Using feet for movement in a VR world

[DaveNagy]

This is a thought provoking thread. I am now going to inflict some of those thoughts on you, since you provoked them in the first place!

If I understand nateight's "Guitar Hero" controller correctly, it is designed to rotate your virtual hips/body? Body-yaw, in other words? It would do the same thing as "rolling" the right analog-stick of a gamepad style controller? Or, does it kind of work kind of like tank controls, where you can control both the forward speed and rotation of your VR body?

If it's orientation-only, I can't help but wonder if it's really bringing enough "newness" to the party. After all, the ol' right-stick works pretty well already. DO we really need to do that with both feet instead? And since head tracking makes it so you don't need to rotate your whole "body" just to look around, adding a third gadget to slew your view around seems kinda like we're heading past diminishing returns land and might even be getting close to dead horse beating territory.

Or, does nateight advocate ditching the right ("view") stick entirely? That could be interesting. All up/down looking would be done through head tracking, and left/right "hip rotation" would be done with the pedals. That means your "gun controller" would only need one thumb stick (for movement), instead of two. That's a worthwhile improvement, IMO. Especially for people that are not proficient with twin-stick game pads already.

But if we are just trying to come up with a foot controller to control only "hip yaw", I think there might be an alternative. (I got this idea after reading somebody's description of gluing a flip-flop to a DJ jog wheel... or something. Frankly, it confused me. But I liked the idea of a big free-spinning disc on the floor.)

So yeah, that's the extent of my idea. A free-spinning disc, with a bit of weight/inertia to it, hooked to some sort of rotary encoder. It would give "direct 1-to-1 control" of hip rotation, much like the side-to-side motion of a mouse. I imagine it being about 20 inches (50 cm) in diameter, and you would spin it with both feet, while seated. For small/slow turns, both feet would stay in continuous contact with the disc, kinda like it was a floor-mounted steering wheel. For bigger/faster turns, you'd use more of a toe-flicking motion to get the disc turning, and then let it free-wheel until you reached your desired rotation, at which point you'd plop you feet back down to stop your turn. 360s or 720s should be do-able if you want to lose your lunch!

I think the pros of this controller would be threefold:

1) It would be simple (cheap) in construction, and pretty bullet proof.
2) It would allow "direct" 1-to-1 control of your hip orientation. While I actual like joystick style indirect controls, it may be that direct controls are a better match with head tracking, which is also direct.
3) I think it would be intuitive, even to novices. "Steering" the disc with your feet is not too dissimilar to actual hip-rotation while standing. And quick 180 (or more) degree spins would be somewhat akin to a "jump turn" where both feet leave the ground after initiating a strong twisting motion.

Alert readers will have immediately noted a forth advantage to this controller: pottery wheel sims!!!

[nateight]

... I want it to be something natively supported and more importantly programmed to by a wide range of game developers.

I'm not sure why you keep reiterating this point, because it's not an issue. All a simple direction foot controller needs to do is emulate an HID gamepad with two analog sticks. No particularly special support required. If the game supports a gamepad, it'll support the foot controller.

Try an experiment for me. Plug two separate gamepads or other controllers into your computer – two 360 controllers, Wiimotes via Bluetooth adapter, flight stick + a second joystick, whatever, just so long as Windows sees two separate devices and not some Saitek monstrosity with multiple parts but one integrated driver. Now go through your game collection, Steam library, whatever, and load up every game with gamepad support - AAA FPS games, indie platformers, the works. Almost all of your games will have some form of gamepad support, the majority of those are likely to be console ports allowing for different button mappings chosen from a static list, and a sizable minority will even let you remap individual buttons freely according to your preference. Of the last group, which I roughly estimate comprises less than half of the games you can currently purchase from Steam or some other commercial entity, you're very likely to own several. Get into the options screens of some of these games, and try to map half the game commands to one of your controllers, and the other half of the commands to the other controller. It's not unheard of for this to work, but unless you exclusively play flight sims or other games with extensive and uncommon levels of joystick support, of most games you're likely to try, even the ones that will let you remap buttons from a single controller in wacky and alien ways like making the triggers control forward and reverse motion, the vast, vast majority will just stare at you blankly, unable to accept a mix of inputs from multiple sources. Remember that even getting to a list of games where this level of experimentation was possible pared away most games you could name, and consider that this may, in fact, be an issue. Game developers of all stripes have long treated joystick and gamepad support as an afterthought; people continue to resort to PPJoy when they're exploring experimental control schemes despite it being buggy, kludgey, and both unsupported and lacking a stable release for something like five years - it's still the only way to shoehorn two whole inputs into most modern games. If devs aren't actively thinking about supporting foot controllers they may indeed work as a sole “gamepad” device but very often will not work as part of a more complex input scheme, and while you may be right that a lack of official support for weird controller combinations may not constitute the death knell of VR, it's still an outcome I believe deserves a certain level of action and urgency as opposed to indifference and acceptance. If you would prefer this be a campaign about urging devs to support a wider range of all forms of input rigs including foot controllers instead of focusing on that one aspect I'm all for it, but to assume they'll get any of this right without significant community pressure belies 30-odd years of the industry screwing up in virtually every way they could.

