Meant to be Seen

LordJuanlo wrote:Going from 800p to 1080p is good, but is not "that" good. The real deal would be something like 2560x1600, four times more pixels than the developer version, now that would be an improvement. That would be perfect for a consumer version of the Rift in late 2013/early 2014. At that time, even medium range videocards should be able to handle that resolution with no problems. Today, any high-end video card can handle 5760x1080 resolutions (triple screen) in almost every game.

Don't forget that on top of the triple screen you need to render at minimum 60 FPS to get 30 FPS 3D (let's say baseline 45 to be safe, meaning you'd need to render at 90 FPS). Asking this at such a resolution might not be realistic except for top cards. And while it'd be awesome if everyone could have that the reality is that very few people have capable hardware and this isn't going to change any time soon. As of September the majority of Steam users are using integrated HD 3000 graphics on second-gen Core i processors! Only 1.44% of the Steam population is using a GTX 670 or 680 which might be able to do what you are asking (5760x1080 @ 90 FPS). The 690 isn't even on the board yet!

Source of stats: http://store.steampowered.com/hwsurvey

@benz145: As I mentioned earlier...aren't high resolution displays a non-issue as long as the video card has a hardware scaler to inflate the resolution to match the device. So if you have a low power rig, can't you just set your resolution lower and have it upsample at practically no cost? And then of course, people with high power rigs could render at the native resolution.

http://www.mtbs3d.com/phpBB/viewtopic.p ... 055#p84613

The inherent loss of definition and artefacts in a scaled image might be a little too distracting when completely surrounded by it. Certainly not the end of the world of course, but it is a consideration.

benz145 wrote:Don't forget that on top of the triple screen you need to render at minimum 60 FPS to get 30 FPS 3D (let's say baseline 45 to be safe, meaning you'd need to render at 90 FPS). Asking this at such a resolution might not be realistic except for top cards. And while it'd be awesome if everyone could have that the reality is that very few people have capable hardware and this isn't going to change any time soon.

Well, the way I see it, the Oculus Rift is not for the average Joe gamer, it's not something for the kind of people that uses an Intel HD 3000 card to play Steam games.

benz145 wrote:As of September the majority of Steam users are using integrated HD 3000 graphics on second-gen Core i processors! Only 1.44% of the Steam population is using a GTX 670 or 680 which might be able to do what you are asking (5760x1080 @ 90 FPS). The 690 isn't even on the board yet!

Source of stats: http://store.steampowered.com/hwsurvey

The majority of steam users aren't using Intel HD 3000, that's just the most used graphics at 4%. (Or 3.55% if you remove Macs from the equation. )

Regardless of the exact percentage with Intel HD "graphics", it is true that you will need a decent GPU to run newer games on the Rift well. My home machine is a i7+680, I promise you won't have any trouble with one of those. I am testing the 660 TI as well, it seems like it performs nearly on par with a stock 680. Definitely a good GPU for the price.

LordJuanlo wrote:Well, the way I see it, the Oculus Rift is not for the average Joe gamer, it's not something for the kind of people that uses an Intel HD 3000 card to play Steam games.

Please don't forget us poor untethered VR users. I would definitely consider myself a "power" user since I take the time to assemble and wear a mobile backpack unit for free motion VR gaming. Unfortunately I can't carry a 2 foot tower and a 60 pound battery supply on my back for long so I have to make some compromises in terms of graphical power and use a laptop. So people like me need the ability to scale down the hardware requirements - even for hard-core usage.

I think there is also no point to keeping the price of the Rift low if anyone using it will need a extremely hefty and pricey pc.

Kajos wrote:I think there is also no point to keeping the price of the Rift low if anyone using it will need a extremely hefty and pricey pc.

While the Rift may be more demanding due to the stereoscopy, we will deal with it like PC gamers always have done - lower the quality a bit. You don't have to run a game on high/full quality to enjoy it.

Mart wrote:

Kajos wrote:I think there is also no point to keeping the price of the Rift low if anyone using it will need a extremely hefty and pricey pc.

