Omni: The "Oculus Rift" of 3D Treadmills | Hacker News Reader

Omni: The "Oculus Rift" of 3D Treadmills

kickstarter.com

160 pointsmikeknoop114 comments

The "aha" moment for me, as an outsider to this industry, is using a special shoe to reduce friction and increase lateral stability versus only focusing on developing a 3D treadmill itself.

Doesn't it push your feet towards the middle since grooves in are radial?

Indeed. I wonder how you strafe.

See ~3:30 in the video for strafing.

Note that in real life you almost never strafe. Just look at any footage ( or a movie ) of soldiers/police. No strafing. Games made us believe that we need it.

Untrue. If wearing a vest, taking a corner involves side-stepping, and in any urban combat situation you'll spend a very high proportion of your time doing that.

Source: Urban combat training, FDF.

Of course in reality you move in every direction, even perpendicular to your direction.( Omni supports side_stepping.)

I define strafing ( in games ) as running sideways at the same speed as running forward ( as in most shooters ) , which is not done in real life.

Except real life uses partial strafing all the time. You don't run in the same direction your head, shoulders, or gun is pointing and people can dodge or rapidly change directions. Which is why strafing feels more natural than being forced to turn all the time.

True, it's really the fact that aim/view and walking direction are tied together in a typical game that puts so much emphasis on strafing. Any 'Mech with torso twist is closer to what we'd see with this device.

>Note that in real life you almost never strafe.

That's fine and good. But in games, I strafe all the damn time. If you're playing some kind of FPS game and you aren't strafing you're going to get your ass beat thoroughly.

I also strafe all the time in games that have spiral staircases. Much faster that way.

I completely agree, I was very excited until I saw the special shoes requirements. While I can see how it works, I don't really understand why they couldn't use a regular treadmill belt with some piezoelectric sensor to detect lateral movements.

Because something capable of achieving that currently looks like this:

http://www.youtube.com/watch?v=d-8eVcN2z3k

http://www.youtube.com/watch?v=msci440q18s

Think about it. How do you accomodate movement along multiple axis simultaneously using treadmill technology in a compact, affordable, game-room-sized package?

There's also this one: http://youtu.be/nQR49JGySTM

That certainly looks expensive.

It was. It cost well over half a million dollars, iirc.

> I don't really understand why they couldn't use a regular treadmill belt with some piezoelectric sensor to detect lateral movements.

Other than the obvious reason as they say, no moving parts(min 4 motors, moving belts etc) How would this work? How would one get coverage on a cone shaped walking zone, sounds to me like an engineering nightmare?

One of the design goals was "No moving parts". (And was done to keep costs down and a more maintainable/repairable/defect-free product)

I'm still excited; I think it's the type of innovation required to bring something like this to a mass market for a reasonable price.

price is an important factor as well. try finding a treadmill that supports lateral movement for $300 USD.

On the omni website it says, "Order on Kickstarter", but isn't kickstarter trying to remind people that it isn't necessarily a store? And so if you order one on kickstarter is it a guarantee that you will get one or is there a chance that it won't happen?

Yes Kickstarter is constantly trying to explain to people they aren't a store, but most hardware projects use it as one anyway. And yes, if you buy on Kickstarter there is no guarantee you will ever receive anything. There have been fairly high profile Kickstarter hardware projects that never delivered. Having said that, this one looks like a pretty good bet to ship, the makers are very engaged and obviously have working models that are pretty far along. Granted, mass production is harder than most people assume, but I'd bet on this one shipping.

Another red flag is that the "Risks and Challenges" Kickstarter mandated doesn't actually disclose any risks or challenges, just generic statement that could apply to literally any manufactured product and a bunch of puffery about how great they are.

Too true...

I've yet to see a decent and honest 'risks and challenges' section for a project, they all end up talking about how great everything is.

They are far removed from the 'risks' section of a financial prospectus (which still get ignored!) and yet kickstarter backers have fewer rights and safeguards than a real financial investor.

They also invest much less. I'm not sure why you would need a safeguard for investing $429. Clearly you can miss the money since the starter clearly states you won't get the project until its done, and that's a minimum of 6 months away.

If the project never gets done and no one receives their reward, what's the harm done?

