VR Game "Armored Ops" Prototype for the Rift

[Owen]

During the kickstarter I was doing some prototyping in Unity to see what kind of gameplay ideas would work well in virtual reality, and I managed to create something that I think has a lot of promise.

I recommend viewing in Vimeo, since youtube downsamples and letterboxes 1280x800 videos down to 720p.
http://vimeo.com/zombieprocess/vrprototype

But here is a youtube version which is also set up for stereoscopic viewing.
[youtube-hd]http://www.youtube.com/watch?v=KjtQwk7zg24[/youtube-hd]

Obviously it still needs a lot of work, as does my public speaking, but I think that there is a lot I can do with this kind of gameplay and would really like to get some impressions.

In particular if anyone at oculus can try this on a prototype I would love to hear how it looks, and what else I could do to make it work properly on the Rift.

From here I will be building my own engine and working on multiplayer. Head tracking will work as you might expect, and I really hope I don't have to resort to a manual centering button for yaw, since the position of the hands relative to your view is absolutely crucial. Movement is controlled tank-style where you go forward by pushing both sticks forward. I think this will be the most intuitive method when you are turning your head freely, but certainly I will need to experiment once I get my dev kit in December.

The overarching game design is still somewhat up in the air, but I am leaning toward small squad based attack/defend missions in urban environments. This should allow for a good mix of ranged and close up action with a lot of content reuse so that the implementation should be manageable.

I am also sharing this in the hope that more people create games along these lines. I think that the Razer Hydra is still undervalued as a near-term input solution for virtual reality, and if enough good games use it I think it might be possible to convince consumers that its a worthwhile purchase on top of the Rift. That could open the door for even more interesting (and expensive) input devices later on.

I am eager to hear any comments or questions. Thanks.

[space123321]

This looks amazing through the eyes of my DIY rift - VERY Intense when the bots head is chopped off and spirals to the ground. Also the sense of depth when the lazer pointers are being used is simply amazing. Outstanding work and I look forward to seeing the progress!

Also - as an owner of the Hydra's, this demo is such a perfect match the technology, and combined with head tracking via the Rift... !

[MrGreen]

Nice work man!

I'm very impressed with the sword combat. Collision detection seems spot on. Do you manage to use your Hydra while standing up? I've heard you had to be like no more than a foot away from the base...

I'm curious as to why you'd wanna write your own engine instead of continuing to use Unity with all the work you've already done?

[IGameArt]

I'd love to test a 2d build of this some time, since i've got a hydra myself, theres really not much out there that takes advantage of the 1-1 tracking.

[defactoman]

What id like to find is someone who is making Razor Hydra in game arm control in UDK so I can talk shop with them. Everyone I see does this in Unity.

[Dycus]

Pretty sweet demo, just doing it cross-eyed! I'll have to throw together another prototype to check this out on. We keep having to give all ours away!

[Owen]

Glad you guys like it. I play in a raised sitting position on a sort of springy adjustable stool I have. Maintaining a fixed forward body facing is important though so standing would probably not be ideal. I guess if you put the hydra base on your head it could work. The fiction of the game will have you as a pilot sitting in the cockpit with equipment similar to the game peripherals

The collision detection was the hardest part, and it really pushes what Unity's physx interface is capable of and required tons of tuning. Even adding a cross guard to the sword made it really unstable, and I am looking to add more complexity. For this I plan to use ODE for a more robust and configurable dynamics simulation, which should allow me to get better results with fewer solver iterations.

Aside from that, I want to avoid Unity because its renderer just isn't that flexible. Using directx 10/11 hardware instancing I can do a whole lot of optimization to make the stereoscopy cheaper. I also want to do some fancy stuff with lighting to improve immersion even more, and Unity gets in the way there. The other things I want to implement Unity doesn't really help much with (usable networking, 3D GUI, dynamic occlusion culling, etc), so really I wouldn't gain that much by keeping it. And the code I wrote is really not tied down, most of it is just matrix math. Plus engine programming is just what I do.

