WMR Analog Joystick Calibration and Drift

[Torzi]

I've noticed several other recent posts on the topic, and thought I'd chime in.

My analog sticks are putting out a non-zero value when physically in the neutral position. This can be seen in the Cliff House... The virtual models of the controllers show the analog sticks in an off-center position:
https://steamcommunity.com/sharedfiles/filedetails/?id=1654919517

It's hard to get a great angle on the screenshots, but the left controller is putting out a value to the right and very slightly forward, while the right controller is putting out forward and slightly to the right. While in the Cliff House, I barely have to nudge to sticks to get them to register a right turn or teleport forward. Turning left, or the backward motions take much more effort/movement on the analog sticks. The deadzone is set sufficiently here to keep these actions from happening on their own in Cliff House.

With the new beta branch of SteamVR (1.3.x), it looks like Valve is working on adding full WMR controller support, but they don't have dead-zone support for the the analog sticks yet. The current problem is that these non-zero values are being passed on to SteamVR, which is causing drift in games if the analog sticks are mapped for movement. If the game or program supports WMR natively, they usually take care of the dead-zone themselves, but in older games without support, or in the SteamVR Dashboard, the lack of the dead-zone causes drift.

While I'm sure the dead-zone situation is temporary, would it be possible to get proper calibration routines for the analog sticks from the driver directly to re-center and set the range of motion? It would be great to be able to set the input sensitivity curve as well, but I guess that's pushing it. Not only would this diminish or negate the need for dead zones, but more importantly it would allow for much greater sensitivity of the analog input.

Thanks

My Specs:
Windows 10 1809, Build 17763.253 - clean install, updated regularly
i7-4770k, 16GB RAM
GTX 1080 w/ driver 417.35
Samsung HMD Odyssey+ XE800ZBA w/ driver 10.0.17763.1069, firmware 1.0.1
Samgsung HMD Odyssey Controller Right AA-HC1HURB
Samgsung HMD Odyssey Controller Left AA-HC1HULB
Mixed Reality Portal 2000.19011.1132.0
Samsung HMD Odyssey+ Setup 1.8.2.0
Windows Mixed Reality for SteamVR 1.0.398

[Girth Control]

There is an adjustable deadzone in the "default.vrsettings" file but it only works when you enable the experimental artificial movement system which breaks many, many games. I made a post about it just yesterday. An adjustable deadzone for WMR should be easy to implement seeing as they've already programmed it, it just needs to be enabled separately from the artificial movement system. Hopefully the WMR team sees our posts and can work towards a solution :)

[Torzi]

I'm betting Valve will actually fix the deadzone problem... they have options for it, it just isn't applied correctly at the moment. They're definitely working on it though, since I'm seeing changes to controller behavior with each new beta release.

I'm more interested in the calibration issue, and was hoping to get a response from someone at Microsoft. Sure, setting a large dead-zone will fix the drift problem, but you're losing precision. If the same control output takes different amounts of effort in different directions, it's hard to be precise... particularly with mouse pointers and aiming, but also if you're trying to control steering or throttle with it.

Ideally, no dead-zone at all would provide a tighter, more responsive control... as you're not losing 25% of the controller action, and a nudge would give the same level of output in any direction.

[WindowsMixedReality]

Hi Torzi and Harmacist,

My name is Julio and I work on the Mixed Reality team at Microsoft. We are always interested in hearing from our customers what they would like to see changed or see next in our products. Please submit this feedback to us using the Feedback Hub. You can find instructions on how to do that here:
https://docs.microsoft.com/en-us/windows/mixed-reality/enthusiast-guide/using-steamvr-with-windows-mixed-reality#sharing-feedback-on-steamvr

Thanks,
Julio from Microsoft

[The Male/Man]

+1

Originally posted by WindowsMixedReality:

Hi Torzi and Harmacist,

My name is Julio and I work on the Mixed Reality team at Microsoft. We are always interested in hearing from our customers what they would like to see changed or see next in our products. Please submit this feedback to us using the Feedback Hub. You can find instructions on how to do that here:
https://docs.microsoft.com/en-us/windows/mixed-reality/enthusiast-guide/using-steamvr-with-windows-mixed-reality#sharing-feedback-on-steamvr

Thanks,
Julio from Microsoft

We all know nobody from the actual microsoft development teams do what we want, otherwise we'd have the aero glass effect from Windows 7 in 10.