All up/down looking would be done through head tracking, and left/right "hip rotation" would be done with the pedals. That means your "gun controller" would only need one thumb stick (for movement), instead of two.

This. I wasted half the day tracking down camera parts and a tripod, but I should be all set now to produce a little video fully explaining how I see the pedal thing working; if I can't find the time tomorrow I should have no difficulty having it together by Saturday night. Incidentally, jogwheel-style spinners are worth some experimentation but I'm having difficulty believing they'll gain widespread adoption even if they work pretty well; I'd hate to live in a world without a micropayment store for virtual pottery glazes and DLC kiln and firing options, but it may be where we're headed.

[WiredEarp]

Interesting discussion here guys. I'm posting from work so haven't had a chance to analyse everyones posts in as much details as I would like, but I do have a few questions and ideas...

Firestorm-style pedals such as I'm proposing are spring-loaded - they return to full height / stick center the instant you take your foot off.

How comfortable is it controlling a character using the foot pedals? I mentioned the centering part compared to flight simulator sticks, as when flying in the simulator, its much harder to use foot pedals accurately when they are not connected physically. I used my racing pedals for a while instead of 'real' pedals, but my flight sim pedals give me much more control, as they have an obvious midpoint and your feet dont end up fighting each other in panic situations.

I think the most intuitive approach would be a horizontally flat pedal under one foot that can be tilted forwards/backwards for moving the avatar accordingly, as well as being rotatable left/right

I think having the ability to move in reverse is a good one - however, its probably not necessary to have a 'flat' pedal. As long as there is say 10% travel for moving backwards, that should be sufficient - we probably dont need full travel for backwards movement, and it would probably be less strain on the ankles.

What youn want for a high precision foot control is getting rid of those gamepad mechanics, you want it precise like a mouse. In the thread Msat postetd, i wrote some ideas down, basically a pedal for turning which is making contact with a disc when pressed down and can be recentered easy. The disc is coupled with a pentiometer for free turning. Those ideas are there waiting to be picked up by people who have the skills of building such a controller.

Its interesting, I have a slightly related idea - you have 2 platforms for your feet, when you press on one enough, it 'registers' that the foot is down, then moving that foot moves your virtual body as though it was balanced above the foot. Lifting your foot lets you reposition for another step. However, my concept was more of a walking simulator, than a foot controller. The job wheel idea sounds interesting, basically if I understand right, its like using a spinning office chair, but having the ring spin around you, instead of you turning? I think that could work. The difficult part would probably be hooking the pedal into it in a way that was comfortable and intuitive.

In a perfect VR World, the mouse would get stripped into 2 independent controls.
1. Aiming = VR gun
2. Turning = foot controller pedal

In a perfect VR world, the mouse wouldn't have even been an input option! I still think the turning can be done simply with a thumbstick on the gun, although I would prefer to use my feet to control my body.

fine aiming should be done with a separate device (ie Hydra), which hopefully would be very accurate.

I agree, IMHO ALL aiming should be done with a dedicated device (gamegun). This will provide the accuracy - and also mean that we don't need mouse type accuracy for our general movements. Even if we do a gross turn and are 20% misaligned with the target, we can just turn our heads and aim our gun - which will mean that the loss of accuracy going from a mouse movement scheme to a proportional turning speed scheme will not be an issue.

That means your "gun controller" would only need one thumb stick (for movement), instead of two.