For UDK, if you are familiar with their tools you should have a real easy time doing arm control. They provide all sorts of IK controllers you can use directly. I originally tried to prototype using that, but it was taking too long to figure out and I already know Unity. Its a little harder if you want to use physical simulation, as opposed to just animation. If you want to take an approach like mine you will need to control the character using a motor-driven ragdoll, which UDK also supports IIRC. For specifics I highly recommend the intro robotics course Stanford has on youtube.

[MSat]

THAT. IS. AWESOME!

I'm really interested in the motion tracking and arm representation as this is really something along the lines of what I've been thinking about lately. Maybe an off the wall question, but how do you calibrate the controls for the game?

Watching that video gave me ideas about ways to implement believable sword fighting mechanics despite the lack any force feedback. Perhaps you can make it so that either yours, or your opponent's sword always gets cut short and has to grow back whatever came off. This would introduce elements in the game where advantage or disadvantage goes beyond just your reflex speed. For example, having performance boosts such as blade strength, and regeneration speed.. Being able to throw your blade could also be cool

Picking up and manipulating objects can also account for the lack of feedback by using something like "The Force". Imagine being able to grab a virtual crate in each hand, and smashing them together to kill an opponent

Damn, this gives me another good reason to buy the Hydra.

[Namielus]

Great work Owen! Looking forward to see more of this!

[rhinosix]

Wow. Nice work! This looks like a really promising proof of concept.
I'm surprised at how accurate and responsive it seems to be with manipulating objects and hitting/ deflecting with the sword.

It reminded me how much I want to see a new Bushido Blade for the Rift - running through forests, chopping bamboo, one hit kills.

Are you planning to make an arcade style game? Or a big project with story, different environments, etc?

Just a few games I was thought of while watching your video:

Tsurugi (Blade of Honor):
https://www.youtube.com/watch?v=LLtJxD_jZeQ

Virtual On:
https://www.youtube.com/watch?v=zNzS60AVrO4

Ghost in the Shell (PS1):
https://www.youtube.com/watch?v=fllxBaIe9aA

[Zaptruder]

This is pretty amazing. I love the programmer art/90s VR art with 2012 shaders.

Personally, I think an ideal match to this idea would be Overgrowth - the game of surprisingly realistic furry combat. Swinging around swords and seeing their blood spurt system in action would be absolutely gold.

[SinSilla]

Fantastic proof of concept! I guess using two Novint Xio with a control scheme like that would be quite convincing.

[MrGreen]

Man I wish I could blame Unity for slowing me down...

/goes back to building %$&*! web apps.

[cybereality]

Looking very nice, Owen. I agree that the Hydra has huge untapped potential. I can see a lot of promise in your demo, even with the place-holder graphics. Can't wait to see this more fleshed-out.

Also, I kind of go back in forth between using something like Unity and writing an engine from scratch. On one hand, the all-in-one packages work really well for prototyping and stuff. And you end up re-inventing the wheel for a lot of basic stuff with your own engine. But then, with pre-built engines you always reach a point where it doesn't support some feature you want to do. Its just a matter of if that feature is a make-or-break element of the project. In my case I am planning on using PhysX hardware heavily, and the software PhysX implementation in Unity I don't think will cut it. There are also some other small things that make me want to roll-my-own. But then again, its kind of crazy to think of the amount of development time that has gone into commercial engines like Unity...

[Owen]

Calibration is done by extending your arms outward in a T pose. I use the positions of your hands to create a transformation matrix from an orthonormal basis where your left hand is at (-1,0,0) and your right is at (1,0,0). The robot has its own basis with the same scale, and the matrix just transforms sensor data into it.

Bushido Blade and Virtual On are definitely sources of inspiration for this project. I will have a few 1v1 arena modes for sure. But the main game will probably be something between Mechwarrior and Modern Warfare. Small teams with lots of cover and attack/defend type objectives. Probably just competitive, no story mode.

