• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

VR team at Valve as busy as ever - "Lots to share!" at GDC 2015

-SD-

Banned
Good to hear that Valve's VR department is choo-choo'ing along nicely, despite several of their VR folks hopping on the Oculus VR train.

...and yeah, Half-Life 3 confirmed, obviously.

https://twitter.com/AlexVlachos/status/560653536581533697
Alex Vlachos said:
I'm looking forward to my "Advanced VR Rendering" talk at GDC. The VR team at Valve has been busy. Lots to share!

http://schedule.gdconf.com/session/advanced-vr-rendering
Advanced VR Rendering
Alex Vlachos | Graphics Programmer, Valve

Valve has been creating advanced prototype VR HMD's since mid-2013 that are more advanced than other developers currently have access to, and this head start has allowed us to gain a ton of VR-specific rendering knowledge that we'd like to share with developers who are actively working on VR or plan to in the near future.

This talk will start with the base requirements of VR rendering, and it will progress into advanced rendering topics focusing on both performance and visual quality.

First generation consumer HMD's are expected to require renderers to shade over 4 million pixels per frame at a minimum of 90 fps. Due to the wide FOV of these HMD's, each pixel ends up feeling lower resolution than viewing the same image on a monitor and requires better shading algorithms than rendering in non-VR. Higher resolution rendering and higher quality pixels at much higher frame rates than games traditionally target is cause for taking a step back and rethinking many aspects of rendering.

Some topics that will be covered include: efficient stereo rendering, reducing rendering latency, saturating the GPU despite synchronization points, reducing pixel cost for low-priority pixels, specular antialiasing, constrained anisotropic lighting, and other tips and tricks relating directly to VR rendering performance and quality.

Game Developers Conference is March 2-6.

http://www.gdconf.com
 
Yeah, anyone who followed them knew their project wasn't dead or stalling or anything. They, or more specifically, Chet Faliszek has talked at several events last year, they had a successful showing at a Boston Game Jam (which led to that infamous Dota 2 VR headset article lol) and Krejlooc has been in touch with them.

I'm just curious if they're going to make any of their experimental stuff public.

Also, Valve Time pointed it out that Gabe in his AMA said Vlachos is responsible to work VR into Source 2. Now, this doesn't mean he's going to show Source 2 here. In fact, this is probably going to be a "boring" tech presentation with algorithms, graphs etc. At least that's what I think. Just trying to keep expectations in check.
 

Sendou

Member
Yeah, anyone who followed them knew their project wasn't dead or stalling or anything.

What is their project? As I understand it they have no desire to launch a VR headset of their own so is all this research for greater good with no strings attached or what?
 
What is their project? As I understand it they have no desire to launch a VR headset of their own so is all this research for greater good with no strings attached or what?

From everything I've read it seems so. IIRC they're sharing their research with Oculus
 
I know it's impossible, but I wish VR wasn't proprietary. I want to play all VR games on one headset.

EDIT: I'm referring to Oculus vs Morpheus and everyone else that decides to jump in.
 
Can someone explain something to me about Valve's VR stuff.

It seems like they had some of the most advanced VR tech and R&D prior to Oculus, but it was all just internal? Did they actually have any plans to develop some kind of consumer VR?

Oculus repeatedly said, Valve's VR room was their benchmark for quality, and they think they've now reached and surpassed it with Crescent Bay. Valve's VR is incredible from what I hear, but it seems like it would have / will never seen the light of day, and Valve have been enthusiastic about giving advice to other VR companies so it's not like they see it as competition.

As I understand it, the Valve VR room uses a much less portable method of tracking (outside-in with tracking markers all over the room) but it seems to me like if they wanted to, they could have been as big as Oculus is now in terms of developing VR for consumers.

Edit - see below post as a TL;DR version of mine:
What is their project? As I understand it they have no desire to launch a VR headset of their own so is all this research for greater good with no strings attached or what?
 
I know it's impossible, but I wish VR wasn't proprietary. I want to play all VR games on one headset.

EDIT: I'm referring to Oculus vs Morpheus and everyone else that decides to jump in.

