Newbie questions on design choice for VR sim

Hi guys. I'm thinking about taking the plunge and building something, I'm just not sure where to start or which direction to go.

This use will be mostly for racing (PCARS) and space sims (ED and someday SC) with some sort of VR HMD (Rift, Vive or OSVR) in the future.

I have purchased the plans from BFFSimulation as I was thinking of going with a 3DOF platform but I have heard there may be an issue with head movement from the motion sim, and the IMU of the HMD interpreting this as intentional head movement so I was looking at something like a G-Seat or Geko GS-105 style as an alternative design to a 3DOF platform or conventional 2DOF seat mover. Would this type of sim help or would there still be too much head movement?

It appears the Gseats are more complicated and are a 4DOF system? Is this type of axis mixing readily available in software or would it require programming knowledge?

The Gecko system seems to be a 2DOF but limits the motion to the seat back. Could this be build the same way as a standard 2DOF seat mover with the seat bottom stationary and only the back moving? Could this be on a simple U-joint or would it require some sort of linkage system to limit it's movements (lateral and heave as opposed to pitch and roll)?

I would like to keep costs down as a first time sim project so was thinking of Arduino/monstermoto(s)/12v wiper motors. Is this unrealistic?

It seems the more I read the less I know and more confused I get. Please set me straight!

Thanks.

I could suggest basic 2 Dof balanced pointvseat mover. Larger 200w 12 DC motors , JRKS ( easy to set up) . Should set you back about $700 ish. Also , I have read , because of the immersion of VR and ive only tried them without motion, you would not need massive thrashing motion. Just precise movement. There would be other VR guys here thar coukd advise though....

I believe the IMU and sensor mismatch can be accounted for by mounting the sensor on the rig. This is a temp solution until motion cancellation is common in HMDs API, but it can work.

There is a topic in the q&a section about VR rigs, check out the posts there. Right now the software for haptics and motion is still in beta, but consider transducers also. It's still open to what is a great setup for VR.

I use a DK2 on a seat mover set up and have the camera mounted on the wall in front of me not attached to the frame.
I don't have any violent movements like lots of shaking on bumpy roads as this does make the view on the rift bad.
Also you really don't need much movement in the rig when in vr.
Having the camera pick up your body moving and translating that to the sim/game you are in(I'm mainly iracing) I find to be actually a good thing, it feels more realistic as even if you were stapped into a real car your head would move around under strong g forces. If the camera were attached to the rig it would not feel right as your view would not move at all when you brake or turn which is not realistic.
That's just my experience though.

I'm very interested to see where this goes. As someone looking to build a 6 dof platform with a vive preorder, I can't wait!

As @noorbeast said, VR has already been successfully done on 2 to 6DOF. Have a look through the Members project thread. Don't know about the Vive. Also have a look at this thread http://www.xsimulator.net/community/threads/building-a-rig-for-vr.7899/#post-97162