And on the kickstarter updates they state 'We have now sold close to 2,000 Omnis'. Kickstarter pledges aren't sales. They should not be describing them as such.

This is a great solution to the problem, and I'm really excited about it.

But there's one very important catch: The "special shoes" are not an option for people with unusual orthotics requirements (e.g. me), because they're not going to be able to make shoes suitable for everybody.

They have to create soles that can be strapped on to our existing shoes.

Judging by their prototypes (and assuming it does catch on) that should be pretty easy for them or someone else to offer. They're already letting people build the support band themselves based on drawings, I don't see why they couldn't (or someone else) release drawings and a bill of materials for the soles, esp. if some of it could be printed.

A better alternative to printing would probably be to hand-cut (or CNC) the shape of the sole out of a thin sheet of whatever plastic is appropriate.

As it's going to be used only in controlled circumstances, they don't really need to be actual shoes, something like a sandal would be enough, and those should have fewer fitting issues... and it could also accept standard inner-soles for a more comfy feel. A larger size without an inner-sole could indeed probably just be used over normal shoes.

Deleted comment

The parent wasn't "spreading nonsense about serious medical problems [they] know nothing about," they were talking about the shoes in video and how they could perhaps be modified to be more comfortable or applicable to other shoe designs.

Your medical issues sound serious and I'm sorry they cause you excruciating pain, but please don't assume people know of and/or were referencing your particular issues before you specified what those issues were.

I'm pretty sure @nknighthb is thinking about something along the lines of strap-on overshoes, such as http://www.amazon.com/Stabilicers-Original-Heavy-Traction-Cl...

I've tried both the rubberband type and the sandal-type (linked to above) and there are challenges to both. The rubberband type fit better, but are hard to put on (particularly when sitting in a driver's seat in the cold - but this wouldn't be as big a problem for the Omni). The sandal-type aren't secure enough, so they flop or slide (this would be a very big deal for the Omni).

From the looks of it, their setup won't be able to accommodate people who are overweight -- not everyone will be able to fit inside that ring. But I suppose that isn't their target audience anyway...

They say the belt is adjustable. Given enough leeway one might be able to use this unless one's downright obese, in which case it is probably not that good of an idea to put oneself through such paces.

Screw the flashy shooters, I would love to hike across Skyrim, although horse riding will definitely feel uncanny.

It seems that the grooves pointing towards the centre of the platform make for an unnatural walking gait. At the start of a step, the foot will move inwards towards the line of travel. As the foot crosses the torso, it will move outwards again.

Whereas in a natural gait, each foot moves parallel to the line of travel.

See 2:10 of the video for how the shoes fit into the grooves using plunger pins, and how they move on the platform.

I thought about that too. I can imagine a whole new category of back and knee problems in the teens who grow up with this or similar devices.

Note that at no point in the movement the torso moves up or down, so the forces on the joints/back are much lighter. So wouldn't that make the movement a light unnatural movement like swimming is? Do people get many knee injuries from swimming? (I had a knee injury that flared up when swimming, but I don't think it caused it)

I'm really looking forward to the next generation of consumer gaming VR once all the kinks have been worked out.

With any luck, the problems and shortcomings of the Oculus Rift and the Omni will be solved within a couple years, and we'll be able to buy a fully usable VR rig for something like $1500 (glasses, treadmill, handheld controller). And we'll have games designed specifically for it.

Hmm, unless I'm missing something it looks like the full cost will be around $900 ($300 for Oculus Rift and $600 for Omni). I definitely agree things will get better naturally after a few years, but if this thing works solidly I will probably bite the bullet and drop the money. I need a damn way to exercise, and this would be FUN.

Right but the point here is, that you dont NEED a specially designed game. if it supports 3d and keyboard/mouse, it supports omni and oculus rift

How about a leapmotion to detect where your hands are so you don't even need a controller at all?

This is a nifty product, and a surprisingly elegant and low-tech solution to the problem of how to allow free movement while in VR.

I had beers a week or two ago with the project founder (really friendly guy!) and he's looking for software developers to help create an awesome SDK. Get in touch with him if you're interested--it's a really cool project!

