real time render when?

[bloodbob]

http://dictionary.reference.com/search?q=render

Computer Science. To convert (graphics) from a file into visual form, as on a
video display.

Hmm I think a few people here must be using lynx or some other text render.

 

[Sage]

Hmm I think a few people here must be using lynx or some other text render.

yeah, my browser renders damn near close to every page I visit

 

[Alejux]

Are you guys lawyers? Only a lawyer has this unique ability to exploit semantics so well and completely twist an argument removing all of it's original spirit.

 

[bloodbob]

Are you guys lawyers? Only a lawyer has this unique ability to exploit semantics so well and completely twist an argument removing all of it's original spirit.

No but I served on a jury of a torture case and saw how it works.

http://tvnz.co.nz/view/news_world_story_skin/473070?format=html

 

[KimB]

Oh, my. You guys must have violent video games over there in New Zealand, too.

 

[Dave B(TotalVR)]

Which end of "real-time" are you talking about? 1fps? 10fps? 100fps? They're all real-time, just different types of real.

Real is just electrical signals interpreted by your brain.

 

[flf]

A few months back I read a Robert X. Cringley "Pulpit" article (from the PBS site) about how the optic nerve's data rate is less than a 56K modem. He followed up this fact with the supposition that it must be possible to pack HDTV into a 56K stream somehow. Now, I can see how it might surficially seem like a correlation that this would be possible, but it completely neglects the fact that, while I may be looking at the upper right corner of a frame, you might be looking at the lower left corner of a frame. So if you think about it, you need a 56K stream of optic data for each set of eyes looking at a frame.

Now here's where the suppositions get tricky: What if you have a computer display that is always centered in the eye (tiny tft contact lenses, if you will) and the computer tracks where your eyes pan in order to determine what we're looking at. How much detail do you have to input into the eye to make it seem as if you're looking at a real scene? Obviously a circular display would be necessary, and the pixel/detail at the center would have to be good, while the peripheral area could be less detailed.

How much processing power do you need for "real time photorealism" if you reduce the display so it matches the true information gathering capability of the eye?

(When you are staring at these individual words, does your brain really need the peripheral detail of the rest of the monitor that you aren't directly looking at?)

Yes, I know this goes waaaay into the realm of science fiction at this point.

 

[flf]

Which end of "real-time" are you talking about? 1fps? 10fps? 100fps? They're all real-time, just different types of real.

Real is just electrical signals interpreted by your brain.

I think the word we're searching for here is "interactive" real-time photorealistic graphics. With that "interactive" word typically indicating frame rates that are fluid enough that we aren't distracted by them.

 

[Killer-Kris]

Yes, I know this goes waaaay into the realm of science fiction at this point.

Not entirely, at least a lot of the display type technology you described is being researched in attempting to reverse some types of blindness. Once you have a working hook-up into the optic nerve, or can stimulate the rods and cones directly, Ati and Nvidia will definitely have to look into all the ideas you just pointed out.

 

[KimB]

Well, eye-tracking is pretty easy. But making a display that would move as fast as the eye would be pretty much impossible (you'd need to have the movement be artificial, i.e. have a full-resolution display in all directions, but only render the highest detail where the user is looking. But then it becomes a question as to whether it's really worth it).

Regardless, the primary problem with such a system would still be that the eye is not digital, whereas pretty much anything connected to a computer still will be in some sense. So you would still need more information sent to the display than your eye receives.

Side comment: I don't think it's so much science fiction as a hypothetical question.

 

[Dave B(TotalVR)]

Which end of "real-time" are you talking about? 1fps? 10fps? 100fps? They're all real-time, just different types of real.

Real is just electrical signals interpreted by your brain.

I think the word we're searching for here is "interactive" real-time photorealistic graphics. With that "interactive" word typically indicating frame rates that are fluid enough that we aren't distracted by them.

you missed the matrix referece then

 

[flf]

Well, eye-tracking is pretty easy. But making a display that would move as fast as the eye would be pretty much impossible (you'd need to have the movement be artificial, i.e. have a full-resolution display in all directions, but only render the highest detail where the user is looking.

That sort of a assumes a normal monitor arrangement, which is why I mentioned contact lenses... the display stays centered in the eye no matter where it looks. But, of course, your eye cannot focus at that point, but you get my meaning.

Hmmm, that being said... what if you did use a normal monitor, and tracked where the person was looking, and only applied maximum detail to that point where they were looking? It's probably possible, but would it be of any use? Probably injects more complexity than the problems it attempts to solve.

 

[Dave B(TotalVR)]

Yes, I know this goes waaaay into the realm of science fiction at this point.

Not entirely, at least a lot of the display type technology you described is being researched in attempting to reverse some types of blindness. Once you have a working hook-up into the optic nerve, or can stimulate the rods and cones directly, Ati and Nvidia will definitely have to look into all the ideas you just pointed out.

Ive seen televised medical experiments where blind people have been able to see a grid of dots , the information of which is sent into the brain in much the same way as a cochlea implant does IIRC.

It was all about the next step being fitting a false eye with a camera and all the required electrical gubbinz. Right now it would only be possible if you had eyes like dinner plates in which case blindness is not really your primary concern

 

[flf]

you missed the matrix referece then

I didn't see the word "spoon" used. There were other lines in that movie that people quote?

 

[Dave B(TotalVR)]

Well, eye-tracking is pretty easy. But making a display that would move as fast as the eye would be pretty much impossible (you'd need to have the movement be artificial, i.e. have a full-resolution display in all directions, but only render the highest detail where the user is looking. But then it becomes a question as to whether it's really worth it).

Regardless, the primary problem with such a system would still be that the eye is not digital, whereas pretty much anything connected to a computer still will be in some sense. So you would still need more information sent to the display than your eye receives.

Side comment: I don't think it's so much science fiction as a hypothetical question.

So like a contact lense that you wear which is actually a tft screen. backlighting would be easy. You then track the eye movements and alter the piont of view of the camera accordingly.

.......Once we make an lcd you can place on your eyeball and not be able to tell it is there.

 

[Dave B(TotalVR)]

you missed the matrix referece then

I didn't see the word "spoon" used. There were other lines in that movie that people quote?

See my previous post, yes, I do

Smell good, dont they

 

[Killer-Kris]

Hmmm, that being said... what if you did use a normal monitor, and tracked where the person was looking, and only applied maximum detail to that point where they were looking? It's probably possible, but would it be of any use? Probably injects more complexity than the problems it attempts to solve.

God, could you see the new flame wars starting about which IHV optimized too much, but when in fact the reviewer just needs new glasses.

 

[KimB]

So like a contact lense that you wear which is actually a tft screen. backlighting would be easy. You then track the eye movements and alter the piont of view of the camera accordingly.

Can't focus on something that's located on the surface of the eye, so need to have it some distance away (as flf said).

 

[Dave B(TotalVR)]

Hmmm, that being said... what if you did use a normal monitor, and tracked where the person was looking, and only applied maximum detail to that point where they were looking? It's probably possible, but would it be of any use? Probably injects more complexity than the problems it attempts to solve.

God, could you see the new flame wars starting about which IHV optimized too much, but when in fact the reviewer just needs new glasses.

 

[Dave B(TotalVR)]

So like a contact lense that you wear which is actually a tft screen. backlighting would be easy. You then track the eye movements and alter the piont of view of the camera accordingly.

Can't focus on something that's located on the surface of the eye, so need to have it some distance away (as flf said).

Hmm, I think holography is the answer.