An editorial by Rob Fahey looks into the possibility of game companies experimenting with modern 3D technology. Over the past decade, advances in the field have been enough for film studios to give 3D another shot, but significant price-related hurdles remain when considering individual consumers. Quoting:
"[The approach presently favored by game makers] has actually been around for some time. It displays the image for the left eye, then the right eye, in quick succession on screen — while the glasses you wear close LCD 'shutters' over your eyes so that each eye only sees the appropriate image. If this is done fast enough, the brain sees no flicker — just a continuous, steady 3D image. The best thing about this final approach is that some televisions already exist which could, in theory, support it. No new display technology is required, but what you do need is a TV screen which can display twice the number of frames per second as a normal screen — since you now need one frame for each eye, where previously you had one frame for both eyes. You also need LCD glasses synched to the television's refresh rate for each viewer. All of this lies in the realms of being moderately plausible."
This discussion has been archived.
No new comments can be posted.
Developers have always loved it, but the glasses have been fragile and expensive. Might as well wait for 3d tv first, then you'll have cheap glasses. Everyone will get a 3d tv when it comes out, when the software (ie movies, tv show) is available and the price is right.
last time i saw a serious discussion about this sort of thing, it came up that when you use 3D instead of 2D, alot of special effect break. E.G.It becomes quite obvious that that choreographed punch went behind the other actors head. Will the worry of special/camera effects breaking put tv/film producers off encouraging the leap to 3D?
A lot of computer game effects - smoke, bloom, etc, are also post-processing effects, and care will need to be taken to make sure that these look correct in a 3D viewing environment.
Yep. The smoke, fire, and (more annoyingly) car glint in GTA: Vice City all render at screen depth. I imagine many postprocessing effects suffer similarly. The HUD for first-person shooters usually renders at screen depth (which is the best place for it to be), but the nVidia stereoscopic driver operates by rendering from two eyepoints equidistant from center, so using the crosshair is like lining up a gunsight to your nose. There was some kind of a workaround where the driver would attempt to render a
Will the worry of special/camera effects breaking put tv/film producers off encouraging the leap to 3D?
Short answer: yes.
Long answer: As an anecdote, a visual effects artist is adding rain to a shot. With a normal film, this would be a simple matter of compositing. For a 3D film it would be very obvious if every raindrop was on a single plane, so the artist has to analyze the scene to compute depth of field of visible actors/props, and generate 3D rain that doesn't cause cognitive dissonance (a raindrop tha
If you look closely at some (not all) 3D movies, you'll notice that anything beyond the near foreground is flat. Instead of rendering everything from two viewpoints, a series of depth planes are rendered, and these planes are shifted horizontally go get the stereoscopic effect. It cuts down on render time significantly, is (relatively) easy to bodge in using the z-buffer, and you can still cheat somewhat with layers of planar SFX rendering.
last time i saw a serious discussion about this sort of thing, it came up that when you use 3D instead of 2D, alot of special effect break. E.G.It becomes quite obvious that that choreographed punch went behind the other actors head. Will the worry of special/camera effects breaking put tv/film producers off encouraging the leap to 3D?
If this means the actors will have to actually take the punch now, I think I might start enjoying Nicholas Cage movies.
It's not that hard but it's less predictable and the beancounters would prefer predictable improvements they can plan around. Of course they don't seem to plan around the bit where dev costs are ballooning too fast and revenues can't keep up...
I don't care about 3D graphics, but 60Hz LCDs are inadequate for displaying perfectly smooth blur-free motion. 120Hz LCDs can show almost as high quality motion as a CRT, so I look forward to them becoming widely available.
A big part of the LCD blur is caused by the sample and hold light characteristic. Watch some 60Hz content frame tripled on a 180Hz CRT (0ms response time) and you'll see almost identical blur to a fast LCD. Switching to 120Hz reduces this blur by half and well as improving the smoothness of the motion. For legacy 60Hz content you can either interpolate new frames (causing motion artifacts and increasing latency), use blank frame insertion (trading blur for flicker), or double the frames (keeping the SAH
Eureka! Thank you. The blur is created by the eye, not the display.
If your eye is following a moving object on a sample-and-hold display, there will be a discrepancy between where your eye expects the object to move to and where the object is displayed on screen. This discrepancy is the blur.
Or on second thoughts, if you're happy with current video technology then you might want to pretend you never read this. Once you've seen the flaws you can't unsee them. Video standards are a ugly mass of hacks, so the more you learn the more annoyed you'll be.
low takeup (Score:1)
Developers have always loved it, but the glasses have been fragile and expensive. Might as well wait for 3d tv first, then you'll have cheap glasses. Everyone will get a 3d tv when it comes out, when the software (ie movies, tv show) is available and the price is right.
Re: (Score:2, Interesting)
Not me, I have no depth perception.
special effects become an issue (Score:4, Interesting)
Re: (Score:2)
Yes.
-:sigma.SB
Re: (Score:1)
A lot of computer game effects - smoke, bloom, etc, are also post-processing effects, and care will need to be taken to make sure that these look correct in a 3D viewing environment.
Yep. The smoke, fire, and (more annoyingly) car glint in GTA: Vice City all render at screen depth. I imagine many postprocessing effects suffer similarly. The HUD for first-person shooters usually renders at screen depth (which is the best place for it to be), but the nVidia stereoscopic driver operates by rendering from two eyepoints equidistant from center, so using the crosshair is like lining up a gunsight to your nose. There was some kind of a workaround where the driver would attempt to render a
Re: (Score:3, Insightful)
Short answer: yes.
Long answer: As an anecdote, a visual effects artist is adding rain to a shot. With a normal film, this would be a simple matter of compositing. For a 3D film it would be very obvious if every raindrop was on a single plane, so the artist has to analyze the scene to compute depth of field of visible actors/props, and generate 3D rain that doesn't cause cognitive dissonance (a raindrop tha
Re: (Score:2)
Re: (Score:2)
last time i saw a serious discussion about this sort of thing, it came up that when you use 3D instead of 2D, alot of special effect break. E.G.It becomes quite obvious that that choreographed punch went behind the other actors head. Will the worry of special/camera effects breaking put tv/film producers off encouraging the leap to 3D?
If this means the actors will have to actually take the punch now, I think I might start enjoying Nicholas Cage movies.
Re: (Score:2)
Re: (Score:1)
It's not that hard but it's less predictable and the beancounters would prefer predictable improvements they can plan around. Of course they don't seem to plan around the bit where dev costs are ballooning too fast and revenues can't keep up...
I hope this succeeds (Score:4, Insightful)
Re: (Score:1)
A 120 Hz LCD with a response time of 8 ms would be a lot blurrier than a 60 Hz LCD with a response time of 2 ms.
Re: (Score:2)
Re: (Score:1)
The above post doesn't make sense to me.
I just don't see how repeating frames on a CRT increases blurriness.
I also don't see how sampling-and-holding increases blurriness on my hypothetical 0 ms LCD compared to using blank-frame insertion.
Can someone explain this to me?
Re: (Score:2)
Re: (Score:1)
Eureka! Thank you. The blur is created by the eye, not the display.
If your eye is following a moving object on a sample-and-hold display, there will be a discrepancy between where your eye expects the object to move to and where the object is displayed on screen. This discrepancy is the blur.
Learning is good.
Re: (Score:1)
Wow, what a brilliant new idea! (Score:3, Interesting)
just saw a 3d movie this weekend (Score:2)