rather then anything deliberate.
