Forum Replies Created
-
Posts
-
Me too I find this problematic. I guess a work around would be to assign the top tab button as touch and use the joystick while resting your index finger against it.
Hard to press trigger at the same time though.
It does feel easier/faster with emulation. I think it’s because the Navi is not as ergonomic as the move: hard to find the buttons with HMD on. The buttons aren’t positioned close enough to finger tips.
Good to have velocity for PSMoveServiceBridge. Looking forward for it to be available in color tracking too. I think Blade & Sorcery is now on that list also, with the magic orbs throwing.
I figured out how to recreate it. If I use the first two section to correct yaw, it will somehow swap roll and pitch. If I use the third section, Drivers absolute offset, it doesn’t.
Therefore, I pick the last one. 😀
By the way, other than re-install to fix the swapped axes issue (pitch becomes roll, roll becomes pitch), i switched to Kinect skeleton for controllers and the problem disappeared when i switch back to colour tracking. Something funky with PSMoveService? I dunno.
I don’t think any of them would fix the drift in position…
I tried the first and third section. But I can double check to be sure. Haven’t paid much attention in game since it isn’t that noticeable.
I’m using ps moves. What I have said works for controllers as well.
Yaw offset in first section would rotate the controllers back to the front.
The positional drift over time happens to the entire setup: hmd and controllers together. So it’s not that noticeable. Unless it’s in beat saber or pirate trainer since game doesn’t recenter.
Yes, it turned out the first section works fine to fix the yaw problem.
Position offset over time is less of an issue. Lots of games recenters on player in menus.
That’s right.
Quick calibrate and fix floor does the job. And hmd standard offset in yaw can fix orientation but not position (somehow, it drifts elsewhere than center).
It’s gone now… I guess my overclocked system was just bugging out? I’ll let you know if it becomes a problem.
Well, I have described the situation and solution to Greg already on Discord. I guess if more users ask for it he will do it.
FYI, I do the following to fix it:
1. calibrate button to fix angle between HMD and kinect v1
2. recenter standing space in steam setting quick calibrate then fix floor
3. when it drifts, i go into Controller Shift/Offset and press up/down in HMD, Adv. Offsets, Driver absolute offset.
4.repeat step 3 mid game when it drift too much.If he put it in Virtual move and rotation menu, we can do it without using arrow in the menu and on the fly. It might conflict with game commands but drift is worst…
Unfortunately no. I can also use your HMD angle calibration to fix it to some extend (orientation will re-align but position will be elsewhere).
In Kinect controller settings window, Greg wrote : “Overwrite controller’s orientation from assigned device like PS Move, Joy-con. Use only device’s buttons.”
@greytoshiaki It IS for ANY device. Not just PS Move.Since re-center is possible in SteamVR via advanced settings or setting menu, kinect/hmd angle auto correct is essentially the same as hmd angle correction. It would be great to be able to assign a key to it to use with this feature.
Hi Greg,
Is it possible to use this feature to trigger the calibrate button in the auto correction in “calibrate angle between headset and sensor” window?
I moved from using Riftcat to a direct to HMD solution. Small drift from HMD is still there(or offset from HMD resting position before initializing). Other than room setup, there is no other way to correct this.
Please advice. Thanks!
Hi Greg,
I notice that it happens frequently in two occasions:
– controllers out of KinectV1 range (i got the big POI balls so it’s better 🙂
– fast movements (i guess it exceeds the 30Hz tracking)Sometimes, it even stay stuck at some position(longer than arms length) until i physically move the controller back to it or Driver4vr frozen and cannot be terminated (Driver4vr fps 0).
Is it possible to improve how Driver4vr error handles *tracking loss or cpu overload)? Crash recovery, restart, etc…
Thanks.
PS: I have to be so far away because of the narrow fov of KinectV1 (to be able to grab objects on the ground and wave my arms in the air). That means my controllers are even further down range when facing slightly away from sensor (~3m to 4m).
Nevermind, PS Navi is recognized as Xbox360 controller… That’s the solution from PSMoveService to use Navi in windows 🙁
I see you are experimenting PSMoveService directly for hip and feet tracking in Discord.
Maybe you can get the buttons from PSNavi via PSMoveService too? I remember that it works 100% in the PSMoveConfigTool once setup per DaleyTech’s video.
Don’t remember whether Steam pick it up as a controller though.
You mean assigning the vrgamepad to a controller in driver4vr? Likr for leap motion?
How about actually replacing the touchpad from psmove with another controller at all?
no… Nothing is working on my end. 🙁
He made it work without an ID at all. I think he completely removed it from PSMoveService api and steamvr.settings. That’s how i get the no ID situation(I did tried that too but it does’t work either). Now, i have enabled it in PSMoveService and shows up as w before you manually change the ID. It worked without an ID for him so I don’t think that is the issue.
Yes. I remove all controller support. Even remove the mapping scheme for 360 controller.
When I try to use it in game, steamvr thinks it’s a gamepad.
But at the minimum, it still have some basic gamepad support and i can’t disable that. I guess it’s some kind of conflict? Did anyone else able to map a normal 360 controller to motion controller?
I still can’t preview navi in device manager of driver4vr even though assignment works.
Better. But unfortunately, still no.
I can see joystick movement on the touchpad of the controller model. However, it’s not a solid dot. It’s flashing like it’s intermittent. I map the rest of the buttons but none of them do anything.
My nav can still move the cursor in the steamvr menu and select with X in the overall menu but not in each dashboard like the trigger.
