Hello all,
I'm Mika Takala, the main firmware developer for the Granite Devices Simucube products.
There have been some reports, that the FFB in City Car Driving is indeed less than ideal, and I've just tested this with our internal development boards that output raw data as the device itself sees it.
Here is my report:
The game creates a DirectInput ConstantForce effect and some other effects, that have no relation to funny happenings. Nothing interesting/wrong happens while the game is loading.
When the world finishes loading, the ConstantForce effect's Magnitude parameters is immediately turned to -10000 (maximum negative value) by the game, and this causes the wheel to turn right at full FFB which can be dangerous with these high torque Direct Drive wheels.
If user presses ESC and then "Start again", the same thing does not happen, and something close to believable FFB starts to be produced
I can repeat this all the time with the latest Steam version of City Car Driving.
In other news, the CCD seems to be using so called effect envelopes for Square wave and Triangle wave effects. This makes CCD the first title that uses these canned FFB effects - congratulations on making FFB implementation extra hard on the device-side
In another news, the CCD seems to send new values AND restart effect on each FFB loop iteration for some of the effects. This is certainly not required by the DirectInput specification, and might cause issues in high FFB detail race driving simulators, but certainly does not cause issue here. We know a simple fix to this for C++/DirectInput SDK side - just leave out the restart flag from the SDK call bitmasks if restarting effect is not strictly required.
I hope that we can get into contact with the game developers via this forum.
Mika Takala
Granite Devices