Ultimately I would love to support as many input devices and tracking solutions as possible, and the Xio sure looks impressive.

Unity is a wonderful 3D prototyping tool, even the best in my opinion. I have been working with it professionally for 2 years. For production it leaves a lot to be desired, but if you are willing to compromise performance and scalability for speed of development and portability then its often a great choice. For mobile especially its usually a no-brainer. But if you want to take your time to make a game with the exact capabilities you want, and optimized at a low level, it can be a net time waster.

[Okta]

This is among a short list of the best VR demos i have seen. Well done sir.

ps, make it into a Mount and Blade type medieval combat genre

[TheLostBrain]

Just tried it out on my diy rift (http://www.mtbs3d.com/phpBB/viewtopic.p ... 255#p81138) as well... looks fantastic!

Love how tall the towers are...felt myself wanting to look up And I second space123321's comment on the gun lasers...sense of depth is absolutely crazy.

Awesome job man! Very immersive!

BTW: Couldn't help but remember Johnny Mnemonic after seeing this... Can you make this a reality for me lol
http://www.youtube.com/watch?v=GmzA2_Ng ... re=related

Even better... make this a reality and you'll give me one of my old teenage dreams lol...
http://www.youtube.com/watch?v=zSmOvYzSeaQ

[zdam]

Hey that is really impressive, well done.

I note you are going to stop using Unity, any chance you could release some of the unity work you have done as a .unitypackage? Perhaps the Hydra control stuff?

Cheers, Adam.

[defactoman]

Someone's actually made a nice package for Unity in the Sixense Hydra dev forums.
http://sixense.com/forum/vbulletin/show ... s-included!)

Give that a shot

[Owen]

I just grabbed some forum code for Sixense, with some modifications to make it behave more like the Unity input system. At some point I will probably package up the robotics code though, since its pretty well generalized.

[BOLL]

Ah, commented on your video on YouTube Owen :D Nice to read about the project here too! I am now subscribed both to the channel and this thread, does that make me a stalker? Uh.

Looks awesome in any case! I especially like how the mech-suit tries to mimic your moves as best as mechanically possible. I have always pondered how realistic a sword-game could be when there is no resistance to your slashes and you can flail around as much as possible. Setting it up like this removes that problem as well, the control scheme ends up being how you would steer the mech from inside! Very smart :D Looking forward to future developments for sure! Cheers!

[AugmentedR]

I've noticed that the swords often hit edge to edge when fighting. This should damage your sword like it would normally, making it a goal to only hit the flat of your blade against their sword.

[Zaptruder]

I've noticed that the swords often hit edge to edge when fighting. This should damage your sword like it would normally, making it a goal to only hit the flat of your blade against their sword.

Wouldn't that increase the chances of the sword been broken, even while decreasing the chances of the edge been chipped/damaged?

[jis]

That would complicate the gameplay for no true value. It will only lead to frustration and decrease the feeling of control you have during the battle. It is much more rewarding to improve the robots IA and animation in itself or improve the visual feedbacks.

[Owen]

I do intend to take that into account somewhat, but it will depend on the type of sword as in the real world. A traditionally made katana can shatter pretty easily with an edge-on-edge impact, while a typical european sword will usually just dull and deform.

My intent is to make this a pretty realistic simulation, within the fiction of course. Proper sword handling technique should be rewarded. Those who lack the patience for it will have other roles to chose from.

[PatimPatam]

Awesome work Owen! I believe something very similar to this should be the default control scheme for most first-person games in VR, where you can basically do whatever you want with both hands: interact with objects, use weapons, cast spells, etc. This could apply to all genres really: shooters, adventure, role-playing or whatever.

The only part that should be different is that we would need to find an intuitive way to simulate walking while seated, without using the hands (or the head). As someone suggested in another thread maybe something like a foot pressure sensor pad could do the trick.

