• Welcome to PiBoSo Official Forum. Please login or sign up.
 
March 29, 2024, 11:40:07 AM

News:

World Racing Series beta14 available! :)


Oculus Rift public release

Started by RiccoChicco, January 06, 2016, 04:31:48 PM

Previous topic - Next topic

HornetMaX

Quote from: h106frp on February 23, 2016, 09:34:02 AM
Something i have been thinking about is the impact on the first person view bike models in GPB. I assume with a VR headset you will have the ability to look around the cockpit area and bike more than you can currently and see areas that would have no model defined.

If someone has a full tracking VR headset it would be useful to get some idea of the extent of the view movements available.
You're assuming that the VR device being 6dof (3 rotations, 3 translations) the game/sim will allow you to use the 6dof.
But that may not be true, e.g it could only allow the 3 rotational or it will allow translations too, but in a limited fashion (because the rider's head can only translated that much when the rider is on the bike).

h106frp

On many current bike just small increases in viewing position would give strange results. Most 1P models in GPB have no lower chassis, wheels, swingarm, front suspension and have heavy geometry culling of unseen faces - could give strange results.

I guess most people who shell out £1000 for VR will be wanting full DOF and personally i think i would find full immersion VR without the 6DOF movement quite weird and possibly very disorientating/vomit inducing. ;)

I did notice that looking at the MXB 1P model definition the default bikes declare full geometry for all components, i guess this is because with no fairing and a narrow tank more of the bike is already visible.

HornetMaX

Quote from: h106frp on February 23, 2016, 10:07:09 AM
On many current bike just small increases in viewing position would give strange results. Most 1P models in GPB have no lower chassis, wheels, swingarm, front suspension and have heavy geometry culling of unseen faces - could give strange results.
But that's easy to fix, it's just a cheapo optimization.

Quote from: h106frp on February 23, 2016, 10:07:09 AM
I guess most people who shell out £1000 for VR will be wanting full DOF and personally i think i would find full immersion VR without the 6DOF movement quite weird and possibly very disorientating/vomit inducing. ;)
But what is full dof when the rider is on the bike and I physically move 5m away ?!

VR without translations is 3dof head tracking, like EDTracker (but better, as the screen is in front of your eyes, no matter where you look at). There's no vomiting with that.

h106frp

Your probably correct, but i am just thinking that if i am genuinely fooled into believing this VR environment and i move my head physically forward/backward/sideways but do not get the corresponding change in my VR environment i think i might start getting a bit queasy as the poor old brain struggles with this new concept.

I do wonder if this is why Occulus implemented the full tracking regime late on in the development cycle.


Hawk

I'd have to agree that the 6DOF is the way to go in my opinion. When you make any movement of your head in any direction you really do need that to reflect in the view. Plus as H says it helps stop people getting motion sickness.

Hawk.

HornetMaX

Quote from: Hawk on February 23, 2016, 12:25:17 PM
I'd have to agree that the 6DOF is the way to go in my opinion. When you make any movement of your head in any direction you really do need that to reflect in the view. Plus as H says it helps stop people getting motion sickness.
There're plenty of other things that could give you (much more) motion sickness: first, you're not sitting on a leaning bike but on a still chair, so trying to have the visuals in sync with your real position/attitude is doomed to fail from the start. Bike will lean, virtual rider will lean, you will not. Also, main reasons for motion sickness were tied to low refresh rates and/or poorly implemented stereoscopic 3d.

I'm fairly sure 6dof for us (and other car/flight sims) is mostly gimmick. We (GPB) don't even need 3 dof strictly speaking: how to handle properly the head tilt is unclear to me. Testing with EDtracker didn't give anything more satisfactory than the basic tilt GPB provides.

teeds

Problem with a DOF system and VR is the movement of the rig, no one's figured out how to pass the movement data to the rift from the platform motion - yet.

Quote from: HornetMaX on February 23, 2016, 12:44:16 PM
Bike will lean, virtual rider will lean, you will not.

But you could, and I do to some degree with my rig on a flat screen. Not sure anyone apart from Pibs has run GP bikes with VR, anyone? One guy ran MXB with a DK2 and he didn't mention a bad feeling at all.

