Interview: Valve's Chet Faliszek On The HTC Vive Pre

[alex davies]

We sat down with Chet Faliszek, the public face of Valve’s VR efforts at CES, to talk about the new HTC Vive Pre VR system.

Interview: Valve's Chet Faliszek On The HTC Vive Pre : Read more

 

[Bloob]

*sigh* Valve / HTC adds a useless camera / room roaming capabilities, and Oculus adds useless headphones and bundled stuff (like remote and controller). I know they are all trying to make VR the best they possibly can, but it would seem much more logical to me to get the core experience of being a 3D HMD for sitting users right first. That's what it will be used for at first for the most part anyway, and the simpler tech would be a cheaper product, which in turn would mean a wider adoption rate.

P.S. I know the features these companies are adding are not useless, but neither do I feel they need to be part of the core experience right now. Nor do I feel that Oculus needs to cover the audio-part when there are a host of great companies dedicated for delivering those experiences, I'd rather they set a specification for VR audio and work together with 3rd parties for that.

 

[sicom]

Room scale VR the standard in 3 years. Do these wackjobs actually believe such drivel? How many people's homes have an entire room they can set aside for VR?

Vive is cool, but it's also doomed to fail. Its cost will be too high for such an incredibly niche product. Oculus is aiming for the high-end sector much more intelligently.

 

[Felixander]

I think room-scale is the way to go, and we need to go it as soon as possible, so the Vive supporting it from the get-go is the right approach. When the ordinary person hears "VR", they think about Star Trek's holodeck, they think about movies like The Matrix and ExistenZ.
Yes, they are blown away when they actually experience a 360° environment for the first time, but the very next instinct is for them to get up and around, and try to grab stuff and interact with this world in a natural manner, just like they've seen on TV, no matter how unrealistic that is. Giving them a gamepad and telling them to sit down and play a game just like they are used to, with only the added bonus of being able to turn your head to look around (which they aren't even required in a "forward facing experience"), will wear them out quickly, and they end up questioning why they spent hundreds of dollars for a headset.
To achieve widespread acceptance, there needs to be more, we gotta bring them as close to their fantasy as possible, or else they grow tired, rather leave the HMD off and just play their games with better graphics, still looking forward onto a monitor, and when room-scale VR finally hits they already wrote VR off as a waste of money. They have to experience the full promise of VR from the start to hook them, to make them follow VR like we do, and to keep supporting this technology.

 

[kcarbotte]

*sigh* Valve / HTC adds a useless camera / room roaming capabilities, and Oculus adds useless headphones and bundled stuff (like remote and controller). I know they are all trying to make VR the best they possibly can, but it would seem much more logical to me to get the core experience of being a 3D HMD for sitting users right first. That's what it will be used for at first for the most part anyway, and the simpler tech would be a cheaper product, which in turn would mean a wider adoption rate.

P.S. I know the features these companies are adding are not useless, but neither do I feel they need to be part of the core experience right now. Nor do I feel that Oculus needs to cover the audio-part when there are a host of great companies dedicated for delivering those experiences, I'd rather they set a specification for VR audio and work together with 3rd parties for that.

You don't feel like these things are necessary, but the people building them definitely feel like these additions are necessary, and there's good reason.

HTC isn't "adding" room roamiing capability. That was the play all along. Room scale tracking was announced the day that Vive was first revealed to the world. The camera wasn't a new addition either. It was always planned, but the early dev kits didn't include the camera. Interanally they have been testing the camera system since day one.

Oculus added its own headphones with built-in DAC for good reason too. Spacial audio is very important, and most people don't have headphones ready for that kind of environment. By including them, it ensures that everyone has a superior audio experience. Sound is half of the experience, so Oculus took this part seriously.

You say that should be focusing on the sitting experience, well I would argue that HTC and Valve are considering that with the front facing camera. It actually helps identify objects in front of you, such as your controllers, or racing wheel, or joystick setup. Keyboard and mouse isn't a great experience in VR, but with the Vive you can at least see them without taking the headset off.

As for the Xbox One controller, what do you expect to use for VR? Games are being programed for navigation with the Xbox controller layout. Another controller style could work, but its easier on developers to have a specific controller in mind.

Getting people excited for the day they can afford something spectacular will do way more for adoption than a half-assed product that many people can afford right off the hop. A limited VR experience will render people bored with VR, which will doom the medium in the long run.

 

[kcarbotte]

Room scale VR the standard in 3 years. Do these wackjobs actually believe such drivel? How many people's homes have an entire room they can set aside for VR?