I'm not sure why we would EVER need two thumb sticks - is it for strafing? One thumb stick should give us forwards, backwards, and turn the body (hips) left and right. We probably can do without strafing, since we will have the ability to shoot at an arbitrary angle to our movement anyway. So we can move at 90 degrees to the target, and still shoot at it. Another cheap/easy option could be a small trackball on the back of the gun, if people wanted more mouse like turning behaviour.

I dont really see any of these solutions working for stand up currently. Pedals will throw off peoples balance when moving them forwards and backwards... However, standup doesn't really require foot input, since you can then use hip sensors and a thumb stick to allow both turning AND strafing.

people continue to resort to PPJoy when they're exploring experimental control schemes despite it being buggy, kludgey, and both unsupported and lacking a stable release for something like five years - it's still the only way to shoehorn two whole inputs into most modern games.

I agree about support for foot controllers etc, but I do have to say, I don't find PPJoy buggy at all. Its just a PITA to get going... but once you have it going, it seems very stable. Its run all my games (I have all my devices mapped into a single virtual joystick) without a single issue since I finally got it setup.
I think the best thing for devs to do is support FreePIE, and then the community can expand that to do what PPJoyJoy does, allow new controllers, etc.

[TheHolyChicken]

Here's an idea just to throw a cog into the machine:

(for seated play) What about using your feet on a wii balance board type device, for WASD movement, and motion tracking your shoulders to turn left/right?

If you are sitting right now, try turning your head far in one direction, while paying attention to your shoulders. After a point (at ~90°+) your shoulders must move to allow your head to turn further. A tracker would pick up on this shoulder movement, translating it into avatar rotation (like pushing gently on an analogue stick left/right).

[MSat]

@nateight

I see the point you're trying to make now. So it's not technically an issue with supporting a foot controller per se, but rather the lack of game support for simultaneous input devices. So, yeah - developers either need to support multiple devices used at once, or gamers will need to use something like PPJoy.

[nanicoar]

Force Feedback!

Add a mighty version of this:

Plus control logic, and the balance board won't simply stay put but instead react to the slope and texture of the virtual world. - Better to render this as some proper prior art for keeping those evil patents away.

Say you have your feet placidly planted on the balance board and you move your character to the foot of a hill. There is some gravel there, so the platform beneath your feet shakes with amazing detail and accuracy to tell you that it's gravel, and perhaps even turns very briefly around the z-axis because you fumbled a balance check. When you start up the hill, the balance shifts so you have to press down with your toes and your feet naturally counter these movements.

Since these sensations come from your feet, they should be very immersing.

[nateight]

I see the point you're trying to make now. So it's not technically an issue with supporting a foot controller per se, but rather the lack of game support for simultaneous input devices. So, yeah - developers either need to support multiple devices used at once, or gamers will need to use something like PPJoy.

Yay, communication! As happy as I am to get us all to an understanding of the problem, it remains a sticky one, and I'm unaware of any current option that is better or even comparable to PPJoyJoy. And that experiment I mentioned? I've become the latest victim of both short-sighted joystick support in every game in my Steam library and PPJoy's inability to be installed on a computer without a parallel port. Mechwarrior Online was the only game I tried that let me dare to mix multiple joysticks, but I discounted it as a worthy example because it's a 'mech game with a throttle and numerous other non-human-avatar trappings. It seems like Unity is perfectly happy to allow this (there are rumors of difficulties under XP, but under Vista that demo works as it should with pedals + gamepad), so rather than give you guys a video lacking the crux of my argument, I posted a topic calling for PPJoy's modern replacement, and I'm now going to disappear for a day or two in an attempt to cobble together a demo in Unity. I'm still not positive pedals are the answer, but they're definitely an answer; if Unity is as simple and easy to dive into as everyone claims I should be able to demonstrate why I like the pedal concept shortly.

(for seated play) What about using your feet on a wii balance board type device, for WASD movement, and motion tracking your shoulders to turn left/right?

Whether it's using a Balance Board + a Kinect or just a Balance Board + a thumbstick, I really think it's a device that has potential, and it has the huge added benefit of being something a dev can find in a brick and mortar store and have running the same day. I just don't know how fine the control is, how bad the notorious latency of the Wii translates for peripherals tied into a PC, or if you can even operate a Balance Board sitting down! Having a small range of devices ready to use is a key part of the pitch we may ultimately be making to Oculus and/or independent game developers, and I strongly suspect the Balance Board will play a major role in that, but I need to see some relevant demos to be sure. PAGING DR. DIRELIGHT! DR. DIRELIGHT, PLEASE REPORT TO YOUTUBE!