While the Rift may be more demanding due to the stereoscopy, we will deal with it like PC gamers always have done - lower the quality a bit. You don't have to run a game on high/full quality to enjoy it.

ditto

Does not surprise me that gtx 660 ti is close to gtx 680 in performance since gtx 680 is not a true successor in the gtx *80 lineup (I also have gtx 680 2gb)
It was supposed to be a second tier card, but was released as their flagship when nvidia saw it was beating AMDs flagship at the time.

So the big kepler card with 2300 cuda cores and twice as fast memory interface is still not out. (No I am not talking about dual gpu gtx 690)
Maybe it will be called gtx 685, but in any case its going to be the true high end card.

Just for my own clarity here - if you want to benchmark your system. If you can run your game stereoscopically at 1280x800 @ 60 FPS - or if you can run it in 2D at 800x600 @ 120 FPS, then that is approximately the performance you will need to use the Rift at the suggested settings. Correct?

Some of the RIFT DIY'ers have reported acceptable usage as low as 30 FPS. How much difference does the extra frame rate make? Is playing uncomfortable (ie. sickening) at frame rates lower than 60 FPS or just annoying I wonder?

brantlew wrote:Just for my own clarity here - if you want to benchmark your system. If you can run your game stereoscopically at 1280x800 @ 60 FPS - or if you can run it in 2D at 800x600 @ 120 FPS, then that is approximately the performance you will need to use the Rift at the suggested settings. Correct?

Some of the RIFT DIY'ers have reported acceptable usage as low as 30 FPS. How much difference does the extra frame rate make? Is playing uncomfortable (ie. sickening) at frame rates lower than 60 FPS or just annoying I wonder?

My thought would be you could better test with 800x600 @ 120 fps, because testing 1280x800 @ 60 fps will not get you the same amount of polygons.

Kajos wrote:My thought would be you could better test with 800x600 @ 120 fps, because testing 1280x800 @ 60 fps will not get you the same amount of polygons.

Stereoscopic SBS at 1280x800 @ 60 FPS should be almost identical to the Rift render pathway (minus the warping).

LordJuanlo wrote:Going from 800p to 1080p is good, but is not "that" good. The real deal would be something like 2560x1600, four times more pixels than the developer version, now that would be an improvement. That would be perfect for a consumer version of the Rift in late 2013/early 2014. At that time, even medium range videocards should be able to handle that resolution with no problems. Today, any high-end video card can handle 5760x1080 resolutions (triple screen) in almost every game.

You are right. 2560x1600 (WQXGA) would be an ideal resolution for the consumer version. 4X the pixels would be a massive improvement in fidelity and high end cards could probably handle that resolution. Just as important though would be that the Rift could then offer the original 1280x800 low resolution mode that would upsample perfectly to the display without any artifacts.

Or similarly to make the Rift more console friendly, they could do a 2560x1440 (WQHD) panel and a 1280x720 (720p) low res mode.

Of course this is all just wish list. Who knows if they could ever find panels at these resolutions at the correct size.

brantlew wrote:

Kajos wrote:My thought would be you could better test with 800x600 @ 120 fps, because testing 1280x800 @ 60 fps will not get you the same amount of polygons.

Stereoscopic SBS at 1280x800 @ 60 FPS should be almost identical to the Rift render pathway (minus the warping).

I've wondered about this actually.. For Rift do you render the same scene twice side-by-side, or do you render it once on a double framerate and then run some filter that halves the framerate and assembles the frames side by side?

I'm kinda inclined to agree with Kajos here, since rendering the same scene two times is different than rendering once for a higher res.

You render two separate frames at half the resolution and then append them side-by-side. So you have to account for the render latency per frame. Thus half screen resolution at 120 FPS or 2 x half screen at 60 FPS should be equivalent.

Did anybody see this Kick Start Episode #1: The Oculus Rift?

... nice ending

Soooooo.... Has it been a "couple weeks" yet?

MSat wrote:Soooooo.... Has it been a "couple weeks" yet?

What are you referring to? Just curious

Namielus wrote:

MSat wrote:Soooooo.... Has it been a "couple weeks" yet?