Honestly I think for the most part these VR games will be platform-neutral. From what I've seen there shouldn't be any games or experiences that you could play on one device but not another. I think you'll just run into different levels of QoP (quality of presence) depending on what device is currently strapped to your head.

Now, if Sony decides to make PS4 exclusive VR games, nothing we can do about that on PC, same as regular exclusive games.
 
Can someone explain something to me about Valve's VR stuff.

It seems like they had some of the most advanced VR tech and R&D prior to Oculus, but it was all just internal? Did they actually have any plans to develop some kind of consumer VR?

Oculus repeatedly said, Valve's VR room was their benchmark for quality, and they think they've now reached and surpassed it with Crescent Bay. Valve's VR is incredible from what I hear, but it seems like it would have / will never seen the light of day, and Valve have been enthusiastic about giving advice to other VR companies so it's not like they see it as competition.

As I understand it, the Valve VR room uses a much less portable method of tracking (outside-in with tracking markers all over the room) but it seems to me like if they wanted to, they could have been as big as Oculus is now in terms of developing VR for consumers.

Edit - see below post as a TL;DR version of mine:

I don't know. Guess I don't see what's in it for Valve.

What's hard to understand? They're experimenting so they get ready for when VR launches. They also want to make Steam a platform for VR. What's in it for them is to get people on board. They believe in VR, they want it to be good so they're sharing their experience with others to make it better for everyone involved.
 
I don't know. Guess I don't see what's in it for Valve.

I know they're a multi million dollar company, so this is hard to grasp, but they really just want what's best for everyone. They want to empower developers to make the best games possible.
Can someone explain something to me about Valve's VR stuff.

It seems like they had some of the most advanced VR tech and R&D prior to Oculus, but it was all just internal? Did they actually have any plans to develop some kind of consumer VR?

Oculus repeatedly said, Valve's VR room was their benchmark for quality, and they think they've now reached and surpassed it with Crescent Bay. Valve's VR is incredible from what I hear, but it seems like it would have / will never seen the light of day, and Valve have been enthusiastic about giving advice to other VR companies so it's not like they see it as competition.

As I understand it, the Valve VR room uses a much less portable method of tracking (outside-in with tracking markers all over the room) but it seems to me like if they wanted to, they could have been as big as Oculus is now in terms of developing VR for consumers.

Edit - see below post as a TL;DR version of mine:

They really don't care about getting anything out of this. Before Oculus even got kickstarted, Valve had been researching and developing VR for years. Once Oculus came online and Valve saw their potential, they gave every piece of knowledge that had for Oculus to work with. Valve never intended or wanted to build VR physically, they knew someone else would fill that role. They just want to be the pioneers of the field technologically and in their personal expertise, and they want to share that with everyone to make VR the best platform possible.

Basically without them VR wouldn't be where it is today.
 
What's hard to understand? They're experimenting so they get ready for when VR launches. They also want to make Steam a platform for VR. What's in it for them is to get people on board. They believe in VR, they want it to be good so they're sharing their experience with others to make it better for everyone involved.
They really don't care about getting anything out of this. Before Oculus even got kickstarted, Valve had been researching and developing VR for years. Once Oculus came online and Valve saw their potential, they gave every piece of knowledge that had for Oculus to work with. Valve never intended or wanted to build VR physically, they knew someone else would fill that role. They just want to be the pioneers of the field technologically and in their personal expertise, and they want to share that with everyone to make VR the best platform possible.

Basically without them VR wouldn't be where it is today.
I see thanks!
 

Nzyme32

Member
What is their project? As I understand it they have no desire to launch a VR headset of their own so is all this research for greater good with no strings attached or what?

They have a bunch of prototype VR sets, and I imagine they are probably still itterating with those as well. They don't intend to release an actual consumer headset (although they might if they feel they have to).

The primary goal seems to be research, ie to inform their decisions and development internally on all kinds of things they may plan in future. Which is why there is this GDC talk on "Advanced VR Rendering" which is likely incorporated into their upcoming Source 2 engine. They also have plans for Steam itself and it's future in VR, and probably other experimental stuff that isn't public - other hardware / input methods etc.

