3D Gaming*

Isn't that the model that halves color depth for 3D? There was a fairly in-depth article on 3D tvs in the C't magazine a little while ago that I think I mentioned here earlier, indicating that Plasma is better at bringing an image into 3D within the short time-frame available creating an image that looks good no matter where on the screen you look and has the least artifacts. At the same time, the overall experience and picture quality, the Samsung LCD was still favored over the plasma. They also cited that even the worst 3D tv equals or outperforms the best we have in cinemas today.

Biggest negative right now in all instances seems to be the reduction in color/contrast, which can be huge (way more than just 50%, because the display simply has less time to show a frame to prevent bleeding). On the gaming side, Gran Turismo for instance automatically way ups the output levels to help compensate, which seems to be very effective going by E3 reactions.
 
All 600 Hz shutter glasses 3D plasmas will have to use 15 bit colour, they will need 960 Hz subfields for 120 Hz 24 bit (2x60 Hz) ... preferably 1152 Hz for 144 Hz 24 bit (2x72 Hz). As far as I know they are still all doing 600 Hz though.

3D in cinemas gets a bad rap because so many cinemas use RealD, which has a huge amount of ghosting.
 
Billy Idol said:
Do you mean the panasonic plasma vt20 ??
Click to expand...

The set I tried was indeed VT20. The 3D demo lasts for a long time. The segments that impressed the most are "Sync Future" and "Coral Wonderland". I spoke of "Coral Wonderland" before. "Sync Future" is a video on 3D art book, but it is wow-worthy (The girls behind me all "wowed" when they saw it today ^_^).

After playing Super Stardust HD and WipeOut on Sony 3DTV, the experience is very close to the Panasonic. I still prefer Panasonic by a thin margin.
 
If it is true it will be for PS4. Unless it is ready for september and launch it as part of the Move system, scrapping the use of the original PSeye. Which wont happen.

We already know they have a system like this in the works called ICU.

They could just sell a base that has slots for two PSEyes, but it will segregate the userbase even more so wouldnt see much use in games
 
Well they could release it as an upgrade for people with 3DTVs. They could not only do 3D video chat but also 3D augmented reality. That's if tracking the glowing ball or balls on two video streams isn't too processor intensive. Then maybe if Kinect is a big hit they could look into 3D interpretation from two image streams for PS4.
 
Stereoscopic 3D in Crysis 2

Is this what Crysis 2 will have for 3D?

Killzone 3, GT5, etc. use traditional 3D technique of render a different frame for each eye. This has massive processing cost (double normal).

Crysis 2 only renders one frame then uses secret low cost technique (parallax occlusion map?) to create second frame.

Do you think this is the technique they use? Can this be done on older GPUs like RSX and Xenos?

Has anyone used this technique before for games?

Look at this for demonstration. Very amazing and only a little bit artifacting.

2D image + Heightmap: http://www.scalari.net/wp-content/stereoscopic.jpg
=
http://www.scalari.net/wp-content/3d.gif

Source: http://www.scalari.net/2008/04/22/generating-stereoscopic-images-with-parallax-occlusion-mapping/

Thank you.
 
Last edited by a moderator:
Aah i think i need shutter glasses to see this in 3D i tried cross eyeing it didn't work.

But i believe they mention how they did it in one of their latest presentation that talked about the future of graphics.
The link kinda describe what was in the presentation if my memory serves me well.

This is the presentation post here
http://forum.beyond3d.com/showpost.php?p=1447192&postcount=859
Starts at slide 27
 
Last edited by a moderator:
Thank you for the link

dragonelite said:
Aah i think i need shutter glasses to see this in 3D i tried cross eyeing it didn't work.

But i believe they mention how they did it in one of their latest presentation that talked about the future of graphics.
The link kinda describe what was in the presentation if my memory serves me well.

This is the presentation post here
http://forum.beyond3d.com/showpost.php?p=1447192&postcount=859
Starts at slide 27
Click to expand...

I really like the statement about "perception" driven graphics. Too much time people waste on accuracy which is only good for screen-shots. What do people really see is what really matters. Also I am glad that he says temporal is good place to "trick" the eye. Very smart.