How comfortable is it controlling a character using the foot pedals?

I haven't used them for much other than Mechwarrior 4, but they do an admirable job of that; I'll do my best to give everyone a good feel for their capabilities once I find/make a game that will let me do it. Your point about decoupling actual aiming off into a Hydra wand or "gamegun" is insightful - even if the pedal thing works pretty well, it's not ever going to be as precise as a mouse would be, but once we stop needing to shoot people with our belt-buckles a slight loss of control in turning won't matter. I do think strafing is sticking around, though, and it should - the pedal concept is complimentary to this, as are others.

[Leahy]

Here's an idea just to throw a cog into the machine:

(for seated play) What about using your feet on a wii balance board type device, for WASD movement, and motion tracking your shoulders to turn left/right?

If you are sitting right now, try turning your head far in one direction, while paying attention to your shoulders. After a point (at ~90°+) your shoulders must move to allow your head to turn further. A tracker would pick up on this shoulder movement, translating it into avatar rotation (like pushing gently on an analogue stick left/right).

I personally like that idea, it simplifies the foot controller to forward/back and strafing. I was thinking of using 2 conductive foam pads placed over the back of a chair centered on each shoulder blade to sense pressure. I suppose it shouldn't be too difficult to make but I have no time and all my budget is going towards a new PC.

[sunvsmoon]

[youtube]http://www.youtube.com/watch?v=c8Pm8EJgwhU[/youtube]

[twofoe]

Awesome work, sunvsmoon! Looks like the "foot joystick" idea is really intuitive for games where your character leans to move, like bicycling, snowboarding, or Tribes: Ascend.

[Okta]

Anyone post yet? http://www.thingiverse.com/thing:30354

[sunvsmoon]

Awesome work, sunvsmoon! Looks like the "foot joystick" idea is really intuitive for games where your character leans to move, like bicycling, snowboarding, or Tribes: Ascend.

Not my work unfortunately

[Arni1984]

All this is certainly interesting ... But I, after a hard day, would prefer to collapse on the couch, put on VR - headset and control the body by thought!

[Leahy]

All this is certainly interesting ... But I, after a hard day, would prefer to collapse on the couch, put on VR - headset and control the body by thought!

Supposedly it's doable with the Emotiv EPOC; I am having a rough time learning, so it certainly takes a fair bit of time training to get it to work

[Arni1984]

All this is certainly interesting ... But I, after a hard day, would prefer to collapse on the couch, put on VR - headset and control the body by thought!

Supposedly it's doable with the Emotiv EPOC; I am having a rough time learning, so it certainly takes a fair bit of time training to get it to work

What are the prospects for this we promise ... This matrix! AVATAR!!!
I think that we expect this miracle 10-15 years ...

[Leahy]

All this is certainly interesting ... But I, after a hard day, would prefer to collapse on the couch, put on VR - headset and control the body by thought!

Supposedly it's doable with the Emotiv EPOC; I am having a rough time learning, so it certainly takes a fair bit of time training to get it to work

What are the prospects for this we promise ... This matrix! AVATAR!!!
I think that we expect this miracle 10-15 years ...

Not at all to that degree. The prospects are limited to mapping certain commands that you can teach yourself to reliably replicate an EEG pattern which in practice is a very literal headache to try and learn. I hope in the next 10-15 years we can see more applications of BrainGate but lets get back on topic

[squibbfire]

I dont know guys...this sound way too easy to use and intuitive...I think im gonna pick one up on ebay.

[youtube-hd]http://www.youtube.com/watch?v=ceNWd4H4SxQ[/youtube-hd]

[nanicoar]

Awesome work, sunvsmoon! Looks like the "foot joystick" idea is really intuitive for games where your character leans to move, like bicycling, snowboarding, or Tribes: Ascend.

Not my work unfortunately

If you position the sensor near the rim, is your accelerometer accurate enough to detect sideways movement so you can turn? The readings I'm getting from the sensors on my phone speak of being on the limit of usability, meaning it may be good for a demo but not gaming in general. If my goal is to re-use old smartphones I'll have to do some really smart sensor fusion server-side, just blasting the airwaves with unmodulated UDP.

