Support NeoGAF

[kenmid]

I recommend these to anybody trying to cover the black cable. The cable was driving me so I had to do something I found these on Amazon and they great and you can paint to match your wall.
http://www.amazon.com/Wiremold-CMK10-Mate-Cord-Organizer/dp/B0015EA3P2?ie=UTF8&psc=1&redirect=true&ref_=oh_aui_detailpage_o02_s01

 

[Zaptruder]

Holy crap I finally figured out a work around for UE4 roomscale + traditional movement problem.

Here are the results.
https://www.youtube.com/watch?v=Q1uwQhzCcns

For those that don't know UE4 doesn't allow collision to work on characters other than the root. Since Roomscale has you walk away from the root (center of the room) collision would not work correctly because the capsule doesn't follow you. I got around the problem by not using collision at all, and instead using capsule traces and tiny teleport distances to move the character smoothly. If a capsule hits something it wont let the player go through it (player will move to the hit point instead of total distance). You can freely walk around in roomscale and use traditional movement.

You can see it working when I go up and down the incline (i am dumb and should have walked into one of the boxes...:/)

What I still need to add:

- Custom Gravity
- hit bounce back (move player back from a hit point so they are not looking into a wall
- teleport the player away if they physically walk into a wall
- try leaning instead of trackpad to move
- add force to dynamic objects that are hit

The cool thing is I can turn this into directional teleporation by increasing the teleport distance for people who can't do tradition movement due to motion sickness.

Well this is interesting. Probably a useful tidbit for me to know.

Currently learning UE4 and C++ so that I can work on implementing my CAOTS idea properly (along with a few other things that I've considered). Sounds like what you've encountered is directly relevant to what I'm trying to achieve.

So this root problem... what does that mean exactly? The collision volume is centered on the room's 0,0,0? If you're to one side of the room, you'll knock into objects only when the 0,0,0 collider hit's them, while ignoring where you are?

Is there no way of moving that collision volume to the HMD position?

 

[Zaptruder]

Developers give out a ton of free copies to the press so can't read too hard into the data in terms of revenue. It is just useful to get a rough idea of how things are going.

How many copies realistically are you giving away to press? In the thousands at absolute most right? Even that is excessive given the amount of work required to give out that many by hand!

 

[Tain]

Well this is interesting. Probably a useful tidbit for me to know.

Currently learning UE4 and C++ so that I can work on implementing my CAOTS idea properly (along with a few other things that I've considered). Sounds like what you've encountered is directly relevant to what I'm trying to achieve.

So this root problem... what does that mean exactly? The collision volume is centered on the room's 0,0,0? If you're to one side of the room, you'll knock into objects only when the 0,0,0 collider hit's them, while ignoring where you are?

Is there no way of moving that collision volume to the HMD position?

In room-scale, the player actor is basically the room. Moving that actor moves the room around the world. This can be an issue, as you probably want the collision volume to be attached to the player's HMD (which is the actor's Camera component) rather than always be in the center of the room.

I think that if you want UE4's sweep functionality to work (an option when updating an actor's location that will move them as much as possible without colliding), the collision checking for that will be based on the actor's root component. The actor's root component has no transform, so it can't be in a "relative" position.

You can still do simpler overlap-based collision checks with non-root components, and I *think* more complex collision events (blocking hits) still work on non-root components (so you'd handle re-positioning on collision on your own). I could totally be wrong there.

You can also specify non-root components to be the "update component" for sweeps, but anything not attached to the update component gets left behind sooo maybe that doesn't help at all.

 

[Zaptruder]

In room-scale, the player actor is basically the room. Moving that actor moves the room around the world. This can be an issue, as you probably want the collision volume to be attached to the player's HMD (which is the actor's Camera component) rather than always be in the center of the room.

I think that if you want UE4's sweep functionality to work (an option when updating an actor's location that will move them as much as possible without colliding), the collision checking for that will be based on the actor's root component. The actor's root component has no transform, so it can't be in a "relative" position.

You can still do simpler overlap-based collision checks with non-root components, and I *think* more complex collision events (blocking hits) still work on non-root components (so you'd handle re-positioning on collision on your own). I could totally be wrong there.

You can also specify non-root components to be the "update component" for sweeps, but anything not attached to the update component gets left behind sooo maybe that doesn't help at all.