Vive is cool, but it's also doomed to fail. Its cost will be too high for such an incredibly niche product. Oculus is aiming for the high-end sector much more intelligently.

You don't need an entire room for room scale. That's the whole point of the chaperone system.
Move your coffee table to the side, the headset will map out the open area and you can play within that space.

Many people don't realize that the Vive doesn't require 15-feet by 15-feet. It can use UP TO that much space. It scales all the way down to seated experiences, but the tracking still has benefits when seated, such as the ability to follow your hands properly and the chaperone system can identify the objects on the desk in front of you, such as your peripherals.

In situation that you only have a seated location, you can still use room scale tracking for a racing game where you enter the car by sitting in your chair, as an example.

 

[alidan]

Room scale VR the standard in 3 years. Do these wackjobs actually believe such drivel? How many people's homes have an entire room they can set aside for VR?

Vive is cool, but it's also doomed to fail. Its cost will be too high for such an incredibly niche product. Oculus is aiming for the high-end sector much more intelligently.

You don't need an entire room for room scale. That's the whole point of the chaperone system.
Move your coffee table to the side, the headset will map out the open area and you can play within that space.

Many people don't realize that the Vive doesn't require 15-feet by 15-feet. It can use UP TO that much space. It scales all the way down to seated experiences, but the tracking still has benefits when seated, such as the ability to follow your hands properly and the chaperone system can identify the objects on the desk in front of you, such as your peripherals.

In situation that you only have a seated location, you can still use room scale tracking for a racing game where you enter the car by sitting in your chair, as an example.

it doesnt matter, roomscale vr will die the moment the first person dies or breaks a bone because they are an idiot, than other people will not do room vr because of it too.

room vr will work great with augmented reality, but not virtual reality.

 

[Bloob]

You don't feel like these things are necessary, but the people building them definitely feel like these additions are necessary, and there's good reason.

HTC isn't "adding" room roamiing capability. That was the play all along. Room scale tracking was announced the day that Vive was first revealed to the world. The camera wasn't a new addition either. It was always planned, but the early dev kits didn't include the camera. Interanally they have been testing the camera system since day one.

Oculus added its own headphones with built-in DAC for good reason too. Spacial audio is very important, and most people don't have headphones ready for that kind of environment. By including them, it ensures that everyone has a superior audio experience. Sound is half of the experience, so Oculus took this part seriously.

You say that should be focusing on the sitting experience, well I would argue that HTC and Valve are considering that with the front facing camera. It actually helps identify objects in front of you, such as your controllers, or racing wheel, or joystick setup. Keyboard and mouse isn't a great experience in VR, but with the Vive you can at least see them without taking the headset off.

As for the Xbox One controller, what do you expect to use for VR? Games are being programed for navigation with the Xbox controller layout. Another controller style could work, but its easier on developers to have a specific controller in mind.

Getting people excited for the day they can afford something spectacular will do way more for adoption than a half-assed product that many people can afford right off the hop. A limited VR experience will render people bored with VR, which will doom the medium in the long run.

The developers feel like the features are necessary, but different developers feel that for different features. Kind of makes their feelings pointless.

I already provided a solution for the audio, and a "Use XBox One controller for recommended experience" -sticker would have been enough for the controller. I would expect quite a lot of people interested in VR already owning a controller (whether XBox one or one that is similar, but can use the same mappings).

The camera on Vive may be a decent addition, to see your controller, but I doubt it is something that makes or breaks the first wave of VR. Especially with dedicated controllers coming soon.

I doubt the omission of the features I mentioned would make the first VR (home) experiences more limited, since most of the first wave experiences are likely to be adaptations of existing products / dual releases for regular displays.

And while we are talking about necessary features for a great VR experience, both headsets lack a high (enough) resolution display and eye tracking.

 

[kcarbotte]

You don't feel like these things are necessary, but the people building them definitely feel like these additions are necessary, and there's good reason.

HTC isn't "adding" room roamiing capability. That was the play all along. Room scale tracking was announced the day that Vive was first revealed to the world. The camera wasn't a new addition either. It was always planned, but the early dev kits didn't include the camera. Interanally they have been testing the camera system since day one.

Oculus added its own headphones with built-in DAC for good reason too. Spacial audio is very important, and most people don't have headphones ready for that kind of environment. By including them, it ensures that everyone has a superior audio experience. Sound is half of the experience, so Oculus took this part seriously.

You say that should be focusing on the sitting experience, well I would argue that HTC and Valve are considering that with the front facing camera. It actually helps identify objects in front of you, such as your controllers, or racing wheel, or joystick setup. Keyboard and mouse isn't a great experience in VR, but with the Vive you can at least see them without taking the headset off.

