http://www.kickstarter.com/projects/1523379957/oculus-rift-step-into-the-game

This Kickstarter project is close to making 10 times the amount of money pledged, beyond the amount they originally asked for.

Could pega-1, or Bigfoot, or henes, or anyone else who is involved with the MorphOS3.1 video card drivers take a look at this thing and determine if it could be adapted for use on our current MorphOS3.1 computers?

From what I found on the Kickstarter site, the minimum requirements don't look hard to meet, as it only uses DVI out and 720p resolution, but it also says it is for PC's only, so that means that they probably are only planning on drivers for x86 computers, to handle the gyroscopic head movement technology that rotates the display in the game to match the movements of the goggles on your head. Since it will work with video cards with a DVI output, maybe a driver could be written for MorphOS3.1 supported video cards.

Of course then it would be useless until some games are written to take advantage of the 3D display, or the games that 2.5 million dollars of Kickstarter money can get lots of existing games ported to use this display device, and the popularity of it will get lots of new games written to take advantage of the 3D virtual reality display & gyroscopic head movement detector technology.

It would be a cool game display device and looks like it is going to get lots of support, so will probably have many games written to take advantage of the display and virtual reality technology. The company creating it appears to already have a couple of game engines written, or adapted to take advantage of the head mounted gyroscope motion detector.

If it is worthwhile project for MorphOS3.1, I would be willing to contribute toward the $3,000 pledge level to get 10 developer units plus the SDK, for the MorphOS Dev. Team to use to create the drivers needed to make it work on MorphOS3.1.

It would be great to have something like this that is cutting edge technology, for MorphOS3.1 users, even though our computers are 5 to 10 years old technology, but only if the MorphOS Development Team could make it work on what we have now, and not some future version of MorphOS for x86.

Another crazy idea, from you know who.

Even without head tracking, stereo 3D Quake would rock `n` roll!

Need some more horsepower than we`ve got ATM though - maybe that R300 driver eh

From what I read, the resolution is just 640x800 per eye, so even an Efika with radeon 92xx should handle that, when speaking about quake, as even bigger resolutions are no problem. e.g 1024x768 34 fps.

I have not much clue about such 3D programming, but I know that for example PS3 games just need a little more cpu/gpu power to render the second eyes screen, as most of the content is identical and it is just a tiny camera move to render.

Geit

Don`t know about that but on my PC (Deneb 3.4GHz / 8800GT), switching to stereo mode on anything beyond Q3 absolutely cripples framerates that are fine in mono.

I guess the render engine needs to be prepared for that, or it will simply half the frame rate to create twice as much images per second.

Anyway, would be cool to play Armagetron using these gadget. :D


Geit

[ Edited by geit 12.09.2012 - 11:42 ]

Sure would!

geit,
Quote:
In OpenGL, using glDrawBuffer() call, you can specify GL_BACK_LEFT and GL_BACK_RIGHT, which specify to which "eye" you want to render. You still need to do transformations (to make image appear as it was from different position) manually. Of course this would need to be supported in TinyGL.


...and even if not:
It depends on how the 3D view needs to be formed. In some cases, you simply put left eye image on left and right to right side of the window (or vice versa, if using "cross-eye 3D", which doesn't need any glasses), this is also rather trivial to achieve by setting different glViewport() for each eye - actually not much more complex than GL_BACK_LEFT/RIGHT support.

Of course it MIGHT be something more complex, like having separate RGB image and depth values for each pixel (which is possible to get from OpenGL, assuming your program DOES use depth buffer in a suitable way, and doesn't f.ex. clear it mid-frame). This is also doable without any tinygl support (same viewport setting as before, only draw the second view without any textures, all objects full white, and using suitable fog) - been there, done that.

I'd definitely be interested in 3D glasses too, especially if I could get them working with MorphOS

Also about head tracking, it's often provided as 3D vectors, which are "easily adaptable for OpenGL use" - the only question is, how well the way you can get them is documented.

Give Trevor Dickinson credit for pointing this Kickstarter project out to Amiga people on a different forum site.

It looked like something that might make it big if they can deliver an excellent quality final consumer edition of this product and the game publishers support it with new games and ports of their old games.

Even if our MorphOS computers don't have the power to generate the display required for this device now, plus the added power required to run the driver for the head tracking software, on games that generate their standard of 720p, maybe MorphOS will be able to handle it in the future with newer video card drivers, or a port of MorphOS to more powerful hardware plus newer video card drivers.

It would be great to have this device working on MorphOS as it will be required for the ports of some games in the future, and hopefully many Open Source games will use it, that can be ported to MorphOS. With a few game engines ported to MorphOS that can use this device, we might even begin seeing some new MorphOS game designers creating their own games using those ported game engines with "borrowed" art and custom rendered characters.

A nice dream for the future of MorphOS, but I have no idea how far into the future we must wait for it to be a reality, if ever.

http://fabiensanglard.net/vr_headset/index.php :D

raistlin77it,
Quote:

Okay, reading the page:
Quote:
This is quite "no brainer", as easy to do in practice (even without any specific support for it in OpenGL implementation) as I said earlier.
Quote:
This, however, worries me more. Apparently the hardware implementation distorts the image in a certain way, and you need to "anti-distort" it (so that they cancel out each other) in your code.

Sure, it might not be that complex with a suitable shader, but that's something we lack currently (except AROS). Also, I'm not very thrilled at the idea of distorting 2D image as low resolution as 640x480... Well, guess you can't have everything, at least at such a cheap price

In any case, the 3D effect WILL probably work even without the (anti-)distortion, but will be... slightly distorted. See the last image on the page, what we would get without this last step would be (more or less) "inverse of that".

So in short:
Assuming it will just work as normal display not needing any additional controls (forgetting head tracking etc. for now), it would be relatively easy to get a 3D image from almost any 3D game on it, but it will need code modifications to every game, and the end result will be slightly distorted.

Maybe it might be possible to achieve some distortion with OpenGL projection matrix? Wouldn't work too well with big polygons anyway...

Few more thoughts:
The 2D image can be distorted "on-screen", but will be extremely slow. Also any 2D stuff can be "pre-distorted" (also in better quality than what OpenGL does), but that won't help much with something that would need to move.

[ Edited by Jupp3 12.09.2012 - 21:47 ]

> Give Trevor Dickinson credit for pointing this Kickstarter project out to Amiga people
> on a different forum site.

I hope you don't mind that I rather give this credit to clusteruk :-)

http://amigaworld.net/modules/newbb/viewtopic.php?topic_id=36157&forum=2

We can make the "distortion" tiling the "screen/eye" texture like the water screensaver from the morphos SDK (not optimal, but we don't have shaders)

in the meantime fabien sanglard is waiting the device to implement it in quake 2 (and we can use the distorting algo in software render )

from https://github.com/fabiensanglard/Quake-2

"This is Quake2 with support for libfreespace motion sensor.
It has been tested with a FSRK-USB-2.
As soon as the Occulus Rift will be released I will start working on pre-warping the output.
The solution is now a Visual Studio 2010 project."

Ok.. now do you really, REALLY, want to sit in front of your computer wearing THAT? What if someone sees..

What happened to good old zip beer on sofa with friends kicking the shit out of eachother in Tekken, jeez...

hooligan,
Quote:
I thought out of all people, YOU would have understood.

Wearing such glasses make it "slightly less obvious" you are watching porn

But sure, it's probably still pretty obvious due to "other reasons"

is there a kick starter for some high technology aspirin ? cause that's often needed after using this kind of device for more than 10 mins.