Not yet sure how the head movement experience will work, or not. From what Pibs said a while ago I don't think we'll have positional tracking at first, only direction. Maybe we can assign the rifts positional tracking like Edtracker, if not I have a my tracker setup for body lean to try. Otherwise i'll be matching what the auto rider does, not sure i'll be able to help it TBH but will I puke if don't, we'll see  8) I'm betting on no ;)

Quote from: Hawk on February 23, 2016, 09:44:02 AM
This is one of the reasons why I personally think a part model of the bike to represent the cockpit is not a good idea(although currently I presume it's for efficiency reasons?)?
I'd have thought that using the full bike model(with detailed cockpit included) with the eye-view(head-camera) working in unisn with head movements would be a better option for VR use?

I've thought the same, not sure why it's not done that way though?

Hawk

@teeds: Klax75 has an Occulus Rift DK2. Not sure how he's got on with it in GPB though?  ;)

Hawk.

h106frp

I thought with VR the concept was that if you tilt the view in the headset your brain is fooled into thinking you really have tilted and then when you move your head you just add the correction and the view levels out.

I was assuming it would be like the old 'cinema 360' where you can watch everyone leaning in sympathy with the projected display even though the floor never moves.

Hawk

Quote from: HornetMaX on February 23, 2016, 12:44:16 PM
Quote from: Hawk on February 23, 2016, 12:25:17 PM
I'd have to agree that the 6DOF is the way to go in my opinion. When you make any movement of your head in any direction you really do need that to reflect in the view. Plus as H says it helps stop people getting motion sickness.
There're plenty of other things that could give you (much more) motion sickness: first, you're not sitting on a leaning bike but on a still chair, so trying to have the visuals in sync with your real position/attitude is doomed to fail from the start. Bike will lean, virtual rider will lean, you will not. Also, main reasons for motion sickness were tied to low refresh rates and/or poorly implemented stereoscopic 3d.

I'm fairly sure 6dof for us (and other car/flight sims) is mostly gimmick. We (GPB) don't even need 3 dof strictly speaking: how to handle properly the head tilt is unclear to me. Testing with EDtracker didn't give anything more satisfactory than the basic tilt GPB provides.

I have Track IR which has 6DOF and I think that's great, unfortunately we cannot use Track IR in GPB which is understandable due to the licensing requirements for Piboso to have to meet(which I think is a big mistake from Natural Point, they should make it free license for developers to promote more sales.  ::)

Anyway, I do understand what your saying there mate about causes of motion sickness and agree with you. But I still feel that in virtual reality you should be able to move in any motion or direction that you can move in reality to get the very best and a more natural experience.  ;)

Hawk.

teeds

Quote from: h106frp on February 23, 2016, 02:14:53 PM
I thought with VR the concept was that if you tilt the view in the headset your brain is fooled into thinking you really have tilted and then when you move your head you just add the correction and the view levels out.

Are you talking about motion platforms + VR or DOF on a tracking system?

Quote from: h106frp on February 23, 2016, 02:14:53 PM
I was assuming it would be like the old 'cinema 360' where you can watch everyone leaning in sympathy with the projected display even though the floor never moves.

That will happen when standing for sure but grab something fixed down and things seem to change in that respect.

h106frp

February 23, 2016, 02:40:20 PM #26 Last Edit: February 23, 2016, 03:08:53 PM by h106frp
Quote from: teeds on February 23, 2016, 02:34:24 PM
Quote from: h106frp on February 23, 2016, 02:14:53 PM
I thought with VR the concept was that if you tilt the view in the headset your brain is fooled into thinking you really have tilted and then when you move your head you just add the correction and the view levels out.

Are you talking about motion platforms + VR or DOF on a tracking system?

Quote from: h106frp on February 23, 2016, 02:14:53 PM
I was assuming it would be like the old 'cinema 360' where you can watch everyone leaning in sympathy with the projected display even though the floor never moves.

That will happen when standing for sure but grab something fixed down and things seem to change in that respect.

Just using tracking feedback to the VR display to allow them to interact.

Quite a few issues considered (some i would not have thought of) in this google guide for its cardboard device;
https://www.google.com/design/spec-vr/designing-for-google-cardboard/physiological-considerations.html


edit: always like a practical experiment, ordering a google cardboard v2 to see what all the fuss is about  :)