What are you referring to? Just curious

Big news from Oculus that was to be revealed in a "couple weeks" several weeks ago. I'm only teasing, but I'm really looking forward to hear it. Luckily, Palmer has been giving us a trail of crumbs as of late to tide us over.

I was really happy just to see the confirmation of 9dof tracker, maybe palmers post number 1337 will reveal something equally awesome.

Just wondering if we know any more about the display driver this is using. It's been mentioned that it takes DVI/HDMI, but I was wondering if it will upscale an image that is not 1280*800? It would be great (for me... not a lot of others, I imagine) if it could upscale from 480p.

My understanding is that Oculus doesn't want to introduce any additional lag in the system which means no hardware scaler, but I could be wrong. Anyway your graphics card can handle the scaling.

brantlew wrote:My understanding is that Oculus doesn't want to introduce any additional lag in the system which means no hardware scaler, but I could be wrong. Anyway your graphics card can handle the scaling.

My idea is for a use that isn't connected to a computer.
Basically two optical sensors http://www.ebay.com.au/itm/Car-Rear-Vie ... 733wt_1072 with frames sync'd and analog multiplexer that switches from one to the other half way each line. This is transmitted over a standard analog video transmit/receive, and then the signal goes into something like this:
http://www.ebay.com.au/itm/320961872476 ... 187wt_1072
And fed into the rift.
The idea is to have the image SBS at transmit end.

... or if it had analog input, but I doubt this.

Hello everybody.
I have a suggestion that if you connect the 2 in 1.
Ie connect vertualny Oculus Rift helmet with brain-computer interface (IMC).

A. A. FROLOV, V. Yu. ROSCHIN
Institute of Higher Nervous Activity and Neurophysiology
of the Russian Academy of Sciences, Moscow
E-mail: aafrolov@mail.ru
BRAIN-COMPUTER INTERFACE. REALITY AND FUTURE
Abstract
The present paper is based on a review on literature data and own research on
implementation and application of Brain-Computer Interfaces (BCI). BCI is a
new way for interaction of a man with a machine based on directtransformation
of a man’s will, reflected in biosignals registered from the brain, into control
commands. The literature data and own preliminary investigations let suppose
that the following three paradigms can be considered as potentially good to
achieve reasonable level of informational performance:
(a) A noninvasive BCI, based on recognition of particular mental states,
induced by imaginary movement performance. Such BCI allows to operate
with a fixed set of discrete commands requiring very simple operator
training and have quite good informational performance.
(b) A noninvasive BCI for gradual multidimensional control. This type of
BCI, after training, allows natural controlling of an external technical
system as an own (virtual) execution organ, in contrast to control by a set
of discrete commands.
(c) An invasive BCI, based on bi-directional brain-computer data exchange
through the implanted electrodes, that allows to incorporate an external
technical system into internal neuronal scheme of the body in process of
training and, hence, to operate the external system as an own executable
organ.
The up to date implementations of BCI of the first type achieveinformational
performance of up to 35 bits per minute after training duringas low as 20 minutes.
Since the second type BCI requires longer training by biological feedback,
potentially they can have much higher performance (100 bitsper minute had
achieved), allowing at the same time to independently control several degrees of
freedom.

Below I will give a couple of sites that describe in detail brain-computer interface. But they Russkoya language, so most likely you will need to use an interpreter, which would further explore this type kamunikatsii:
http://brain.bio.msu.ru/bci_r.htm
http://neurolectures.narod.ru/2008/Frolov-2008.pdf
http://www.dislife.ru/flow/theme/12479/
http://trv-science.ru/2009/07/07/ovlade ... yat-minut/

And now a couple of sites with YouTube, so to speak significant factor:
http://www.youtube.com/watch?v=na-qyomedLI
http://www.youtube.com/watch?v=H1QJiESX ... =endscreen
http://www.youtube.com/watch?v=uhEDhAHq ... re=related

I see that the IMC system is not yet fully elaboration. But Imagine if they combine. Then in games in general are no longer needed will be, no joystick or keyboard and mouse, or something else. All actions will be carried out with the help of brain impulses.