It is also because he says they have 3D on all consoles and describe type that I ask this question about specific technique that will be seen in the final console games.

I remember someone said the "brute force" 3D method of KZ3, GT5, etc. will not be in Crysis 2.
 
ihamoitc2005 said:
I remember someone said the "brute force" 3D method of KZ3, GT5, etc. will not be in Crysis 2.
Click to expand...

It will be present in engine and seeral other types of 3D methods. I'll assume only their cheap 3D is user accessible in console version.
 
Yes

Nebula said:
It will be present in engine and seeral other types of 3D methods. I'll assume only their cheap 3D is user accessible in console version.
Click to expand...

I think this is correct. They said they have found a technique which is cheap and can be on all consoles.

I wonder if it is the parralax occlusion method and also, if this is it, how they do this on old hardware like RSX and Xenos. I thought parallax occlusion mapping is only able only on directx10 hardware but maybe I am wrong on this.
 
ihamoitc2005 said:
I wonder if it is the parralax occlusion method and also, if this is it, how they do this on old hardware like RSX and Xenos. I thought parallax occlusion mapping is only able only on directx10 hardware but maybe I am wrong on this.
Click to expand...

Parallax occlusion mapping is no DX10 specific feature. Many games have this atleast in DX9 mode regarding PC games.
 
I have a question. Can't they just have a cpu to convert it to 3d after the frame is rendered kinda like what 3dtvs are doing to 2d signals ?
 
Why not it should be cheap. Well I have 3D plugin for TMT3 and it costs barely anything from CPU to make movies 3D on the fly.
 
Yeah what you get with on-the-fly 3D conversion. Though different methods in TMT3, red/cyan anamorphic, 120Hz frame sequential, line interleave 3D output, DLP 3D checkerboard output. Left/right stereoscopic vision or top/bottom with option to swap left/right eye view to relieve eye strain and configurable depth perception strength. The wonders of software! :LOL:
 
Any technique that doesn't render separate images for both eyes is going to have artifacts for certain cases, mainly where shifting the camera position over allows you to see something behind an object that wasn't visible from the other eye position. In these cases the information simply doesn't exist, which means you have to "make it up". Probably for a lot of scenes (especially scenes where the camera or the geometry is moving a lot) you could probably do it well enough that it wouldn't be seriously noticeable. But for a long slow pan where there's something in the foreground that partly occludes the background...I don't know what you could do to not make it look like crap. You can see it really badly in this pic from that blog. I'd definitely like to see what CryTek is doing for these cases (if anything at all).
 
MJP said:
Any technique that doesn't render separate images for both eyes is going to have artifacts for certain cases, mainly where shifting the camera position over allows you to see something behind an object that wasn't visible from the other eye position. In these cases the information simply doesn't exist, which means you have to "make it up". Probably for a lot of scenes (especially scenes where the camera or the geometry is moving a lot) you could probably do it well enough that it wouldn't be seriously noticeable. But for a long slow pan where there's something in the foreground that partly occludes the background...I don't know what you could do to not make it look like crap. You can see it really badly in this pic from that blog. I'd definitely like to see what CryTek is doing for these cases (if anything at all).
Click to expand...

"Crysis 2 uses a concave 3D effect so that objects do not pop out of the screen, which gives the game “more depth” according to Yerli."

I think that means that they 'push the whole scene back', reducing the 3d effect, which in turn reduces the 'gaps'?
 
You must log in or register to reply here.

Similar threads

F
Dedicated gaming GPUs bottom of the priority list?
2
Replies
34
Views
3K
Seanspeed
S
F
RDNA5 Speculation
Replies
2
Views
1K
Erinyes
E
R
Gaming and gaming hardware trends
2
Replies
30
Views
2K
eastmen
E
B
Game Engine Convergence And The Problem With UE5
4 5 6
Replies
104
Views
17K
cheapchips
C
Cyan
Playtron : new Linux based platform-agnostic open gaming OS
Replies
10
Views
2K
MfA
M
Back
Top