[WiredEarp]

@ TheHolyChicken: Using the shoulders is a good idea IMHO. Its quite intuitive, and if we get away from using absolute movement and into using the shoulders more like a stick (ie, turn of shoulders controls speed of turn) then I don't really see any problem with it. That would then allow you to use a foot controller to do the forward/backward/strafing.

I was originally planning on using a zero gravity chair and foot setup that would be reclined, not sitting, but I've come up against the fact that it could put your head at a poor angle once you have the weight of a HMD on it. I do have an idea that a reclining position may lead to simulating standing more easily while seated, assuming that your mind can be tricked to believing you are upright when you are not.

Part of the original idea for this involved separately pivotable pieces for your butt and shoulders, so you can twist around while seated. If you hooked up a simple hardware sensor (potentiometer, hall effect, etc) to the shoulder support part of the chair (assuming suitable pivots), then you could use that to easily register your shoulder movements. Same thing could apply to more conventional 'office' type chairs and a more vertical seating position.

[Okta]

I dont know guys...this sound way too easy to use and intuitive...I think im gonna pick one up on ebay.

[youtube-hd]http://www.youtube.com/watch?v=ceNWd4H4SxQ[/youtube-hd]

lol wow that thing is awful.

[squibbfire]

No actually thousands of fishermen use them without looking at their feet. Its perfect. Its stationary. You only need one appendage to use it. Lets compare it to the other stuff.

Disk that requires two feet to use. Well we need something stationary and a disc joystick sounds like it would HAVE to be positioned perfectly. If its a few degrees rotated by accident the user will have to take off the rift and reposition. Once someone's taking off the rift to adjust it looses it value to them.

Rudder pedals. This sounds like a good idea. Stationary and no required position. User can figure out if they are in the right direction with out have to look at them. Downside is that forward reverse will still require forward or tilt movement or both feet.

Locomotion device...Well...they look cool but if we wanted to get that crazy well go outside. Still way too much effort for little reward. Look no further then DDR which showed when you make users use too much effort you get shoved in a closet fast.

Foot/shoes sensors...requires both feet possible complex movements. Can be re adjusted/used without looking.

Single kota foot paddle. Can be used stationary..includes forward movement and tilt. Use of heel buttons. Can be readjusted with out having to look at it. Proven design. I don't see anything else that compares to it. I think everyone is trying to design the next Powerglove or something that is not intuitive enough for them to want to use.

[nateight]

kota foot paddle

I think the idea has promise, with the added benefit of only having to engineer the electronics instead of the entire device. There still may be latency issues, precision issues, or whatever else, but it's definitely worth some experiments. I don't know what your previous experience is but HIDUINO firmware on an Uno continues to impress me, and if you can get raw potentiometer or similar output out of one of these paddles it wouldn't be far to go to have a complete sketch driving an HID joystick in Windows. Check out the Instructable I started from, ask any questions you may have here or in private, and post your results!

[WiredEarp]

The trolling controller looks quite decent, although the video doesn't demo it very well. When he talks about leaning the foot left and right, it looks more like hes pushing forwards and backwards. Also is there a forwards backwards on that thing, or do you need to adjust the 'speed controls' with your foot?

The basic idea is what we are looking for IMHO (I think a small amount of lean left/right could serve to provide a rudimentary 'strafe' control), but I dont think it is quite right. Could be a good starting point for some modification though...

[squibbfire]

The trolling controller looks quite decent, although the video doesn't demo it very well. When he talks about leaning the foot left and right, it looks more like hes pushing forwards and backwards. Also is there a forwards backwards on that thing, or do you need to adjust the 'speed controls' with your foot?

The basic idea is what we are looking for IMHO (I think a small amount of lean left/right could serve to provide a rudimentary 'strafe' control), but I dont think it is quite right. Could be a good starting point for some modification though...

Yes I don't think we need speed control but they have all sorts of types...including a floating pedal. I am going to stop at the nearing marina someday and take a look at what kind of controls they have. Most of these devices are simple closed loop switches so they shouldn't be too hard to get hooked up to a pc.

[shiva]

Im and still shocked no one has look at trolling boat motor controllers for this?