Thanks for that info. Sounds like something that they'll hopefully fix given how much they're hyping the VR stuff... but for now, it means the rest of us will have to resort to hacky work arounds?

 

[Tain]

Thanks for that info. Sounds like something that they'll hopefully fix given how much they're hyping the VR stuff... but for now, it means the rest of us will have to resort to hacky work arounds?

I haven't dug into it (my game mostly uses those simpler overlap checks) but it does sound like if you want sweep-like functionality you'd have to roll your own.

It seems like a good solution for Epic, again talking out of my ass a bit, may be to allow multiple components to be set as "Update Components" instead of having just one.

 

[Croatoan]

Well this is interesting. Probably a useful tidbit for me to know.

Currently learning UE4 and C++ so that I can work on implementing my CAOTS idea properly (along with a few other things that I've considered). Sounds like what you've encountered is directly relevant to what I'm trying to achieve.

So this root problem... what does that mean exactly? The collision volume is centered on the room's 0,0,0? If you're to one side of the room, you'll knock into objects only when the 0,0,0 collider hit's them, while ignoring where you are?

Is there no way of moving that collision volume to the HMD position?

I actually already did your CAOTS idea. I can show you a video if you would like to see it working in unreal. I am currently trying to add leaning into it (currently just uses pointing like you described).

UE4's character movement component requires a capsule to be the root.This is fine for traditional games since the player generally doesn't move off the root (unless an animation forces it to). When developing for roomscale the player(technically the camera) , a "child" of the root, can freely move away from the root though. Which means collision and player position are out of sync until the player walks back to the root position.

Think of the root as the center of your chaperone bounds (that is exactly what it is).

If you think about the parent-child link, you can move a child to the position of the parent without moving the parent, but if you move the parent to the position of the child the child will move as well. Since the player camera is a child to the capsule collider (root) you cannot ever move it(the parent) to the position of the player (cam position).

Now, you might ask, "why not just add a collider to the camera"? Well this doesn't work because collision components on children of the root ignore collision with static meshes. Why is this? Fuck if i know but epic is silent about a fix, probably because this is mostly a vr specific issue and they don't really care about vr yet.

Furthermore, you actually want to use a pawn instead of a character for roomscale vr. With a pawn the root is a scene component and not a capsule. Pawns cant use the character movement component though. Which means you can't use the built in movement system.

Thus i created my own movement system that works with room scale 100%. It uses a pawn and capsule traces from the camera to check for hits before moving a characters root (which moves the camera as well). This is a great workaround until epic decides to write room scale specific stuff into ue4 source.

I imagine they will eventually have a special vr controller that fixes all of the current deficiencies. Until then it is up to us.

Probably the oddest thing about the collision issue is that child colliders WILL collide with dynamic actors which have simulate physics on.

If you want specifics i can share my blueprint graph with you via pm.

 

[Zaptruder]

UE4's character movement component requires a capsule to be the root.This is fine for traditional games since the player generally doesn't move off the root (unless an animation forces it to). When developing for the roomscale the player, a "child" of the root can freely move away from the root.

If you think about the parent-child link, you can move a child to the position of the parent without moving the parent, but if you move the parent to the position of the child the child will move as well. Since the player camera is a child to the capsule collider (root) you cannot ever move it to the position of the player (cam position)

Now, you might ask, "why not just add a collider to the camera"? Well this doesn't work because collusion components on children of the root ignore collision with static meshes. Why is this? Fuck if i know but epic is silent about a fix probably because this is mostly a vr specific issue and they don't really care about vr yet.

Furthermore you want to use a pawn instead of a character for roomscale vr. With a pawn the root is a scene component and not a capsule. Pawns cant use the character movement component though.

Thus i created my own movement system that works with room scale 100%. It uses a pawn and capsule traces from the camera to check for hits before moving a characters root (which moves the camera as well). This is a great workaround until epic decides to write room scale specific stuff into ue4 source.

I imagine they will eventually have a special vr controller that fixes all of the current deficiencies. Until then it is up to us.

Probably the oddest thing about the collision issue is that child colliders WILL collide with dynamic actors which have simulate physics on.