As for the Xbox One controller, what do you expect to use for VR? Games are being programed for navigation with the Xbox controller layout. Another controller style could work, but its easier on developers to have a specific controller in mind.

Getting people excited for the day they can afford something spectacular will do way more for adoption than a half-assed product that many people can afford right off the hop. A limited VR experience will render people bored with VR, which will doom the medium in the long run.

The developers feel like the features are necessary, but different developers feel that for different features. Kind of makes their feelings pointless.

I already provided a solution for the audio, and a "Use XBox One controller for recommended experience" -sticker would have been enough for the controller. I would expect quite a lot of people interested in VR already owning a controller (whether XBox one or one that is similar, but can use the same mappings).

The camera on Vive may be a decent addition, to see your controller, but I doubt it is something that makes or breaks the first wave of VR. Especially with dedicated controllers coming soon.

I doubt the omission of the features I mentioned would make the first VR (home) experiences more limited, since most of the first wave experiences are likely to be adaptations of existing products / dual releases for regular displays.

And while we are talking about necessary features for a great VR experience, both headsets lack a high (enough) resolution display and eye tracking.

Microsoft partnered with Oculus a year ago for this. The controller that is being included doesn't add any meaningful cost to the package. Even if they are paying the cost a retailer would for the controller (which they likely aren't), the cost is at most $45-50. If people still need to have the controller to play with the Rift, then the cost savings are moot. That's like selling a console without a gamepad. It's an incomplete package. People don't like having to buy extras if it's a required item to use the thing.

The camera in the Vive is a massive addition to VR. You'll see when you get to try it eventually. Those dedicated controllers you mentioned, they're not visible to you while the headset is on withought Vive's chaperone system. It doesn't require the camera for that, but it's a huge advantage over what Oculus has provided for finding your controllers while in VR (read: nothing).

Your assumption about the games is way off too. There are very few games announced that are being released in VR and for regular screens. The way games need to be created for VR is very different, so no one is putting in that effort.
The only exceptions that I know of are a handfull of simulation games, such as Project Cars, Eliteangerous and Star Citizen.
Vive games are certainly not being developed for regular screens. Many of them won't even be offered on other VR platforms.

You seem to be contradicting yourself though. What do you want? A basic VR headset that is cheaper, or a super advanced headset that costs even more?
The reason the resolution isn't as high as you believe it needs to be, is because the technology isn't there yet. You can't have a 4K screen in a VR headset, even if the displays were available. Graphics cards can't produce that kind of resolution at the frame rates you need for smoothe, motion-sickess-free VR.
Also, eye tracking is just begining to be a thing in the PC market. It's cutting edge technology that isn't yet supported by game developers very much, and would add even more cost to a VR headset.
Eye tracking will be a benefit, but its a long way from a make or break feature.

Oculus and HTC are taking on the problems that they can solve right now. The other features that you think are so necessarry would raise the cost far higher than anything they are including for the first round, including the front facing camera and room scale trackers that HTC and Valve are working on.

 

[DrakeFS]

*sigh* Valve / HTC adds a useless camera / room roaming capabilities, and Oculus adds useless headphones and bundled stuff (like remote and controller). I know they are all trying to make VR the best they possibly can, but it would seem much more logical to me to get the core experience of being a 3D HMD for sitting users right first. That's what it will be used for at first for the most part anyway, and the simpler tech would be a cheaper product, which in turn would mean a wider adoption rate.

Except VR is not just about the image presentation. VR is about immersion, which requires more than your eyes to be fooled. What you are describing is basically Gear VR. Its just not same, immersion wise.

 

[Bloob]

Microsoft partnered with Oculus a year ago for this. The controller that is being included doesn't add any meaningful cost to the package. Even if they are paying the cost a retailer would for the controller (which they likely aren't), the cost is at most $45-50. If people still need to have the controller to play with the Rift, then the cost savings are moot. That's like selling a console without a gamepad. It's an incomplete package. People don't like having to buy extras if it's a required item to use the thing.

The camera in the Vive is a massive addition to VR. You'll see when you get to try it eventually. Those dedicated controllers you mentioned, they're not visible to you while the headset is on withought Vive's chaperone system. It doesn't require the camera for that, but it's a huge advantage over what Oculus has provided for finding your controllers while in VR (read: nothing).