Supplementary: The 3D Printed Occulus Rift killer.

http://www.durovis.com/index.html

Good luck with that. What makes the Occulus Rift important is how fast it translates real world movement to signals software can process to give a true VR feel. The accelerometers and magnetometers in phones are decent for their intended purpose, but their intended purpose is not VR. This is no Rift killer.

I had the opportunity to ask Palmer Luckey (the founder of OR) directly about the possibility of smartphone based HMDs. It turns out he's looked into it quite a bit, but ultimately the unsolvable latency of the sensors made him move on. It could work for viewing 3D content or for situations where the head will only be moved fairly slowly, but it's not a great solution for most applications. I would note that a dedicated handset maker could create the hardware and drivers which fix this.

I highly appreciate your feedback. Especially that the founders looked into that possiblilty, just shows how much class they have. Carmack doesn't support them just because they're cool. There is more to it.

That's speculation. Do you have factual evidence to prove that current accelerometers aren't responsive enought? Like a datasheet, a publication or something similar that isn't based on gut feeling. No offense, I appreciate the Occulus Rift as a high tech device that sets new limits in VR Technology, but Smartphones have come a long way sir.

Are you saying that the iPhone4s, iPhone5, Galaxy S3, Galaxy S4 and Google Nexus phones accelerometers are really unresponsive? Btw. just in case your theory is right (I admit when I'm wrong), then please checkout how long that theory will have legs ;)

  * Quantum Limit, freaking yeah!!

http://arstechnica.com/science/2012/12/an-accelerometer-so-g...

See my sibling comment to yours for more context. There are two main problems with the sensors in current devices. The first is that the communication layers (hardware and software) between the sensors and your applications can add 60ms or more. The rendering loop itself will add another 16ms (60fps) or 33ms (30fps). The second problem is that they fast, but not very fast. Generally phone position sensors can run 100 times a second. The OR sensors, on the other hand, run at something like 1000 times a second. This lets a processor in the goggles themselves do predictive algorithms which guess at future head motion allowing the higher level software to compensate for expected head motion.

The key is that humans can perceive latency around 20-30ms, and the phone is delivering around 80-100 in the best case. For what it's worth, I did real measurements on all this before giving up on the idea myself.

I just looked through the Datasheet and I have to correct my post. It HAS DATA about the latency, but I don't fully understand it. It says something about 2000 times a second.

Please open the PDF on page 14, chapter 2 Module specifications.

Latency is measured in seconds, not Hz. I can film the sun at 2, 2 thousand or 2 million images a second, but when I do that from earth, the latency will never (1) be below 8 minutes.

(1) nerd remarks: yes, it will eventually go below it, when the sun expands to become a red giant. Also, earth might be knocked out of its orbit earlier, or maybe we move it, etc.

hmm.. I know about the humans limit, that's important to begin with. I don't know details about the latency. My busted post doesn't have that backing, but the sensitivity of 16G is enough for VR applications.

OK, I just busted your theory. Sorry. :) It has 16G

The LSM330 has an user-selectable full scale acceleration range of ±2 g/±4 g/±6 g/±8 g/±16 g and angular rate range of ±250/±500/±2000 d.

Samsung Galaxy S4 accelerometers/gyro Datasheet: http://www.st.com/web/en/catalog/sense_power/FM89/SC1448/PF2...

Here's a FULL Teardown of the insides of the Samsung Galaxy S4. http://www.chipworks.com/blog/recentteardowns/2013/04/25/ins...

Even if there were a single phone that outperformed the Rift in real world responsiveness (and there isn't), that wouldn't really change anything because the whole value proposition of a generic smartphone printed solution is you use it with whatever phone you already have. If you have some phone other than that super phone, as the majority of smartphone users would, you then have to buy some specific $650 phone for the thing to give you comparable performance to the Rift, and you just spent more than double what a Rift costs.

Who said that your ordinary iPhone4s or iPhone5 hasn't comparable, equal or better sensors? Am not here to speculate and I know it's tedious to prove everything, but I welcome it in your next post :)

Everyone benefits, if one bright mind comes up with a solution or a fact that others have to live with.

Note that the attributes of the phone hardware sensors are important, but not sufficient—generic "use any phone, no root required!" solutions like you mentioned will have to send the data through a full stack of software layers that are very unlikely to be optimized for high-rate / low-latency uses like this one.