Hmm. Thanks for all that! Hope you don't mind if I PM you for notes in the future when I'm at that point

 

[Croatoan]

Hmm. Thanks for all that! Hope you don't mind if I PM you for notes in the future when I'm at that point

I don't know if you noticed my edit but i have done your caots idea in vr. Its pretty simple using my system. I can share my blueprint graphs/ make a video if you wish.

Warning Math! To make it work you just get the linear velocity of the headset and convert that into a vector length. You should set a cap value (max vel) and then divide the current headset vector length by that (will give you a number between 0 and 1.

With that number you can get a desired speed by multiplying it by a max speed.

Now just move the player using this speed in the direction the controller is pointing.

Everything i do is in blueprint right now. Moving from c# to c++ hasn't clicked for me.

 

[Zaptruder]

I don't know if you noticed my edit but i have done your caots idea in vr. Its pretty simple using my system. I can share my blueprint graphs/ make a video if you wish.

Warning Math! To make it work you just get the linear velocity of the headset and convert that into a vector length. You should set a cap value (max vel) and then divide the current headset vector length by that (will give you a number between 0 and 1.

With that number you can get a desired speed by multiplying it by a max speed.

Now just move the player using this speed in the direction the controller is pointing.

Everything i do is in blueprint right now. Moving from c# to c++ hasn't clicked for me.

Oh nice. Yeah, I'll definetly have to hit you up for that offer (i.e. please PM me your blueprint graph) - blueprints is what I'm working in now too... seems pretty powerful. I'm hoping I can implement most of my ideas with blueprints, as that'll save me time from having to actually code (although really blueprints is a form of coding in and of itself).

Looking forward to seeing your take on the implementation - once I'm a bit more up to speed, I'll actually be able to tweak it more exactingly.

 

[Croatoan]

Oh nice. Yeah, I'll definetly have to hit you up for that offer (i.e. please PM me your blueprint graph) - blueprints is what I'm working in now too... seems pretty powerful. I'm hoping I can implement most of my ideas with blueprints, as that'll save me time from having to actually code (although really blueprints is a form of coding in and of itself).

Looking forward to seeing your take on the implementation - once I'm a bit more up to speed, I'll actually be able to tweak it more exactingly.

I can pm you the graph tomorrow after 12 pm cst. Gonna be away from my work comp until then.


Btw, you can actually convert blueprints to c++ now which helps with performance. In case you are interested.

 

[Zaptruder]

I can pm you the graph tomorrow after 12 pm cst. Gonna be away from my work comp until then.


Btw, you can actually convert blueprints to c++ now which helps with performance. In case you are interested.

Cheers!

I am interested, but lacking sufficient knowledge to properly exploit that information for now

 

[Trojan]

FYI for Team Couch Gaming: Corsair Lapdog just released. It's getting good reviews from sites, with the major drawback being price. If you've ever tried to game with KBM on the couch, you know there aren't many options and this feels like the thing I've been waiting for.

http://www.corsair.com/en-us/landing/lapdog?utm_content=&utm_source=VerticalResponse&utm_medium=Email&utm_term=&utm_campaign=It%27s%20Here%21%20Unleash%204K%20Gaming%20In%20Your%20Living%20Roomcontent

 

[zoukka]

Wow the Blue demos absolutely blew me away. Especially the abyss one, the darkness helps to sell the illusion. When you get close to the anglerfish I nearly shat my pants. Kinda missed opportunity to not have sharks or other scary fish there, but most people would freak out I guess :b

 

[Zalusithix]

Briefly started up Elite Dangerous last night as it was the last VR game that I have which I hadn't gotten around to. I think I can safely say that I have little interest in playing it in VR until I have a computer than can handle it at 2x supersampling with the settings I'd like. This has nothing to do with text legibility issue and everything to do with the horrid amounts of aliasing everywhere. TrackIR is good enough for that game in the interim. The gains from VR vs TrackIR are far outweighed by the image quality reduction it brings.

 

[Onemic]

I accidentally started up A-10 and it surprisingly plays without a hitch on my R9 270. Guess I just played my first 'made for VR' game if you dont count AltSpace.

After playing it though, it seems like SPT would be very similar to this. To the people that have both, is this true? If so should I just return one of them?

 

[GuitarAtomik]

I accidentally started up A-10 and it surprisingly plays without a hitch on my R9 270. Guess I just played my first 'made for VR' game if you dont count AltSpace.