So far, I have tried with or w/o using PSMoveService, with or w/o manually adding mapping in steamvr.vrsettings and it’s still not functional. It only works as a gamepad in SteamVR.
If I don’t enable it in PSMoveService in DeviceManagerConfig.json, it shows up as gamepad W in Driver4VR. Button assignment are in steamvr.vrsettings (like you guys had back in October).
If I enable it in PSMoveService, it shows up as gamepad with no ID. It works fine in PSMoveServiceConfig and in Windows Device as Xbox360 Controller.
Can you let me know those details concerning the setup you are using right now to make it work? And how do you avoid it behaving as a gamepad in SteamVR? Even without Xbox 360 controller support, it still behaves as a gamepad(Dpad and Joystick moves cursor in SteamVR menu but trigger is not available).
I want to use it for TouchPad and forward/backward move locomotion.
I saw the tutorial for the old gamepad button assignment. And I also tried the new one. Even though the assignments worked. It doesn’t in SteamVR.
Is the PS Navi supposed to show up as a PS Move device in device manager that can be previewed? Mine doesn’t. But in button assignment, it detects the buttons presses.
In SteamVR, after bringing up the menu with PS Move, I can move around in menu selection. But I think that’s just 360 gamepad behaviour right?
Hi Jonny,
Can you tell me how you disabled 360 pad support in Steam? I did it in General Controller Settings in Setting/Controllers but the navi is still working like a 360 controller.
I can assign buttons in Driver4VR but it doesn’t work in SteamVR. Did you went through the entire setup in Daley Tech’s tutorial or just the SCPTool part?
Thanks.
For most part, it works well with some tweaking. You can use Advanced settings in SteamVR to compensate position, orientation offsets.
Me too. Since the controllers swaps randomly at restarts, reassignment was required. In my case, rotation was almost always reversed even in previous releases.
Got the POI balls installed and tracking is way better at long range and almost 360. A bit heavy even after removing the original ones but it works great. 🙂
Not all the time since the gyro is not as well calibrated. It would be great to have a joystick at some point.
Yay! Rotation is working fine now and tracking in general is a lot more stable than before in spite of max CPU load. Along with RiftCat 2.0, it’s working great. Thanks Greg!
Hi Greg,
I think this issue was fixed for a while before the OpenVR update. Unfortunately, it is back in the current release… Can you take a look?
Thanks.
Skeleton tracking alone is playable in a lot of titles with the right expectations(PS move service/PSVR level and not WMR/Oculus/Vive level). Most people would be thrilled to be able to play VR with their old hardware/tech for such a low cost thanks to Greg.
The trade off for precision is CPU load for colour tracking. Precision is only crucial for games that requires very accurate aiming: pushing multiple small buttons on virtual console/keypad consecutively, picking things up off the ground, less forgiving shooters like Arizona Sunshine,etc. At this the current state, it is already playable.
That’s my 2 cents after trying skeleton and colour tracking with PS Move controllers 🙂
After more gameplay, for seated position, colour tracking with ROI for everything is still the best option.
For some reason, skeleton tracking for seated position gets worse with time. After a long while, it’s more shaky than colour tracking HMD.
The only problems left are the minor jitter for colour tracking and sometime crashing/glitches because CPU utilization is maxed out.
It’s definately playable for long duration with colour tracking for now!
Btw, it’s not just HMD. Colour tracking in general is jittery. The controllers are vibrating when I’m using skeleton head tracking. It’s just that when it’s the HMD, its more apparent because the whole Fov shakes. It does’t matter whether it’s my or ps move’s led ball.
Nevermind about the vibration, it was a game glitch.
My bulb is an orange table tennis ball glued to a led. It is a dimmer than the PS Move orb but I found an exposure suitable for all three with the lights off in the room. I have wrote my experience in the announcement thread.
I had a better experience in comparison. It is definitely way way better than first beta. Thank Greg for implementing the solutions/improvement and i’m looking forward to try the reflector and larger POI balls. 😀
Skeleton tracking is good for short to long range. It’s the less precise and hence affects orientation of controllers when swinging around 360 but with CPU at ~50%, plenty of power left for the game to run smooth. Even if the skeleton is messed up, it recovers quickly. It is prone to occlusion problem when limbs/body overlaps. Ex: bow and arrow, dual yield anything at any angle. Good for all games in the lab but don’t expect precision.
ROI disabled is the next more CPU intensive one @ ~70%. Great for seating position. It is range sensitive: precise and jitter at short range and shaking a lot/loose tracking for a while/or permanently at long range. Tracking is worse during fast movement but handles dual yield guns pretty well (less prone to occlusion problem). Therefore, it’s good for controllers precision but not HMD(constant jitter still).
ROI enabled is the most CPU intensive one @ 80-90%%. It is smoother than the last one and the best of all. But since CPU is almost maxed out, when the game become more CPU intensive, tracking/buttons will freeze/glitch sometime temporarily, sometimes permanently. Bulb range issue does affect the tracking negatively still.
Just tested it a bit i few games at close range without moving the sensor.
Wow, it works better! It was really a range issue: now that the CPU load is less, it recovers faster and I can practically dual yield guns close to 360 all around even when there are occlusions.
My calibration was actually like you have explained. Jitter is worse with distance from sensor.
It is playable at room-scale but with jitter.
PS. did you enabled vibration? One of my games(using Unreal engine) starts to have it! But the controllers tracking don’t work(stayed on the floor at default position but with functional buttons). Don’t know whether it is a problem with the game or Driver4vr.