Xbox calibration data masked in Xidi-WinMM?

[rccraigb]

Using Windows 10's "Game Controllers" calibration function, I intentionally mis-calibrated several controllers in order to understand application behavior using Xidi-WinMM. Interestingly, the bad calibration (sticks only allowed to travel left or forward, leaving right and backward fully saturated) was completely ignored by Xidi-WinMM. That is, stick movement was 100% smooth, whereas tests without Xidi-WinMM reflected expected behavior - movements jumping wildly in the right and backward directions.

Removing the DLL and using the default OS library yielded testing results consistent with the bad calibration. Does Xidi-WinMM mask or ignore calibration data?

[samuelgr]

I'm guessing the "bad calibration" is being applied to XInput controllers specifically?

For XInput controllers, Xidi gets controller data directly from XInput. I imagine XInput might ignore the calibration data, which is why Xidi-WinMM also appears to ignore it.

[rccraigb]

Yes, apologies for not specifying in the original question: the devices are all various flavors of Xbox controllers (360, One, Razer), so yes XInput.

Thank you, by the way, for Xidi and your support of it: very much appreciated!

[80T]

Is it not a thing that XInput controllers are reporting 2 different sets of controls to the OS, like XI and DI independently.. ?
For example exposing two axes (one per each Trigger) in the XInput set, but one combined axis as both Triggers in the DInput set ?
Therefore, the axes calibration for the DInput controls set (which I assume is what we're seeing in the Gamepad CPL) is independent from XI ?