Holy sh*t, this is the most promising idea I've seen concerning VR feet movement
It's simple, functionnal, already on the market, doesn't take space, cheap. It's a wonder no one thought about this earlier
Definitly interested in hearing more from your tests if you choose to customise one

[squibbfire]

Don't worry once I get to Bass Pro Shop..I'm sure they'll have tons for me to try!

[nateight]

Your Arduino-based HID pedal demo, as requested. This is the first time I've ever posted something to Youtube, so please let me know if there's anything I could be doing better. Video quality is poor but surprisingly decent considering I only own a “still” camera; apologies for occasionally snorfling into the mic, but this was a better alternative than the decidedly not-decent mic in the camera.

[youtube]http://www.youtube.com/watch?v=EvtFn73LgDg[/youtube]


So, did that explain everything sufficiently, or are there further questions? Who wants one of these rigs and what would you pay to have one? Should I spin this off into its own topic? Does Oculus need a “Senior Foot Pedal Guy”, or are they counting on third parties to fill in the controller peripherals picture? Would a Kickstarter even be a good way to get these devices out to devs or would we need to think smaller while the Rift is still nominally an experimental product?

[STRZ]

I think it would be cool to have a controller like this asap until other solutions may pop up in the future. First come first served!

Respect for putting up the video, lots of geeky stuff there in the background, even a OLPC

Regarding the Guitar hero pedals, i think Palmer knows the person behind Guitar Hero, he possibly knows where the factory is

If you're still interested in testing the 360° disc control for turning i was talking about (for one foot operation), those pentiometers could possibly work for this (build also into wind direction measurement stuff) http://www.conrad.de/ce/de/product/4240 ... -k-1-W-20-

English data sheet: http://www.produktinfo.conrad.com/daten ... TYP357.pdf

[nateight]

If you're still interested in testing the 360° disc control for turning i was talking about (for one foot operation), those pentiometers could possibly work for this

For continuous rotational use you'd want nothing other than a decent optical encoder. The owner of the Battletech: Firestorm franchise graciously let me borrow one of the most precise encoders available, and it's another thing I managed to hook into an Arduino. Potentiometers are okay for a lot of situations, but I'm still shocked at how much more precise the encoder feels - pots are slow and sloppy in comparison, and in widespread use only because they're passive and somewhat cheaper.

Good eye, by the way. I'd still like to get a little Python script running on that OLPC that constantly updates my email status and the like, but the thing is just so slooooow. If I was a kid in Africa and that was the only computer I'd ever seen I'd seriously wonder what all the fuss was about.

[nateight]

anything

Despite ~50 views according to Youtube's counter, my video has generated one comment worth of feedback (thanks, STRZ!). I don't think a self-bump is out of line considering the effort I've invested. Feedback is important here not to bolster my ego but because putting a physical foot pedal in the hands (or feet) of developers and gamers simply won't happen without a significant group of people literally buying into the concept and building software that supports it, and I have no way of knowing how crazy this idea may be without a chorus of support and/or constructive criticism. I'm hard at work developing a couple of ideas complimentary to the pedals, so within a few days I'll push this video into a new thread about my overall vision for VR control, but for now let's try this once more:

[youtube]http://www.youtube.com/watch?v=EvtFn73LgDg[/youtube]

I don't mean to appear needy or pushy here, but: What do you think?

[MSat]

I'll let you know tomorrow when I have enough bandwidth (don't ask) to actually view youtube videos.

[Aabel]

It's interesting, but it looks overly complex. I am not entirely sure what is gained by such a system as none of the actions you do to use the interface mimic what is happening in game. I just don't see this adding to immersion, but rather becoming a distraction as you try to remember whether to step on the pedal, or move with the stick. Then again I haven't used a system like this, I could be wrong.

[nateight]

it looks overly complex. I am not entirely sure what is gained by such a system as none of the actions you do to use the interface mimic what is happening in game.

This is a good point, but consider the alternatives: "Simple" control in a VR environment means either using both thumbsticks on a gamepad or pegging head motion to character rotation. I can't speak from a position of authority or even experience as I've yet to stick my face into a modern HMD, but neither of these options seems to play to VR's strengths - any competitor with a 2D/3D monitor and a typical mouse+WASD setup would seem to have a sizable advantage in either situation. You're right that the demo does little to sell the idea, though - I intend to spend the next several days building a small game demo that properly illustrates what becomes possible if you decouple aiming from body and head orientation.