Your assumption about the games is way off too. There are very few games announced that are being released in VR and for regular screens. The way games need to be created for VR is very different, so no one is putting in that effort.
The only exceptions that I know of are a handfull of simulation games, such as Project Cars, Eliteangerous and Star Citizen.
Vive games are certainly not being developed for regular screens. Many of them won't even be offered on other VR platforms.

You seem to be contradicting yourself though. What do you want? A basic VR headset that is cheaper, or a super advanced headset that costs even more?
The reason the resolution isn't as high as you believe it needs to be, is because the technology isn't there yet. You can't have a 4K screen in a VR headset, even if the displays were available. Graphics cards can't produce that kind of resolution at the frame rates you need for smoothe, motion-sickess-free VR.
Also, eye tracking is just begining to be a thing in the PC market. It's cutting edge technology that isn't yet supported by game developers very much, and would add even more cost to a VR headset.
Eye tracking will be a benefit, but its a long way from a make or break feature.

Oculus and HTC are taking on the problems that they can solve right now. The other features that you think are so necessarry would raise the cost far higher than anything they are including for the first round, including the front facing camera and room scale trackers that HTC and Valve are working on.

You most certainly can have 4K on VR, IF you have eye tracking so you can limit the area (ie. you would only need to render at 4K-like density in 1/8th of the screen). Had Oculus and Rift focused on eye tracking from the get-go, as much as they did on other features, we would have it.

As for contradicting myself, regarding pricing, I just stated my preferred area of focus for VR. As in, if they are focusing features instead of price point, eye tracking and the 4K resolution allowed by it, would be my preference. I guess we will have to agree to disagree on our preferred features.

Except VR is not just about the image presentation. VR is about immersion, which requires more than your eyes to be fooled. What you are describing is basically Gear VR. Its just not same, immersion wise.

Yes, VR requires more than your eyes to be fooled, but with 1/4-1/3 of our brain dedicated to vision, our eyes are the first ones needing to be fooled. Gear VR also has issues with low resolution, high latency sensors (relatively), power, input options and high price (if you don't own a compatible phone). Oculus did not need to provide the audio experience themselves (which does not mean that the audio experience would not need to be there (AFAIK, it's not there in the Vive), and I fail to see how Vive's camera is likely to increase immersion (except a few scenarios involving racing-sim controllers or similar, it might still be a good additional feature, as pointed by kcarbotte above, but I don't feel like it needs to be a first-gen feature).

 

[kcarbotte]

Microsoft partnered with Oculus a year ago for this. The controller that is being included doesn't add any meaningful cost to the package. Even if they are paying the cost a retailer would for the controller (which they likely aren't), the cost is at most $45-50. If people still need to have the controller to play with the Rift, then the cost savings are moot. That's like selling a console without a gamepad. It's an incomplete package. People don't like having to buy extras if it's a required item to use the thing.

The camera in the Vive is a massive addition to VR. You'll see when you get to try it eventually. Those dedicated controllers you mentioned, they're not visible to you while the headset is on withought Vive's chaperone system. It doesn't require the camera for that, but it's a huge advantage over what Oculus has provided for finding your controllers while in VR (read: nothing).

Your assumption about the games is way off too. There are very few games announced that are being released in VR and for regular screens. The way games need to be created for VR is very different, so no one is putting in that effort.
The only exceptions that I know of are a handfull of simulation games, such as Project Cars, Eliteangerous and Star Citizen.
Vive games are certainly not being developed for regular screens. Many of them won't even be offered on other VR platforms.

You seem to be contradicting yourself though. What do you want? A basic VR headset that is cheaper, or a super advanced headset that costs even more?
The reason the resolution isn't as high as you believe it needs to be, is because the technology isn't there yet. You can't have a 4K screen in a VR headset, even if the displays were available. Graphics cards can't produce that kind of resolution at the frame rates you need for smoothe, motion-sickess-free VR.
Also, eye tracking is just begining to be a thing in the PC market. It's cutting edge technology that isn't yet supported by game developers very much, and would add even more cost to a VR headset.
Eye tracking will be a benefit, but its a long way from a make or break feature.

Oculus and HTC are taking on the problems that they can solve right now. The other features that you think are so necessarry would raise the cost far higher than anything they are including for the first round, including the front facing camera and room scale trackers that HTC and Valve are working on.

You most certainly can have 4K on VR, IF you have eye tracking so you can limit the area (ie. you would only need to render at 4K-like density in 1/8th of the screen). Had Oculus and Rift focused on eye tracking from the get-go, as much as they did on other features, we would have it.

As for contradicting myself, regarding pricing, I just stated my preferred area of focus for VR. As in, if they are focusing features instead of price point, eye tracking and the 4K resolution allowed by it, would be my preference. I guess we will have to agree to disagree on our preferred features.