After playing it though, it seems like SPT would be very similar to this. To the people that have both, is this true? If so should I just return one of them?

On the surface they seem the same but they're actually really different. A-10 is more about defending the platform you're on by killing enemies before they get too close and shooting as many as possible, while SPT is more about literally dodging enemy fire while taking them down. Another way to look at it is SPT is really fast paced where enemies are always shooting at you, while in A-10 the enemies are a little more numerous and looming and rarely shooting at you. Personally, I found A-10 kind of boring but a lot of people think otherwise.

 

[Plasmid]

Longbow from The Lab is probably one of my favorite games right now. I REALLY hope Valve is developing their own VR games.

 

[MaDKaT]

Longbow from The Lab is probably one of my favorite games right now. I REALLY hope Valve is developing their own VR games.

Me too. This and Xortex I play all the time. I need more of both but especially want more of the bow game. Flesh it out to allow randomly generated or allow custom building, different types of arrows. Maybe add the ability to move to any hex space not on a path. Active tower defense game.

 

[Onemic]

Im probably in the minority here, but I much prefer traditional movement vs teleportation. If given the option to pick only one, I will use traditional movement everytime vs teleporting. I really hope that at the very least devs allow the use of both traditional and teleportation based movement. Altspace does this pretty well. Traditional movement is just way more immersive to me.

One thing I wonder though is why devs arent trying to at least adapt traditional movement to how a real person would move. VR games with traditional movement still feel as if you're floating, which I think is a large factor in people being motion sick. Even though Im pretty immune to motion sickness in VR(framerate drops dont affect me either), I can tell if something feels disorienting. With traditional movement now, it feels as if you're standing in a vehicle due to how floaty the movement is. So when you turn(with your feet, not your head) it feels akin to standing in a moving bus that's currently going into a turn while you are also turning at the same time. Naturally in that situation you would likely fall or at the very least lose your balance, so that feeling of disorientation remains when you do it with traditional controls. If devs would instead make movement a lot more weighty, and make it so that it feels like legs are actually moving rather than it feeling like you're floating, while giving the first person view an actual body(I dont know why in VR of all platforms, devs are still making bodyless first person cameras) it would probably go a long way in adapting traditional controls for VR.

On the surface they seem the same but they're actually really different. A-10 is more about defending the platform you're on by killing enemies before they get too close and shooting as many as possible, while SPT is more about literally dodging enemy fire while taking them down. Another way to look at it is SPT is really fast paced where enemies are always shooting at you, while in A-10 the enemies are a little more numerous and looming and rarely shooting at you. Personally, I found A-10 kind of boring but a lot of people think otherwise.

Ah, Ok. Hopefully I'll be able to try SPT soon.

 

[rje]

Enemy Starfighter is now called "House of the Dying Sun" and it looks rad. Sounds like it should be compatible with steamvr from day 1?

https://www.youtube.com/watch?v=56fCuO7JEy4

 

[PolishQ]

Enemy Starfighter is now called "House of the Dying Sun" and it looks rad. Sounds like it should be compatible with steamvr from day 1?

https://www.youtube.com/watch?v=56fCuO7JEy4

Yusss. I tried this years ago at PAX East on an Oculus DK1. Excited to play the final product on my Vive!

 

[Samaritan]

Enemy Starfighter is now called "House of the Dying Sun" and it looks rad. Sounds like it should be compatible with steamvr from day 1?

https://www.youtube.com/watch?v=56fCuO7JEy4

Sweet Jesus this looks fantastic. Any idea if it'll have HOTAS support?

 

[Zalusithix]

Sweet Jesus this looks fantastic. Any idea if it'll have HOTAS support?

*NOTE: This game supports rebinding controls, but only dual analog gamepads and keyboard & mouse modes are officially supported.

The wording makes me think it's possible. So long as they support DirectInput, everything should be good.

 

[bj00rn_]

Wow the Blue demos absolutely blew me away. Especially the abyss one, the darkness helps to sell the illusion. When you get close to the anglerfish I nearly shat my pants. Kinda missed opportunity to not have sharks or other scary fish there, but most people would freak out I guess :b

Yeah, it's great. Besides Budget Cuts theBlu is probably my favorite Vive experience.