You might get lucky and have a phone so incredibly powerful that even non-optimal software stacks are fast/low-latency enough to work well, but I suspect it will be a long time before most people do...

I risk the dangers of writing to a 10days old post..

You can use LLVM and CLANG to write the sensor tools, bypassing the entire "Stack of Software", could result in sub-millisecond reaction time.

I would love to hear a medical professional opinion on contusion probability with extensive use of this thing.

Also they didn't show if it's comfortable to run backwards - a critical move in any first person game.

While running backwards is a popular tactic in FPS games, it's one that is obviously completely unrealistic. I think given that this is a virtual reality device, it's more likely to be used for more realistic games that wouldn't require backwards running.

If it's a fun and intuitive device, it seems likely it will be used for all sorts of games, even those that are completely unrealistic...

[My impression is that the running backward technique is profitable largely due to limitations on user control and enemy AI, and those are probably not going to go away any time soon (even if developers might really want to make a game that feels realistic, it's, well, hard :)...]

If by realistic games You mean "Jogging In The Park Simulator 2013" then i guess it's perfect. This could be a nice device to try out on some gaming expo but i think most gamers will be disappointed with the lack of actions they can perform in it.

Yeah, strafing and walking backwards at least you can do. How would you crouch and go prone for example? And I wonder how good you would be able to aim with Omni, that's gonna be a deciding factor.

They mention Kinect integration - that should be able to detect crouching movements.

But would the belt allow you to crouch? It's designed to keep you from falling, so I assume it's there pretty tight.

Maybe a ducking motion?

This isn't intended to be the Holodeck - but it's a step in the right direction.

For that matter, how would you represent a jump?

That said, I'd use the hell out of it for a game like Skyrim, or other open world environments. And it can be so much more than that, .e.g. ocean/space environments for instance.

The gameplay is secondary - there are a lot of motions that can't be expressed well using this device (although with the kinect, you can unlock a lot of arm motions - good for sword play, I'd wager) - but it's superb for walking in fantasy environments.

> For that matter, how would you represent a jump?

You just jump, I think there are videos showing that. I cant jump like the hero in the game but you can do a little hop.

Given sufficiently low friction, there should actually be very little force exerted on the belt. Some sort of larger harness might be an option if it does become an issue.

A diaper-like seat would be a nice addition. With that kind of support user could be waving his legs forwards, backwards and even sideways with little to no problem.

Then all we'll need are astronaut diapers and, oh wait, things get weird when such diapers are involved http://abcnews.go.com/TheLaw/astronaut-love-triangle-attacke...

How is this 3d? You can only go in two dimensions. It doesn't generate obstacles or even hills...

Don't get me wrong, it looks cool. I just don't understand why people are using the term 3D.

I think there may also be a market for it in health clubs as well, they are constantly looking for new gizmos to entice new members in. Also, I see people at the gym try to do lateral movement on the treadmill by going sideways on it so I think people are looking for more than the boring 1D experience of a simple treadmill.

I could see how in the future instead of "going to gym", we simply "play games" to exercise, and it's a lot more fun to do it, too!

Isn't that how most people have exercised for years?

"...to exercise"

Yes, people have played games instead of going to the gym, but I wouldn't call that "exercising". I guess you could put Wii/Kinect in this category, but even those are pretty static compared to this. You're actually running on this.

I think he may be thinking of more traditional forms of games. Soccer, Lacrosse, etc.

You're actually running on the football field or lacrosse pitch; actually swimming on the water polo field. You're actually throwing a Frisbee around with your friends.

With regard to the Omni as a way to exercise, I wonder how it stacks up to actual running/walking since the user does not actually have to propel his or her own weight.

It must be somewhere in between. But definitely closer to running than walking. When running most of you effort is directed against the gravity, that is why the treadmills work.

Anything would be better than sitting.

I'm dying to see what effect something like this will have on the obesity crisis. We could end up with a super fit generation of game players.

I wonder how running/jumping feels, since there are a lot of games where your character running/jumping skills are far better yours in real life.

From the "Oculus Rift" demos I've seen, it seems people already feel weird by the character height, so I guess it might feel a little awkward to run twice as fast as expected.