You are missing one very, very important part.
VR requires at least 90Hz/90FPS to ward off motion sickness. This has been proven, and corroborated by basically everyone working in the industry.
Currently there are no displays available that offer that high of a resolution, and that high of a refresh rate. You can't even get 1440p 90Hz displays. Hell the screens being used in the Rift and Vive, 2160x1200/90Hz, is a custom panel designed specifically for VR.
The technology is simply not there for that high of a resolution in VR. It will come, but you are talking about second generation stuff. For it to be first gen would require a much longer wait to launch the medium.

Same thing with eye tracking. Sure it has been shown to work, but theres a reason its not on the market yet. There's still work to be done to perfect foveated rendering. Its a problem that many companies are tackling, but again, its just not there yet for VR.

 

[Ve1ox]

You most certainly can have 4K on VR, IF you have eye tracking so you can limit the area (ie. you would only need to render at 4K-like density in 1/8th of the screen). Had Oculus and Rift focused on eye tracking from the get-go, as much as they did on other features, we would have it.

Hmm interesting idea. So 1/8th of the screen is rendered in 4k and the rest is rendered in lower res for peripheral vision? Sounds cool but it'd presumably just hike the cost up far higher than it currently is due to requiring 4k screens and the additional tech for eye tracking?

Room scale VR the standard in 3 years. Do these wackjobs actually believe such drivel? How many people's homes have an entire room they can set aside for VR?

Vive is cool, but it's also doomed to fail. Its cost will be too high for such an incredibly niche product. Oculus is aiming for the high-end sector much more intelligently.

Definitely don't agree with this. Having been an avid watcher of VR since early last year, HTC's system is far more attractive to me than the rift. I only have at most a 5x5 feet area and that's if I shift things around but I still want the ability to move around, bend down, duck and so on which seems much more limited on the rift. It seems like moving out of the camera range would totally break immersion & bug the hell out of me. A cone of available area just seems so limited by comparison to Vives room capable system. Sounds fine for cockpit games but I wan't the whole package!

 

[kcarbotte]

You most certainly can have 4K on VR, IF you have eye tracking so you can limit the area (ie. you would only need to render at 4K-like density in 1/8th of the screen). Had Oculus and Rift focused on eye tracking from the get-go, as much as they did on other features, we would have it.

Hmm interesting idea. So 1/8th of the screen is rendered in 4k and the rest is rendered in lower res for peripheral vision? Sounds cool but it'd presumably just hike the cost up far higher than it currently is due to requiring 4k screens and the additional tech for eye tracking?

This concept is called Foveated rendering. Several companies are working on trying to make it work. The idea is that you only render the point you are focusing on in full detail.
Nvidia has been trying to make this work, but it will require eye tracking to work properly, and be unnoticeable to you.

Fove is working on headset that will have eye-tracking built in, but it still have a lot of work before its ready for a consumer release. http://www.tomshardware.com/news/fove-vr-first-look-ces,30964.html
StarVR is also putting eye tracking into its headset. Starbreeze announced a partnership with Tobii-tech last year. But again, that headset is a while off. It uses two UHD panels, but you can't get those in 90Hz displays yet either, so that one won't be out for a while either. It's a second generation headset.
http://www.tomshardware.com/news/immersed-toronto-starvr-hands-on-experience,30234.html


When these headsets become available, then foveated rendering will become a possibility, and it will be a big leap in GPU performance. VR is just in its infancy right now. There's a lot a poeple working on improving the experience, but it will take some time.

Room scale VR the standard in 3 years. Do these wackjobs actually believe such drivel? How many people's homes have an entire room they can set aside for VR?

Vive is cool, but it's also doomed to fail. Its cost will be too high for such an incredibly niche product. Oculus is aiming for the high-end sector much more intelligently.

Definitely don't agree with this. Having been an avid watcher of VR since early last year, HTC's system is far more attractive to me than the rift. I only have at most a 5x5 feet area and that's if I shift things around but I still want the ability to move around, bend down, duck and so on which seems much more limited on the rift. It seems like moving out of the camera range would totally break immersion & bug the hell out of me. A cone of available area just seems so limited by comparison to Vives room capable system. Sounds fine for cockpit games but I wan't the whole package!

Your 5x5 foot space will work just fine for Vive, but you may be restricted from player certain experiences. The defined space can be shrunk down to a seated position, so you should be fine.

Even a game like hover junkers, that requires you to move around, ducking and shooting, can be played in a small space.