Anyway, Wevr call it the "season 1" of theBlu. I think that is a safe clue leading to more seasons, and I'm 99% sure there will be sharks..

The talent and production value potential within the Wevr network is exciting for future VR experiences. AFAIK a bunch of Hollywood people was involved making theBlu, including the animation director of Avatar.

Here's an interview with the CEO of Wevr, talking about..VR:
WEVR: Where Hollywood meets VR

 

[Felix Lighter]

Longbow from The Lab is probably one of my favorite games right now. I REALLY hope Valve is developing their own VR games.

It is shameful how tired my shoulder gets holding the Vive controller straight out like a bow.

 

[d00d3n]

I started rearranging furniture and was disappointed to learn that I can only free up about 2.1-2.5m x 2.1m for my play area. I know that is is more than the minimum spec (1.5 x 2), but how is the experience with a small play area like this? My hands will be able to swing past three of the limiters (bed, sofa, desk), but one is a wall.

 

[Zalusithix]

Yeah, it's great. Besides Budget Cuts theBlu is probably my favorite Vive experience.

Anyway, Wevr call it the "season 1" of theBlu. I think that is a safe clue leading to more seasons, and I'm 99% sure there will be sharks..

The talent and production value potential within the Wevr network is exciting for future VR experiences. AFAIK a bunch of Hollywood people was involved making theBlu, including the animation director of Avatar.

Here's an interview with the CEO of Wevr, talking about..VR:
WEVR: Where Hollywood meets VR

TheBlu was one of the few experiences where I've felt like it was a true alternate reality instead of an approximation of it. It was, headset limitations not withstanding, as if I were at the bottom of the ocean as some incorporeal entity. The modeling, textures, animation, lighting, sound... damn near perfect considering the limitations of today's technology. It's basically what convinced me that VR has major applications in education and documentaries when taken in tandem with the Apollo 11 experience. Take the goal of Apollo 11 and apply the polish of theBlu. Heck, imagine an interactive VR equivalent of the Planet Earth documentary another decade or so from now.

Even if VR flops dead in the gaming realm, the possibilities outside of gaming are immense.

 

[| Praxis |]

Enemy Starfighter is now called "House of the Dying Sun" and it looks rad. Sounds like it should be compatible with steamvr from day 1?

https://www.youtube.com/watch?v=56fCuO7JEy4

This looks amazing, but no hotas support means no sale. It's a fucking travesty if this comes out and only supports controllers/mouse & keyboard.

 

[Dreams-Visions]

This looks amazing, but no hotas support means no sale. It's a fucking travesty if this comes out and only supports controllers/mouse & keyboard.

no HOTAS support?

RIP.

 

[datamage]

Wow the Blue demos absolutely blew me away. Especially the abyss one, the darkness helps to sell the illusion. When you get close to the anglerfish I nearly shat my pants. Kinda missed opportunity to not have sharks or other scary fish there, but most people would freak out I guess :b

As much as I would love a shark scenario, I'm not sure I'd be able to handle it, lol.

The Whale Encounter videos I saw online had a Great White in it, I had so much anxiety during that scene waiting for it..

 

[Wallach]

This looks amazing, but no hotas support means no sale. It's a fucking travesty if this comes out and only supports controllers/mouse & keyboard.

Really no HOTAS support? Why? Who, why. Fuck. That looked really cool.

 

[GuitarAtomik]

This looks amazing, but no hotas support means no sale. It's a fucking travesty if this comes out and only supports controllers/mouse & keyboard.

no HOTAS support?

RIP.

Definitely a bummer. It's going to have X-input support and advanced control rebinding so I think you can manually configure it though (correct me if I'm wrong someone). Still, I'm hoping official support comes eventually. It is technically an early access release so I doubt it's totally off the table for the future.

 

[Trouble]

This looks amazing, but no hotas support means no sale. It's a fucking travesty if this comes out and only supports controllers/mouse & keyboard.

I was confused by this statement, so I hunted this down on their website.

What controller should I buy for the game?

House of the Dying Sun works great with just a keyboard and mouse, but was designed with a dual-analog gamepad (Xbox 360, DualShock 4) as the "lead" controller. We highly recommend one of those. Don't buy a HOTAS specifically for this game!

It sounds like they are going for arcadey controls. Still, I'd be surprised if they don't add full HOTAS support at some point.

