cancel
Showing results for 
Search instead for 
Did you mean: 

Oculus Certified

Leonard_Powers
Protege
After Reading the VR Sickness Blog entry below:
http://www.oculusvr.com/blog/vr-sickness-the-rift-and-how-game-developers-can-help
For others we need the help of developers to do the right thing, use the correct mathematical models, and resist the completely understandable temptation to just “get something on screen”. We’ll go into that in more detail in the SDK and blogs in the future.

it has become clear to me that there is ALOT Developers can do wrong to make the VR Experience visually stunning, yet very poor for the average user.

The current tools out there are great to get the ball rolling, but because there are many technical steps that need to be taken to make sure that the user experience is "technically correct", I feel that there should be a certify program by Oculus that Developers can pay for to have their game checked.

This would provide consistency and an easy way for end users to ensure that what they are about to buy has been "validated" by the people who know best.

Quality control is nothing new or bad. Both Microsoft and Sony do this very heavily for understandable reasons.
The PC Market is an open playing field and this will be entirely optional for Developers, but I think that this would be very important to deliver a consistent experience.
18 REPLIES 18

DrSnake
Protege
How about a Oculus Training Course where they address these issues and the Developers can get certified?? Give a man a duck and he can eat for a day, teach a men to fish and he can eat for a lifetime 😉

Leonard_Powers
Protege
I don't like fish! :lol: :lol: :lol:
Yeah, that sounds like a good idea 😄

tomf
Explorer
"Build a man a fire, he'll be warm for the night. Set a man on fire, he'll be warm for the rest of his life."

We plan on expanding the SDK with a suite of tests to make sure you did get your math right. One idea we had was for reference scenes of known-sized blocks at known distances. You build this scene in your engine, take a screenshot, and compare it to a reference screenshot that you need to match with pixel-perfect accuracy. If you don't - there's a bug somewhere!

This stuff is hard to get right, and even when you know it's wrong, it's often hard to figure out which part is wrong. It's not just a case a of getting an "even number of sign errors". Any suggestions on simple debugging methods you like to see, fire away...

geekmaster
Protege
"tomf" wrote:
"Build a man a fire, he'll be warm for the night. Set a man on fire, he'll be warm for the rest of his life."
... Any suggestions on simple debugging methods you like to see, fire away...

The most accurate non-subjective testing method would be to measure barf-bag contents after a testing session... 😉

That "fire" quote was my email signature in the "pre-web" days of the internet (email, newsgroups, ftp, gopher). The interwebs are for noobs.
:lol:

jherico
Adventurer
"tomf" wrote:
We plan on expanding the SDK with a suite of tests to make sure you did get your math right. One idea we had was for reference scenes of known-sized blocks at known distances.


That's an awesome idea. If I have time I'll make a go at integrating this into the GitHub version of the SDK this week.

I'm also planning on creating PNGs of 2D textures that will, for any given pixel hold the X and Y coordinates of the pixel that should be rendered from the scene texture, for a given set of FOV values. There was discussion a while back on using such a lookup texture rather than doing the computation per pixel every time, so example code to compute them easily might be helpful.

tomf
Explorer
"jherico" wrote:
There was discussion a while back on using such a lookup texture rather than doing the computation per pixel every time, so example code to compute them easily might be helpful.

That's on the list of things to investigate. My current thoughts are to keep things function-based and generate the texture at start of day, rather than actually shipping the textures in the SDK. We'll see how well it works out and what machines each technique is faster on.

xSpektre
Honored Guest
With retrofitting games in mind, how difficult is it for developers to convert old games to be compatible with the Rift? Would the availability of Stereoscopic 3D be sufficient? Also, would fitting an engine with Rift capabilities speed up the process? And if so, by how much? The topic of things being "Oculus Certified" and Tom's response got me thinking as to how difficult it would be for even the average coder to get recent games fitted properly for the rift. I understand Vireio and Vorpx can do it, but not to the extent that 'native support' from devs can do it.

What I've really had in mind is replaying Bioshock Infinite when I get my Rift, simply because I loved the art direction and the cityscape.. and would really like to experience it the way one experiences New York for the first time.. (ironic, if you've played the game.) I'd like to know if it's easy/worth the effort for the devs to code Rift support into their Bioshock series natively. Like I said, I know Vorpx can do it, but Cymatic Bruce + me going cross-eyed saw that the 3D effect isn't as profound or convincing as in other natively supported games.

jherico
Adventurer
"xSpektre" wrote:
With retrofitting games in mind, how difficult is it for developers to convert old games to be compatible with the Rift? Would the availability of Stereoscopic 3D be sufficient?


I'm currently working on getting Homeworld working with the Rift, and it's non-trivial for someone who isn't familiar with the codebase to chase down where the insertion point for making sure the scene gets rendered twice from different viewpoints should be. Just having stereoscopic 3D in your game isn't good enough, because you need to have a shader based post-processing step that does the distortion (or you have to be ready to emulate a per-pixel distortion in your game engine in some other fashion, but retrofitting GL or DirectX shader support is probably the easiest way to do that).

"xSpektre" wrote:
The topic of things being "Oculus Certified" and Tom's response got me thinking as to how difficult it would be for even the average coder to get recent games fitted properly for the rift.


There's no such thing as an 'average coder' in my experience. Software development is just too large a field with too many variables to make any sort of reasonable decision or estimate based on the idea that someone is an 'average coder'.

"xSpektre" wrote:
I'd like to know if it's easy/worth the effort for the devs to code Rift support into their Bioshock series natively.


It's easy in that it's not an intractable problem. There might be hiccoughs, but there's certainly enough information in the SDK and examples for any developer familiar with the Bioshock Infinite graphic engine to do it. However, it's almost certainly not worth the effort, because it doesn't add anything to the bottom line. Right now the only benefit you get from supporting the Oculus Rift is making a bunch of other crazy developers happy.

jackwilliambell
Honored Guest
"DrSnake" wrote:
How about a Oculus Training Course where they address these issues and the Developers can get certified?? Give a man a duck and he can eat for a day, teach a men to fish and he can eat for a lifetime 😉


I prefer "Start a fire for a man and you keep him warm for a day. Start a man ON fire and you keep him warm for a lifetime!"