I don't know. Guess I don't see what's in it for Valve.

Staying relevant, being able to come up with new features / hardware / software / technologies to help continue and maintain their growth / business, staying ahead of any competitors, keeping the PC ecosystem progressing ie the one they are reliant on for business. Gabe gave a talk about this sort of stuff a while back I think
 

Seanspeed

Banned
What is their project? As I understand it they have no desire to launch a VR headset of their own so is all this research for greater good with no strings attached or what?
Well, no strings attached in that they don't seem to be asking for money for it or anything. But if Valve still have the same plans they used to, I think Valve want to be involved in the VR ecosystem, probably doing something with Steam. Maybe they actually go into competition with Oculus, but not the Oculus Rift, but instead the Oculus Store. Would be smart. They can both offer a lot of the same applications, but Steam already has such a huge mindshare in terms of where games(and developers for that matter) like to be with their games that they could definitely pull something off. Maybe even have a completely VR-focused Steam interface, analogous to how Steam do Big Picture mode for use with living room TV's.

And of course Valve could absolutely do games. I know Half Life 3 is the popular thing to mention on this subject, but really, I think DOTA2 might actually be the more immediately obvious game to bring VR to.
 

Krejlooc

Banned
What is their project? As I understand it they have no desire to launch a VR headset of their own so is all this research for greater good with no strings attached or what?

Their project is Steamworks VR, a unifying API. Steamworks VR is to VR headsets what DirectX and OpenGL are to video cards. 1 API that talks to many headsets - developers write one code base and it'll communicate with many different type of headsets in the same way. So, theoretically, when it's all done, I write an application using Steamworks VR, then set build options to, like, Morpheus, Oculus Rift, Google Cardboard, OSVR - and it pops out with different executables for each headset, all from one code. Valve has been working on this API for years already.

Valve wants to be the software behind VR, not the hardware. To develop their software, they've built prototype hardware they have no intention of selling.

When you understand this, their steam universe initiative makes a lot more sense - the Steam Universe initiative is Valve's strategy to get Steam -- and thus steamworks -- everywhere. Linux, Mac, Windows, and in both the office and living room. Soon enough, I am willing to bet that Valve will expand steam to Mobile, too.
 

Krejlooc

Banned
Which is cool, but I'd like HL2 VR to not consistently repeat frames and I kinda wonder if the lack of Direct Mode vsync is due to their API. I get that it's early days though.

Those would be problems with oculus' api not Valve's. HL2VR does not use Valve's API yet, it's just a c implimentation of the oculus api.
 
Sweet baby jesus I can't wait.

from the Boston VR Jam

“[Valve] mentioned offhandedly that they have a Dota 2 VR experience where you see the entire game arena sitting on a table in front of you and can bend down to inspect any piece of the action,” said Jonomf. “They also mentioned a life-size Dota 2 VR experience where you’re hanging out in a lane watching the heroes fight; they said it was very scary.”

would be cool to have it in by TI5
 

Tain

Member
Those would be problems with oculus' api not Valve's. HL2VR does not use Valve's API yet, it's just a c implimentation of the oculus api.

Oh, huh. I think I just assumed the lack of a true full screen mode had something to do with Valve's API, but it obviously must not be related.
 

scitek

Member
I know they're a multi million dollar company, so this is hard to grasp, but they really just want what's best for everyone. They want to empower developers to make the best games possible.


They really don't care about getting anything out of this. Before Oculus even got kickstarted, Valve had been researching and developing VR for years. Once Oculus came online and Valve saw their potential, they gave every piece of knowledge that had for Oculus to work with. Valve never intended or wanted to build VR physically, they knew someone else would fill that role. They just want to be the pioneers of the field technologically and in their personal expertise, and they want to share that with everyone to make VR the best platform possible.

Basically without them VR wouldn't be where it is today.


