Hacker News Reader

Comment by bglazer | Hacker News Reader

I developed a mild fascination with omnidirectional treadmills a while back. It's such an interesting engineering challenge. Anyways, the large size and hefty cost of these treadmills seems almost unavoidable, and therefore makes them unappealing to a large consumer market. It also seems like this technology is the obvious next step for VR.

So, that leads me to the conclusion that perhaps there might be some future resurgence of the arcade. Except they'll have VR headsets and omnidirectional treadmills instead of Pacman.

Wait, maybe there will be Pacman, but you'll be running.

View on HN

One of the big differences between Oculus and Valve's philosophies is that the latter believes that there must be a 1:1 correspondence between motion in the virtual world and the real one. No doing so kills presence and causes the nausea that people talk about. Having spent a fair bit of time in an Oculus DK2 I couldn't agree more.

Unfortunately, this kiboshes treadmills and things like the Virtuix Omni. At least until we manage to fool the vestibular system as well as we can the eyes.

Serious treadmills are designed so you move around just as you would normally but a small bias is created below your perception threshold that centers you on the treadmill with time.

https://www.youtube.com/watch?v=Af0Skxi4ftw

I remember seeing something about VR-enhanced roller coaster rides which took advantage of this, too -- the path of the virtual track and the real roller-coaster diverged considerably, and (iirc) there were places in which you were travelling "backwards" in the VR world.

> At least until we manage to fool the vestibular system as well as we can the eyes.

If we can trick the neurons on site, then might as well trick it right before it plugs in at the brain...in which case, we probably will have optics covered as well. Thence we can get rid of the headset altogether.

It's possible to stimulate the vestibular system electrically:

http://en.wikipedia.org/wiki/Galvanic_vestibular_stimulation

This is sort of a dumb question, but why develop omnidirectional treadmills instead of omnidirectional footwear?

Or even better; I never understood why ppl were not looking into these http://www.thebabyloft.co.uk/images/products/door-bouncer.jp...

With sensor lights picking up the foot movements. I bet the engineering challenge of giving sensory feedback to the foot is easier than making the treadmills work.

Babies are expensive!

These things were all explored over the last 30 years in the VR space. Body suits, hanging rigs, chairs, caves - you name it someone experimented with it.

I fully expect we'll see a massive re-exploration of all of it in the next 30 years, especially as VR becomes a $100 billion market across consumer/enterprise/govt. Eventually a few optimal approaches for the home will come out of that market process, similar to gaming today: you might have a basic controller, you might buy a driving wheel, you might install an entire flight simulator environment, all depending on how far you want to go to boost your experience.

So people don't run into walls while wearing their headsets I think.

When you make a sharp turn while running (or even when you start and stop running), your body works against inertia. It isn't simulated properly by slippery shoes, but can be on a (large, military grade) omnidirectional treadmill.

Omni treadmill shoes used over a large 12ftx12ft are or so could work. But too heavy.

I think natural walking is more likely to take off. Why abstract it when you can just walk. Sure there will be issues with running out of real world space. But that can probably be solved via some sort of body based gesture.

I say this, after going through the original Crescent Bay demo at Oculus Connect, where being able to walk around even within a 4x4 foot cube of space it was amazing having real world movement match your characters.

I think it's running that opens up a lot of possibilities. That's really not possible in the vast majority of spaces, especially with a headset.

Maybe I'm overestimating the amount of physical effort people are willing to take for video games, but there's some precedent with Dance Dance Revolution for example.

No, I think you are right. You can't play Counter Strike or Team Fortress and casually walk around the map. If you're making a dash for the flag you need to run.

Also, walls.

Not sure how we're going to mimic that without force-feedback bodysuits or wiring directly to perception through our brains. e.g., make me convinced that I am walking or falling or climbing a wall.

It's already established that traditional FPS games don't translate well to VR (it's even mentioned in the article). VR is a relatively unexplored medium and it comes with new constraints as well as new possibilities. Long range locomotion is one restriction that doesn't seem easily solvable.

yeah the problem is you want to somehow augment physical actions. no game is going to take off if all you do is move around at walking speed..

as palmer luckey has said a guy playing COD is not actually capable of doing all the stuff a real marine does, so then the question is how far do you take it?? like if i'm playing an nba game it seems fairly plausible that my 1 foot vertical could become a 3 foot vertical, or that my poor mechanics shooting a basketball where in real life i only sink the 3 pointer 10% of the time, when i'm playing as stephen curry the game interprets as 50% of the time..

i think one of the problems is that what makes video games so compelling and addictive is that they are an order of magnitude easier to get really good at then real life... e.g. sword fighting in a game versus sword fighting in real life.. if games become too realistic in their mechanics e.g. i actually have to master good footwork and how to swing a sword properly the game because too difficult

Self-adjustable difficulty. You might get some slack, that slowly fades away and you get closer and closer to the real thing. I believe it is doable with certain aspects.

I would love to play such a 1:1 movement rendering game. I am not going to swordfight people in real life (well, I don't think I will) but I would love to play to a simulation of such a fight. Force feedback is a tremendous challenge for such a thing though. Other activities I would love to have VR for : wingsuits, photorealistic freefall from space, parkour, gun fights, reckless driving ... Yes, I would get destroyed at many of these activities when playing against professionals but I am not a professional player able to play dozen of hours each day so I don't see that argument as that much of a problem. Furthermore, most people seem to favor offline play where the difficulty is up to the developer. It can even be tuned in real time (cf Max Payne) but I don't think it is a great idea. I replayed to Half-Life 1&2 recently, I have got the feeling that this game would be 10 times more fun with a VR experience : crawl through vent and glimpse across a corner gun first, slalom with the airboat.. The technology is far from being there, but I am already sold. The list is pretty much endless.

well i guess it's a question of how you make the game easier... either you make it easier in the virtual world (i.e. in a boxing game the punches come slower than a real life boxer would face), or you make it easier in the real world where your actions are enhanced somehow (force of the punch would be an obvious example)

yeah force feedback would be really cool but it's unclear how you manufacture those, it's like 4d cinema,, hasn't really caught on right... developers seem to talk a lot about trying to avoid breaking immersion, and maybe that mean no 'force feedback' is better than bad force feedback. but i don't know. makes me think that it might be an all or nothing problem.

The simplest solution to this problem I have ever seen involved you wearing a harness to keep you in place, while standing in socks on a super slippery pad. That allowed you to walk around without an omni-directional scrollable surface underneath. Cheap as hell, and works.

I'll just leave this here: http://en.memory-alpha.org/wiki/Holosuite

Reply on news.ycombinator.com