Nevertheless, it sure beats being sitted.

Ugh. This will eventually destroy my knees or hips someday. It's bad enough that I go running an hour every other day. I just can't imagine doing that for 6 to 8 hours. Although, I think this will be awesome for out of shape people who love games. Might be the thing we need to goad these people into shape!

Actually no.

Most of the problems associated with running come from the unforgiving impact of your foot on the ground (often a hard surface). Devices like this and other treadmills are much better if you have joint problems.

Mildly better.

An obese person should not be running in any case, even if they're able from a cardiovascular standpoint, the impact that running & jogging has on joints is unacceptably severe until a person reaches a normal weight range - even the ubiquitous-in-the-US 'Overweight' status is pushing it.

Ellipticals, cycling, walking, leg lift machines, etc are much better.

I could definitely envisage some heavily inflamed tendons and soft-tissue from the weird crouch/walking motion though.

If this becomes an issue of picking your evil, I'd prefer the devil that allows me to play extremely immersive games for hours while maintaining a high quality fitness level.

> It's bad enough that I go running an hour every other day

Well you're in luck. Running at sub-elite levels has not been correlated with osteoarthritis.

http://www.ncbi.nlm.nih.gov/pubmed/9489830 http://www.ncbi.nlm.nih.gov/pubmed/22632690

Well it sure as shit trashed my knees and I'm sub-sub-sub-elite. Osteoarthritis is not the only potential problem.

Is it possible your bad knees make running unpleasant and not the other way around?

Does your condition have a name? I've not seen claims of long-term injuries (as opposed to acute injury) that have names other than osteoarthritis.

Thanks.

I see that X, the Y of Z is still going strong as an elevator pitch.

What's wrong with "Omni: a 3D Treadmill"?

Because people would ask; "Why do I need a 3d treadmill? I am perfectly happy running in a 2d line." It is not immediately obvious what it is for.

I think it would have make a lot more sense to say "Omni 3d Treadmill: The perfect complement for the Oculus Rift."

Agreed. I guess I'm just one of those curmudgeons who finds piggybacking on other brands distasteful.

But life isn't a tastefulness contest. I suck at it. (And I'd suck if it was).

I can't comment on why it resonates so well with people, but I noticed that the same Kickstarter was posted earlier today using the original OP title "Omni: Move Naturally in Your Favorite Game" and it received hardly any attention.

Yea that title is pretty bad. "Move naturally in your favorite game" could mean any one of millions of things, when the most important aspect of it is that it goes naturally with an Oculus Rift.

I think they meant to say "Omni: a 3D Treadmill for Oculus Rift"

A 4D Treadmill would be crazy awesome, just imagine that! Hey or 6D, altough I don't know how 6D would be like xD hehehe.

Omni + Oculus + kinect + horror game = shit a brick. I for one can't wait.

The problem for me is that I will probably have a dog by the time all these come out.

If the dog were to brush across the back of my leg while I played some terrifying HL2 mod or other it would be kicked clean across the room.

Watch the first couple of videos.. It's pretty clever and cool. Doesn't look like it'll be fat friendly. All that aside, let's pretend like we're responsible adults for a second. The video, especially the end of the first, sort've makes me think this might be the "line". People have talked about violence in video games, and realism, and for the most part I don't think there is any real issue today, it's all malarkey. Something like this, however, changes the game. It's suddenly a very different experience when you have the VR goggles on and the VR running board and you gun down civilians in CoD, or kill XYZ in random FPS.

tl;dr. Does the omni + oculus rift increase realism enough that violent games require extra caution?

There's some dejavu for you. I remember this exact same discussion when "Wolfenstein 3d" hit the scene. That 3d killing of Nazis was sure to be "the line".

The secret jumped out at me when I read that drone pilots were getting real, bona-fide combat PTSD just from pushing the button while looking at those grainy monochrome heat images. People (almost all of them) know the difference between real and imaginary. "The Line" is in us, not on the screen.

One of the documented side effects of virtual reality is a mental disease called derealization , which makes people feel like real life isn't real, and causes a lot of anxiety.

Also there's the fact that military training conditions people to kill automatically without thinking and the fact that virtual reality has a proven ability to decondition people with PTSD or fear of spiders(with real results in real life), which is basically the same psychological process of military conditioning.