It's hard for people to realize that Valve is supplemented well enough by Steam to not have everything they do be motivated by directly improving their bottom line. They can do things because they want to rather than because they have to, and that's what this is. Of course, they'd benefit financially from VR taking off, but there's no guarantee it will, and it wouldn't be immediate even if it did.
 
from the Boston VR Jam



would be cool to have it in by TI5

Imagine if they could set it up so that you could spectate from a virtual grandstand with other spectator avatars, like a live virtual sporting event, and have the tourney commentary streamed in to direct your attention.
 

A-V-B

Member
If Valve released a VR device for every game they release, everyone would buy every single one. Even me!

At this point, Valve VR would probably make them a boatload of money if that was what they wanted to do. I know I'd choose their device over the Oculus like *snap*
 

AmyS

Member
Oculus CV1 - Morpheus - Valve VR - Microsoft's AR HoloLens

The VR / AR next gen wars cometh!

Here comes a new challenger!
 

Durante

Member
Oculus CV1 - Morpheus - Valve VR - Microsoft's AR HoloLens

The VR / AR next gen wars cometh!
Valve isn't building hardware. AR and VR devices are fundamentally different, at least for the next decade. CV1 and Morpheus aren't competing in the same market.

I don't foresee any wars in the near-term.
 

AmyS

Member
Valve isn't building hardware. AR and VR devices are fundamentally different, at least for the next decade. CV1 and Morpheus aren't competing in the same market.

I don't foresee any wars in the near-term.

VR and AR aren't competitors, they're very different, but the hardware platforms that power them very well could be.

PC
PS5 vs DirectXbox12
 

Krejlooc

Banned
Well, there is some Valve magic in the Rift anyway I'm sure, since Oculus were consulting with them a lot.

Valve is the one who demonstrated the usefulness of low persistence to oculus rift. John Carmack details how Michael Abrash built a demo specifically to show him why Oculus needed low persistence.

Valve's presence in VR is awesome, they basically act as a neutral independent researcher. They've shared their stuff with Sony, too.
 
Valve is the one who demonstrated the usefulness of low persistence to oculus rift. John Carmack details how Michael Abrash built a demo specifically to show him why Oculus needed low persistence.

Valve's presence in VR is awesome, they basically act as a neutral independent researcher. They've shared their stuff with Sony, too.
That's nuts, I didn't know about the low persistence coming from Valve. It's interesting to see how many things have come together to make VR what it is this time around. Haven't Oculus been open with Sony, too? They at least invited some Sony guys to Connect IIRC.

Also I can't think of anything else, product or service, where companies have shown this level of respect to each other. And I know why it is, and that each company has their own agenda, but it's refreshing AF to have a break from all the anti-competitive shit that usually plagues the tech industry.
 

Nzyme32

Member
That's nuts, I didn't know about the low persistence coming from Valve. It's interesting to see how many things have come together to make VR what it is this time around. Haven't Oculus been open with Sony, too? They at least invited some Sony guys to Connect IIRC.

Also I can't think of anything else, product or service, where companies have shown this level of respect to each other. And I know why it is, and that each company has their own agenda, but it's refreshing AF to have a break from all the anti-competitive shit that usually plagues the tech industry.

Specifically about the low persistence thing - here is John Carmack back in 2013 (while working at ID) explaining how Abrash (at Valve) and their team, changed his thinking on the importance of low persistence - http://youtu.be/Uooh0Y9fC_M?t=54m23s
 
Carmack also mentioned Valve in the DK2 trailer.

Don't really care about VR but I wonder if they are going to show support for Source Engine 2.

At this session? I doubt it. Maybe if they announce something related to it beforehand.

I don't even think this is going to get streamed? Will be probably included in the GDC Vault and shown to public much later.
 
Specifically about the low persistence thing - here is John Carmack back in 2013 (while working at ID) explaining how Abrash (at Valve) and their team, changed his thinking on the importance of low persistence - http://youtu.be/Uooh0Y9fC_M?t=54m23s
JC is such a hero god damn.

It's really weird and awesome to see talks like this from a couple of years back, and then realize all this 'hypothetical' tech is right next to me and is working as predicted. I'm so glad guys like him and Abrash are in this field.
 
Top Bottom