Nice Marc. Remember this though, these results of the initial ffb deadzone are deadzones from a static steering wheel start position which means these are the deadzones for static friction. Static friction in higher than rolling friction, so the actual initial ffb deadzone for rolling friction will be less than for the static friction which is what your wheelcheck graph results are showing. On my t500 it was 8% but when/if you use stm in rf2 you don't go higher than 4-5% because when you add just a little movement to the wheel you take it out of the static friction and into the rolling friction and now the amount of force needed to keep turning the wheel by itself is reduced (because rolling friction is less than static friction). Now as you can see from your results, using 110% instead of 100% reduces the initial ffb deadzone 75% and 120% reduces it 100% down to no initial deadzone. But whilst you've found a value to fix the initial ffb deadzone in the profiler alone, you've caused another problem in the process…..ffb non-linearity. Just look at (for 110% overall effects strength curve) from 0% - 22% force sent to the wheel correlates to 0% to 60% force output from the wheel output. The remaining 22%-100% force sent to the wheel resulting in the remaining 60-100% force output at the wheel. It's quite non linear indeed. Now even 100% is not ideally linear as well but at least it regains some more linearity than 110%. Now if you can find a value between these two (maybe 106-107%?) that still gives you a very minimal deadzone from the profiler, then you can add less stm from rf2 and combined you can achieve the best of both worlds. This is what i do with my t500 (using 70% overall effect strength where the default is 60%) and add only 3% stm instead of the 4-5% i would have to otherwise. Alternatively, if you don't wish to use stm, perhaps you could try using the 110-112% in the profiler and add some negative ffb non linearity in the controller.ini file. The curves for all the overall effects strength on your g27 are characteristically positive, i.e. like this: So you want to add a layer of opposite non linearity by choosing a less than "1.0" value in the controller.ini file that applies a layer of negative non linearity like this: and hopefully you can come to some good values that'll equalise each other out to produce a relative more linear ffb response curve. I still believe in using the stm though but if you have non linearity issues with your wheel's ffb response curve regardless of the overall effects strength % you use, then perhaps it's worth a look at attacking the problem from the reverse non linearity solution.