That sinking feeling..

I never had a sinking issue with my rig. In fact the tracking had always been perfect. Just using a RAM mount with the sensor centred at the back of the steering wheel base, just above the top of the wheel. Since the top of the Oculus sensor can still tilt up or down, I just mushed a bunch of blu-tack up in that part of the sensor to stop it from moving. Seems to be effective.

But since the 1.9 update from Oculus the tracking has been terrible. It doesn't really sink, it just randomly snaps the head position up to the roof level, or to the bottom of the seat. I haven't been able to pin it to any particular movement, maybe it happens more often with the more violent movements. Completely unplayable. Pretty pissed off about it TBH.

I've tried all sorts of things, different USB ports, different USB PCIE cards, reinstalling drivers, enabling ASW, disabling ASW, toning down my motion profile. Nothing made any difference. I'd love to be able to roll back to previous runtimes, like 1.8 or 1.7, to at least confirm if the 1.9 update has caused the problem. Does anyone know if that's possible?

Failing that, suggestions on a gimbal setup?

Thanks. Have ordered one of these: http://www.ebay.com.au/itm/201711927857 will see how it goes.

Can anyone explain WHY this happens? I don't understand why it happens and can't make any sense out of the fact that the gimbal helps.

Don't get me wrong, I'm already bought the gimbal and going to use this fix but I really hate not understanding what the heck is going on.

If the camera is fixed to the platform the pilot is sitting on the relationship between them should be the same as if both were on solid ground. This just doesn't make sense!

Yea like @noorbeast said, but in testing heave does not seem be be effected at all, roll/pitch is the most sensitive, yaw will actually catch up but it does it in annoying steps. Quite easy to test yourself - just mount your tracking camera and headset to a common plane (a meter long piece of wood/cardboard tube will work fine) and move these two together in space while watching the headset copy on the computer screen - all the issues will become apparent. Note that I never had the sinking issue with DK2. FI the latest update seem more stable and while it has not stopped the rolling/pitching it does seem to has stopped the jump around and sinking - so maybe the need for a camera gimbal is now not worth the effort . I feel this will be a moving beast .

Yes I knew that you were using heave as an example, but FI try it yourself, heave is not affected at all. I can set the test rig up and move it up and down over a meter - the image does not shift/move at all. Why is heave not affected - don't know?

Thanks for the input guys! A question and also some clarification of my confusion:

Is Vive currently viable for motion sims via using the hand trackers?

As for my confusion: I'm aware of the gyro and the camera, I didn't know an accelerometer was still being used. The main reason I'm confused is that I thought pitch and roll were handled 100% by the gyro and heave, sway, and surge all handled by the camera. I guess I never gave yaw any thought and that is an interesting one. They must use an accelerometer for yaw still, then have the camera correct for error.

What I don't get is why the camera doesn't have the final say in correction. The Rift got some conflicting data... fine, so things get a little funky until the next camera update.... then it should put yaw and positional back where they belong instead of deciding the Camera is wrong?

Probably best I just keep quiet, but normally I have a pretty good understanding of how things like this work and this one is throwing me.

@SeatTime here is your Rift camera mounted that you are finding this problem mostly gone with the newest Rift patch?

My camera is (very temporarily) mounted about 45 degrees to my right on the front of my "armrest" and it's moving around a LOT even with the newest patch. I'm actually about to build my first version of my Rift camera mount and wondering what the best height for it is (as close to head height as possible?) and whether there are less drift problems if it is close to the HMD or far away from it. My current design puts the Rift camera three and a half feet in front of my head, excellent for tracking area but is it good or bad for the drift problems?

As we both said, Oculus has designed it so that the pitch and roll of the Headset systems have priority over the Camera tracking. Not great for what we want to do, but that's how it is . You wont stop the image rolling and pitching, what the original issue was that the image use to sink or just miss-track, which does not seem to be an issue with the latest patch. My camera is basically right in front of my headset - that works for me, but I don't think it really matters in regard to image movement - unfortunately all the rolling/pitching is due to the systems in the headset not the camera. I still keep the camera on the rig though, as it does remove heave, surge and sway movements. Until Oculus introduce some type of motion cancellation into the SDK (they currently have no plans to do this) and all simulator developers implement these functions, then it is what we are stuck with. My current solution - I'm building/installing a G-seat into my rig, so that I can reduce the required roll/pitch of my sim (still have plenty of surge/sway/heave though).

With the newest patch and the camera relocated (not on a gimbal currently) the problem seems to be gone as you guys said!

I know about the pitch/roll, and never had any ideas we'd be able to get rid of that, and I don't find it hardly noticeable anyway. It was the view drifting and jumping around that was the problem. I just hope it stays gone!

Can confirm, tracking is back to normal in V1.10.

What a relief, ugh. Oculus better not do that again, that made me sad for a while there. They broke my toy

Well the gimbal is still in the post, let's hope I don't need to use it on this.

Something changed since last update for sure, I'm starting feeling nausea quite soon. Never had before with DK2 and CV1.
I don't know if they changed something with the ASW implementation, let's hope it's not that, I wouldn't get rid of it...

I just play Assetto Corsa with the Rift. It happens in single player and with the AI.
With my rig I have ASW on only at races start for the first corners, otherwise I'm always at 90fps.
I was thinking that lately I'm using Post-Processing effects, I'll try turning them off to see if they're the problem

No issues here so far but I'm not playing Assetto Corsa yet. I had someone SUPER sensitive to motion sickness play Prepar3D on the motion sim in my CV1 Sunday and he stayed in the rift for longer than five minutes for the first time ever! He flew for nearly an hour.

Based on that I think it's safe to say I'm not having whatever issue you are experiencing, at least not in P3d. Note that my rig is a bit of a beast, 4790K @4.5Ghz and GTX 1080 OC'd.

I turned off PP today resulting in a nice, non-nausea simdrive.
I had turned PP on because with ASW I could finally handle it, but the stero rendering is still bugged, mainly in the far distance with objects or effects that sometimes appear only on one lens.

Let's wait for a PP VR fix...

Kunos' history with VR support isn't all that great. Given how poor it was in the DK2 days, I'm delighted with where it is now. It would be nice to have it 100% working, including PP and the full menu UI, but I think we could be waiting a while for that.