I hope you like the idea that can turn the entire computer games industry and bring them to the next level.

I like the idea of having a brain computer interface, particularly for walking. It would be good to just be able to look where you want to go and think about walking to walk there. From the research I've seen, it looks like that is possible and effective. I tried to persuade Emotiv to add a Cognitiv detection for walking, but I don't think they have. The Emotiv Epoc was too expensive for me, and I haven't been able to try any sort of direct BCI yet.

I think it's unlikely the Oculus Rift will include a BCI though, since Palmer Luckey has a very low opinion of the Emotiv Epoc.

Still there is a good object lesson in the TV series "House MD" in 5 season 19 series.

I want eventually to get around it as shown in the Anime Sword Art Online.
http://www.animespirit.ru/anime/rs/seri ... nline.html

All this is possible and feasible, the main start of the experiment (test) and then, if everything will turn out that Palmer Luckey will be 2-3 steps away from the immersive vertualno reality.

From John Carmack tweeter: "Got approval for GPL release of Doom 3 BFG code (minus third party bits)! @idBrianHarris has already done most of the work."

It's almost been a month since the last Kickstarter update. Throw us a bone, Palmer!

I'm starving for news

Brain computer interface would make Oculus Rift "the best invention ever" for me.

Sadhu wrote:From John Carmack tweeter: "Got approval for GPL release of Doom 3 BFG code (minus third party bits)! @idBrianHarris has already done most of the work."

But the Doom 3 BFG code hasn't finished being written yet. We're still waiting for the final Oculus Rift dev kit to be finalised so they can support it properly.

... unless...

Maybe when he said Doom 3 BFG would be supporting the Oculus Rift, what he really meant was that we would be expected to add the final support ourselves.

The idea of combining this happened vertualno reality and points of IMC will lead to neuro helmet vertualno reality. I understand that the developers make their lives easier and use this device with gamers (people) went from the idea of ​​the helmet, and came to the creation of simple points. And I think that Return to the good old helmet can probably not radyvat Most of the people. But otherwise as. Some points of total immersion in the reality vertualno not achieve.

I also understand that there may be difficulties with the perception of information and its transmission to NSHVR (Neuro Helmet vertualno Reality). But more about that later.

2EyeGuy wrote:I like the idea of having a brain computer interface, particularly for walking. It would be good to just be able to look where you want to go and think about walking to walk there. From the research I've seen, it looks like that is possible and effective.

That sounds too good to be true. Man that would be awesome. What I've seen seemed pretty unreliable and made Kinect look flawless in comparison though.

jaybug wrote:It's almost been a month since the last Kickstarter update. Throw us a bone, Palmer!

I'm starving for news

+1

2EyeGuy wrote:

Sadhu wrote:From John Carmack tweeter: "Got approval for GPL release of Doom 3 BFG code (minus third party bits)! @idBrianHarris has already done most of the work."

But the Doom 3 BFG code hasn't finished being written yet. We're still waiting for the final Oculus Rift dev kit to be finalised so they can support it properly.

... unless...

Maybe when he said Doom 3 BFG would be supporting the Oculus Rift, what he really meant was that we would be expected to add the final support ourselves.

It's finished. I'm guessing the Oculus integration hasn't been finished yet. If you want another copy, you can get BFG edition on Steam right now.

I'm getting all my friends Doom 3 BFG for christmas, so I can show off my Rift at their places :p

And by all my friends, I mean all my friends with a capable computer.

Which is just one person... Maybe he knows more PC gamers, I'll give him an extra copy :p

Wouldn't it be be cheaper to just take your computer with you?

Would be even cheaper to just stay home. Just thought I might prefer not to drag too much stuff around

Don't try to talk me out of not having to find a proper christmas gift for this person. :p

One of the guys working at Unity reply to question about Oculus RIft in video hangout.

Interestingly he mentions something about the dev kit having higher screen res? Im not sure, what do you guys make out of it?

https://www.youtube.com/watch?feature=p ... S8#t=2180s