All this raises interesting questions about "the line" between real and virtual , that at least be tested before wide scale deployment of VR.

Where is that documented? 'cause that's a pretty big claim.

Derelealization:

http://www.ncbi.nlm.nih.gov/pubmed/20712501

Efficacy of VR exposure therapy in anxiety disorders , in cluding PTSD:

http://en.wikipedia.org/wiki/Virtual_reality_therapy#Efficac...

Exposure therapy probably works via habituation or conditioning.

http://www.nelsonbinggeli.net/NB/CBT-Exposure.html

The knowledge on how military training works is found on dave grossmans's book : "on killing".

I think this is an interesting thought, but I would probably say no. I don't think that just because the Omni makes violent gaming more realistic that it changes its effect on players.

Is it just me that can't see the videos? On Chrome with Ghostery disabled, I can play it and hear the sound, but the image remains a black screen.

If you integrate it with Google Streetmaps, I think I will be able to go jogging in a different street every day for the rest of my life. Woah!

Looks fun, but seems a bit gimmicky to me. Why would you use something that is extra physical effort to use and makes you worse at gaming? I presume people will only really play games against other people using these, which means that the community will be very small.

On the other hand, I could imagine these or something like these making the gym more fun as a fitness thing rather than a gaming thing.

> Why would you use something that is extra physical effort to use and makes you worse at gaming?

Fun.

Did you miss the Dance Dance Revolution mats, the Guitar Hero guitars, the light gun games, DK Bongos, etc etc?

Most of those were for playing games that you couldn't reasonably play with more normal controllers, or at the very least you would be worse at with more normal controllers. There may well be games like that for this device too, but they aren't FPSs which is all anyone ever seems to talk about when describing it.

I think something less gamey and more second-lifey might have potential, but I don't think these will make much headway with hardcore FPS gamers.

Also lots of people only play single player games.

So imagine you're playing half life or something and then you realize that you have to go way back over half the level because you've forgotten something. Are you going to walk? Or imagine you get to a difficult part of the level and you realize that this would be so much easier with a keyboard and mouse. How many times will you retry before you give up and go and sit at your desk?

> So imagine you're playing half life or something and then you realize that you have to go way back over half the level because you've forgotten something.

For long and boring walks you do the same as in any MMO. Hit some kind of lockbutton or something.

> Or imagine you get to a difficult part of the level and you realize that this would be so much easier with a keyboard and mouse. How many times will you retry before you give up and go and sit at your desk?

Well the same argument can be made for playing a game at the lowest difficutly mode. If I like a game and enjoy playing I want to beat the game on the highest possible difficulty. Of course you die often but it is also fun. Nothing is more boring then a shooter where you are never challanged.

Challenge is not about having frustrating input devices.

People want to play a challenging shooter, but nobody goes out to get the worst mouse they can in order to make it extra challenging.

Well people will use a good mouse and the will want to use a good treadmill. If your input divice is really frustrating, say a old school mouse with a ball in it, then you will just stop playing. If you have a treadmill that that is just frustrating and inaccurate you will also stop.

But as soon as a new mode of input is good enougth, people have fun doing things with these input modes even if there is a more efficent way to input. People play with controllers even if it is clear that a shooter can be played better with a mouse. People like to play wii because the like hitting the air instead of moving the finger to the button.

Its fun to use a wheel to drive a car instead of a controller even if the controller works much more accuratly. The same goes for the treadmill, running around and shooting stuff can be more fun then sitting and shooting stuff better.

Also nobody belives that everybody will always use a treadmill, just as nobody belives everybody will use mouse, controller, wii remote, kinect or anything else. Some people only like the mouse/keyboard other people have all the above.

I think lots of people will have fun with actually walking around and moving the same distance in the game. Also people will have fun when the look to the left while keeping there gun pointed at some door.

Having another independend multidirectional input will make it harder if you are untraind but it can improv the max skill. I remember playing Goldeneye on the N64 where your gun was always point strait and you had to switch between walking and pointing the gun. Then it changed and you could do both at the same time. Now we will have another such change, your head, your body and your gun all move independently and that will be fun.

Great device! I can't wait to try one out.

Reply on news.ycombinator.com