Thanks for the feedback - if everyone were too shy to step up and say what they really think about an idea we'd live in a world even more dominated by bad ideas.

[Aabel]

"Simple" control in a VR environment means either using both thumbsticks on a gamepad or pegging head motion to character rotation.

I disagree, I think an interface can be simple and robust. If the interface is simple enough that a 3 year old, a 65 year old, and everyone in between, can come to grips with it and use it proficiently in a matter of minutes at most I would consider that intuitive enough to enhance the sense of immersion and be ready for VR. That's my own personal benchmark that guides my thinking on what our interfaces need to be so that VR can FINALLY emerge onto the main stage of human society. I believe that in order to be successful a VR interface needs to be so intuitive to use that it is nearly invisible, you should never have to stop and ask yourself "how do I use this thing again?".

I really don't think you will get much argument from anyone here that decoupling head orientation from body and hand movement is a must, because it absolutely is!

I look forward to seeing what you come up with in Unity, you are obviously on to something and I can't wait to see how your idea develops more. While I don't know that pedals are ideal for controlling a character, they are undoubtedly an input device with a ton of potential yet to be tapped. Heck even if pedals aren't 'ideal' for character control they could still be good! I am willing to try anything that isn't a keyboard and mouse.

[nateight]

I think an interface can be simple and robust.

Okay, I agree with all of that. One thing that can still sink the Rift faster than a cannonball to a rowboat is a lack of powerful yet intuitive controls, and while I think something like my pedals are a viable alternative to gamepads alone I'm being very careful to reserve any final judgement until I can test them properly. "There must be a better way" is nevertheless not the same as "perhaps this is a better way" - what input scheme are you proposing as superior to foot control + a thumbstick?

It's true I'm a slavering fanboy of the highest order, but I say Battletech: Firestorm is a near-VR experience that still has many lessons to teach us about VR control. It's hard to get a sense of it from the website, but because so much of the VR enthusiast community overlaps with the Battletech community I might be able to describe my reasoning with something approaching brevity. "Torso-twisting" is what a Battlemech does when its operator wants to move in one direction and shoot in another; it therefore bears mentioning in discussions about strafing and head tracking. Anyone new to the Firestorm pod experience is actively discouraged from using the pedals, turning with the joystick and handling the throttle being enough of an initial challenge. This leads to a lot of newbies charging directly towards their targets until they crash together - an experienced pilot in the mix can rack up incredible scores by choosing an agile, lightly-equipped 'Mech, turning the pedals on and tying them to either turning or torso-twisting (leaving the other for the joystick), doing figure eights around these clusters of moshing newbies, and getting points for blowing up one 'Mech AND other 'Mechs caught in the first's explosion. Circle strafing is a similarly powerful technique employed by mouse&keyboard gamers, but I believe it's a technique born of the limitations of 2D screens rather than some tactical advantage. I'm of the opinion that decoupling aiming from orientation and providing fully analog character motion (through foot pedals or whatever turns out to be a better option) represents a similar step up in player control.

In my view, the goal for game developers and VR hardware producers alike needs to be non-exclusionary upselling. No game will succeed by demanding a foot pedal or Razer Hydra or whatever controller be used, but devs should always aim to create games with a high skill ceiling, and each of these various peripherals can (I think) be used to give players ever more control. Once some high-profile titles exist that utilize new VR-centric control schemes, gamers can still join in the fun with gamepads and other limited inputs (and may be encouraged to do so as they acclimate), but upon being thoroughly schooled by a Hydra-wielding, pedal-turning powerhouse, I predict a new form of ragequitting will emerge: Angry VR peripheral shopping.

[WiredEarp]

I dont really see learning a control scheme as being that limiting. I mean, if its just using a thumbstick on a gun, thats as close to intuitive as you will probably get, and learning to use pedals would probably be not much hassle, within an hour it would become instinctive. I don't think we need to get too caught up on the word 'Reality' in terms of user interfaces just yet.

[Aabel]

what input scheme are you proposing as superior to foot control + a thumbstick?