E: Any word on price for it?

 

[Hawkian]

it's heeeeeeeere time to see how the setup goes and if my 780 can muster anything fun

 

[Hawkian]

Is it potentially possible to install your base station too high? Any risk to this?

 

[wwm0nkey]

Got my shipping confirmation!

 

[Mazre]

Is it potentially possible to install your base station too high? Any risk to this?

Not practically. The power cable is a little longer than 10 ft so if you don't want to use an extension cord for power keep that in mind. I think the install guide recommends 6.5 ft and mine are around 7 ft high.

 

[Tain]

Is it potentially possible to install your base station too high? Any risk to this?

as long as line of sight is all good, I doubt it.

 

[Hawkian]

Thanks. The length of the power cable makes it actually a lot more flexible than I expected it to be. I'm going to err on the side of putting them higher up than necessary and just wanted to make sure that wasn't a problem.

Edit: Totally see what you guys meant about the included drywall anchors btw. Just what? Why so cheap? It's mindboggling for such a pricey product considering the good ones are like maybe $0.50 apiece...

Had equivalent ones from HD that will work just dandy though.

edit again: Hmmm, how important is the X-axis tilt? Will be able to test them shortly, just get so paranoid with stuff like this.

 

[Tain]

edit again: Hmmm, how important is the X-axis tilt? Will be able to test them shortly, just get so paranoid with stuff like this.

I'm pretty sure the angle doesn't matter at all, which is neat. One of mine is basically in a 45 degree roll and then a bit of a pitch downward.

 

[Hawkian]

woooooaaah this is crazy shiiiit

 

[Wallach]

woooooaaah this is crazy shiiiit

Let me know if you pick up BigScreen or AltSpace!

 

[Hawkian]

I grabbed both, dunno what they are yet at all but I got em!

Lots to talk and ask about (wow this is seriously wild stuff), but I do seem to be having some sort of glitch or issue. As I move my head the screen seems to jitter or flinch side to side every second or two. It's extremely noticeable in the tutorial, haven't spent enough time in anything else to see if it's noticeable there. Don't really even know where to start to troubleshoot it, if it's in the hardware somewhere, settings, or if this is just my graphics card not really being cut out for this.

 

[jellies_two]

As I move my head the screen seems to jitter or flinch side to side every second or two. It's extremely noticeable in the tutorial, haven't spent enough time in anything else to see if it's noticeable there.

There are some settings available from the SteamVR control window, the little one that shows green icons for each component. In the settings are some for debugging and warning the player about frame-rate issues. Try playing around with those.

 

[pj]

I grabbed both, dunno what they are yet at all but I got em!

Lots to talk and ask about (wow this is seriously wild stuff), but I do seem to be having some sort of glitch or issue. As I move my head the screen seems to jitter or flinch side to side every second or two. It's extremely noticeable in the tutorial, haven't spent enough time in anything else to see if it's noticeable there. Don't really even know where to start to troubleshoot it, if it's in the hardware somewhere, settings, or if this is just my graphics card not really being cut out for this.

Steamvr has settings for showing a message in vr if your computer is missing frames. If you aren't hitting the framerate then it's probably your gpu. Also try plugging the headset into a usb2 port if it's using usb3

 

[Radnom]

Lots to talk and ask about (wow this is seriously wild stuff), but I do seem to be having some sort of glitch or issue. As I move my head the screen seems to jitter or flinch side to side every second or two.

Could it be due to reflective surfaces in the room messing with the Lighthouses? Or perhaps the Lighthouses are wobbling?

 

[Defect]

So I let my sister try it yesterday and she said that she saw a red and blue 3d glasses effect on text.

What does that mean?

 

[mrchampipi]

I started rearranging furniture and was disappointed to learn that I can only free up about 2.1-2.5m x 2.1m for my play area. I know that is is more than the minimum spec (1.5 x 2), but how is the experience with a small play area like this? My hands will be able to swing past three of the limiters (bed, sofa, desk), but one is a wall.

I have pretty much the same space available and I find it way too small. If you count your extend arms in the equation, you end up with barely a meter to move in each direction. I still think 4x4m is golden for a proper rommscale experience.

 

[tokkun]

So I let my sister try it yesterday and she said that she saw a red and blue 3d glasses effect on text.

What does that mean?