[benz145]

Hey guys, I talked to Owen about this game, you might like to have a look!

http://www.roadtovr.com/2012/09/27/virt ... hydra-1372

I really hope he continues development with this!

[Oneironaut]

This looks stunning! Any progress?

[Libertine]

Heres the Mythbusters episode on slicing/breaking a sword with another sword. I don't think its very realistic based on this. http://www.youtube.com/watch?v=OHSXDbIhViI

[German]

Heres the Mythbusters episode on slicing/breaking a sword with another sword. I don't think its very realistic based on this. http://www.youtube.com/watch?v=OHSXDbIhViI

Katana are made much differently than long swords, it's not a good comparison.

[EdZ]

Katana are made much differently than long swords, it's not a good comparison.

Numerous varieties of sword (Katana, Claymore, Rapier, Viking) were compared against each other. While some knicked the edges of others, the only breakages were caused by snapping from sideways flexing, rather than any swords actually being able to cut each other. Even if you took a modern industrially hardened steel wedge and a bar of mild steel, you're going to need a press in order to cleave one with the other.

[Owen]

I have been doing some engine work for it, with pretty much everything but the weapons implemented in my own system outside of Unity. Since the Rift got delayed I decided there wasn't much point in more demos/prototyping yet.

And with swords it varies quite a lot. I have myself broken a saber clean in two while fencing so trust me it happens. A traditionally made katana is actually one of the most likely to shatter, because its edge is made from a much harder and more brittle steel. Because of this, you are taught to never directly absorb a hit with a katana, and always deflect at a shallow angle with the flat of the blade.

Mythbusters almost certainly used a machine manufactured blade, which are stamped from high-tensile steel. Those will bend easily and do not hold an edge very well, so not very practical. A traditionally made one is almost unobtainable in the US, there are only a handful of qualified smiths and only a small quantity of the correct type of steel is smelted each year, with regulated access to it, and export is illegal.

[EdZ]

IIRC they used a differentially hardened blade of high-quality steel (no need for the extreme folding process if your steel isn't riddled with impurities to start with). Your broken Sabre was likely the result of it snapping from being bent and snapping (the Mythbusters segment has some nice high-speed footage of flexion and springback resulting in what appears to be a clean cleave when watched in real-time), rather than another sword actually cutting through it.

[Owen]

In a real katana the edge actually has a different chemical composition, it isn't just tempered.

And yes a sword can't literally slice through another sword like a loaf of bread, but the impact can fracture it and cause a catastrophic structural failure.

[unsilentwill]

Also they are giant robots.

Watched the video again recently, the arm motion is so amazing, glad I bought the Hydra on sale. This demo is still expected to be my GOTY, just because of what it can do. Speaking of, you probably have but just in case, check out Martijn Zandvliet's work in Unity. So much amazing stuff that'd be perfect for VR, and I think he managed most of it in a week.

[Baristan6]

A traditionally made one is almost unobtainable in the US, there are only a handful of qualified smiths and only a small quantity of the correct type of steel is smelted each year, with regulated access to it, and export is illegal.

I recently watched a TV show featuring a Japanese smith crafting a sword in the traditional manner. Was very interesting to see how, and why they differ from other blades.

https://movies.netflix.com/WiMovie/Weap ... a/70209062

It's worth watching if you can stand the very annoying host.

[shiva]

F*CKING WOOOOOOW
It's amazing to see how quickly the VR community works, we already have the Vireio driver, the VorpX and now THIS, all done without even disposing of the actual dev kit.
I have no idea what will happen when people like you will ACTUALLY have your own Rift. Maybe the world will split in half or something

[vrvision]

Is there any way that i can download this and try it?

[lossofmercy]

This was one of the videos that I really wanted to see after I got my Rift, and it did live up to my expectations. Has there been any new work done for this game?