[SintKebab]

Originally posted by WindowsMixedReality:

Hi Torzi and Harmacist,

My name is Julio and I work on the Mixed Reality team at Microsoft. We are always interested in hearing from our customers what they would like to see changed or see next in our products. Please submit this feedback to us using the Feedback Hub. You can find instructions on how to do that here:
https://docs.microsoft.com/en-us/windows/mixed-reality/enthusiast-guide/using-steamvr-with-windows-mixed-reality#sharing-feedback-on-steamvr

Thanks,
Julio from Microsoft

Any news on this issue?

In SteamVR Dashboard and Home, my right hand controller's stick seems to report slightly north of center when it should be neutral.

In Cliff House it's working fine, and I can't say I've had any issues in any games/apps.
But it's incredibly annoying in SteamVR to have to hold the stick back a little, for it not to scroll randomly.
The amount of error varies as well, so some times it only scrolls a line every 10 seconds, but other times it'll suddenly bounce to the top of the page.
And that variation makes it hard to compensate.

Should the controller be replaced, or should the software improve somehow, as it works perfectly in certain situations?

[Torzi]

It works now whenever deadzones are applied. The underlying issue is still there... it just gets masked because the non-zero input is ignored below a certain threshold. If you look at the analog sticks in the Cliff House, you can see they aren't perfectly centered.

The point I was making was that using large deadzones reduces precision on the controls. This would be fine if were some cheap $20 controller, but they are not, and there is NO alternative.

This IS easily fixable with a firmware update. But since Julio from Microsoft expects me to go create an issue in some other board, and I haven't had time to do it, it's likely that this issue hasn't gone any further than this post.

[The Male/Man]

Originally posted by Torzi:

It works now whenever deadzones are applied. The underlying issue is still there... it just gets masked because the non-zero input is ignored below a certain threshold. If you look at the analog sticks in the Cliff House, you can see they aren't perfectly centered.

The point I was making was that using large deadzones reduces precision on the controls. This would be fine if were some cheap $20 controller, but they are not, and there is NO alternative.

This IS easily fixable with a firmware update. But since Julio from Microsoft expects me to go create an issue in some other board, and I haven't had time to do it, it's likely that this issue hasn't gone any further than this post.

We need Linux to look better.

[SintKebab]

Originally posted by Torzi:

It works now whenever deadzones are applied. The underlying issue is still there... it just gets masked because the non-zero input is ignored below a certain threshold. If you look at the analog sticks in the Cliff House, you can see they aren't perfectly centered.

The point I was making was that using large deadzones reduces precision on the controls. This would be fine if were some cheap $20 controller, but they are not, and there is NO alternative.

This IS easily fixable with a firmware update. But since Julio from Microsoft expects me to go create an issue in some other board, and I haven't had time to do it, it's likely that this issue hasn't gone any further than this post.

Ah, that's too bad.
I'm not going to do that kind of workaround, potentially breaking app functionality.
Thank you for the heads up.

[The Male/Man]

Originally posted by Torzi:

It works now whenever deadzones are applied. The underlying issue is still there... it just gets masked because the non-zero input is ignored below a certain threshold. If you look at the analog sticks in the Cliff House, you can see they aren't perfectly centered.

The point I was making was that using large deadzones reduces precision on the controls. This would be fine if were some cheap $20 controller, but they are not, and there is NO alternative.

This IS easily fixable with a firmware update. But since Julio from Microsoft expects me to go create an issue in some other board, and I haven't had time to do it, it's likely that this issue hasn't gone any further than this post.

How do you apply deadzones?