Whether an input scheme is superior or not is going to determine on the type of application first, and user expectation second. There is not going to be one input scheme to rule them all in VR, well at least there won't be for a very long time. The pedal foot control is interesting but we must consider user expectation, most people have lots of experience using pedals every day, in their cars. When people use a pedal they are expecting it to control a vehicle, not their bodies. When it comes to input for controlling a character walking around a game I personally think ODT's will be very strong for this, though they will have some real limitations as their design is more skewed towards simulation than entertainment. I am also a fan of balance boards, they are cheap, elegant and easy to use with minimal safety concern. If you snoop around the web you can find several instances of people using the wii board as a controller for movement in everything from roomba's to World of Warcraft. However neither of those examples show the full range of control that the wii board is capable of capturing.

I am going to have to look into Firestorm. I loved playing pen and paper Battletech back in the day. I agree with you about circle strafing, decoupling head movement, body movement and aim, largely resolves the need for it, however it's what people are used to and they expect it. Simply hearing a control scheme doesn't directly support it is enough for most gamers to simply write it off. An ignorant position, but that's the beast of user expectation. Hopefully the Rift will be a big enough paradigm shift that we can start to slip in new, more immersive control schemes and get them accepted.

No game will succeed by demanding a foot pedal or Razer Hydra or whatever controller be used, but devs should always aim to create games with a high skill ceiling, and each of these various peripherals can (I think) be used to give players ever more control.

I am kind of torn on this. I have a hard time seeing how a compelling VR experience can be developed for multiple control profiles and remain true to the designers intent. Granted I am of course new to the world of VR development, but we've all seen the challenges of designing a game for PC then porting to console and not 'dumbing it down' while keeping it intuitive to play on the more limited control scheme. A part of me wonders if it isn't worth the risk to just try and shove VR peripherals down peoples throats, they will already need to buy a Rift, so what's an extra $60-$70 for a Hydra? While initially the market will be smaller, the word of mouth should be quite powerful and more will follow the early adopters. My wife and partner in development severely disagrees with my view on this and like you thinks the initial applications need to support as many traditional schemes as possible. I've been exploring the VRPN as a possible solution to this but I still haven't wrapped my head around how exactly it all works, or if it's even a solution to supporting multiple control schemes with minimal headache.

upon being thoroughly schooled by a Hydra-wielding, pedal-turning powerhouse, I predict a new form of ragequitting will emerge: Angry VR peripheral shopping.

Oh how I hope this is the case. In firestorm it very well may be that the game is designed to offer players who master more peripherals an advantage over those who don't. However most games, especially competitive player vs. players games do not confer that kind of advantage with anything that deviates from the standard keyboard and mouse. In fact it most likely winds up being a handicap when going up against someone with a keyboard and mouse. Gabe Newell recently had some harsh words for motion controllers, and I agree with his observation that motion controllers are less 'efficient' I don't necessarily see that as undesirable when aiming for deeper immersion and a greater sense of fun. Personally I prefer controls that are, in order of importance: Fun, immersive, then efficient. Pursuing efficiency at the expense of either is a mistake in my mind. However you can sacrifice efficiency if it winds up being more fun and immersive, right up until the point that the lack of efficiency results in the control scheme no longer being fun or immersive.

[mrklaw]

Could you use leaning for motion? Think about someone riding a Segway - they lean slightly and it moves. Dorms teem too disorienting for the rider.

If you could track body leaning, could that work? Maybe for WASD equivalent, and a twist of the shoulders for rotating?

I think maybe a lot of this is about anticipating movement too. A suddenly view movement from a joystick input might be jarring because its difficult to anticipate the motion fully.

Maybe those of you with HMDs could try an experiment - use WASD or a gamepad for strafe movements, but just as you are pressing a button, lean in the direction you are going to move in. Does that help?

[Aabel]

If you could track body leaning, could that work? Maybe for WASD equivalent, and a twist of the shoulders for rotating?

Yes, you can do this with a wii board. You can get all the movements you get with a standard keyboard and mouse. If you use the board as a device to interpret balance and not just pressure you can do some awesome things with it.

[WiredEarp]

How do you turn with a Wii board? lean left/right? If so, then that prevents you using that same movement for strafing.

Also, how comfortable are wii boards for sitting down use?

I dont really see ODT's as having much of a future outside of arcades and simulations. Too bulky, expensive, and tiring to use.nNateight's idea is more for consumer VR.

I thought the video was good nateight, you presented it well. I think its a good idea that has potential, the only problem is there are existing pedal systems already out there, so you would be competing against those. If you could make pedals cheap enough (and that provided strafing) I think you'd have a market though...