30" monitor = 2560x1600 resolution in this story. Any smaller monitor, and most larger ones, will only go to 1920x1200 or 1920x1080.
IBM had 3840x2400 monitors, but I haven't seen much about them lately. ISTR that they required multiple specialized video cards to drive a single display. I think AMD was trying to make the point that, should displays with resolution this high become widely available, they'll have no trouble driving them.
Actually, seriously, it seems like it would be more useful to have a standard 30" display centered in your FOV, and a projected 90" display surrounding it at lower resolution. You still get the peripheral cues, but you're not wasting resolution (and expense) on parts of the display where you can't perceive it. The math and logic is fairly simple, but I've never heard of a card that supports it. (There were some esoteric simulators many years ago that did this, but it never caught on in the wider market.)
I don't insist on creatures that can actually alter the half-life of radioisotopes. Just ones that ingest them, do isotopic separation, and excrete the separated isotopes into segregated containers. We can let the uranium and hydrogen/deuterium separators drive the economy, with all the other separators running as boutique suppliers.
I love the notion of writing off most of the thermodynamic inefficiencies of an IC engine by using it to replace an existing furnace. But an IC engine requires considerably more maintenance than the typical furnace. Deploy 100 000 of these, and you're going to need an awful lot of mechanics.
Give me a way to expand time, so I can experience an hour of clear, cogent subjective thought and action in five or ten minutes of "real" time, and I'll be all over it.
6. Repeat steps 1-5, producing a second, identical ball covered in magnets.
7. Roll the balls past one another.
8. Observe that there's no net repulsion between them.
9. Go back to the parts of the electromagnetism textbook you skipped previously.
Let's see if the third time's the charm. The linked article goes into detail on how "real" monopoles would be expected to catalyze proton (and neutron) decay. It's not deep detail, but I have a feeling that any explanation much deeper would be lost on anyone but theoretical physicists.
If they were real, physical, isolable monopoles, they might turn out to have some minor applications in energy production. (Yeah, I linked the same article upthread; it's interesting enough to repeat.) The claim is that they would make protons (and neutrons) decay promptly. Of course, if these folks were seeing that kind of monopole, they would have noticed side effects, starting with a sudden inability to keep their samples below 1 K.
I've been reading for decades about the search for subatomic-particle-type monopoles, and all the wondrous things one could do with them. This sounds more like some kind of group phenomenon, an emulation of a monopole, if you will. Sort of like holes in a semiconductor, which behave in some ways like positive "things", but are actually just the absence of an electron in a lattice.
I'm guessing that these aren't the kind of "real" monopoles that would let us build super-powerful motors, or compact proton disintegrators, or whatnot. On the other hand, even though the semiconductor folks can't isolate and sell bucketloads of holes, they do turn out to be quite useful.
If you could build a printer yourself, it would be more than ten times slower than a commercial printer, probably have ten times poorer resolution, and cost more than ten times as much.
The DIY crowd has wisely skipped over 2D printers, and moved directly to 3D fabricators.
No, not really. Photons have to come from somewhere, and they travel in straight lines unless they're diffracted, refracted or reflected. For holograms, you've got to have a panel either in front of or behind the virtual image. Same thing for "3D TV", either with or without glasses. You can focus enough laser power to make the air itself go nonlinear and emit or reflect light, but lasers at that power level entail, shall we say, certain risks. And at that point, you could argue that the air has become a "volatile gas".
So, sorry, no hologram projectors that shoot a three-foot 3D image into midair from a little lens.
Where? They certainly aren't different shapes in the US. Here, at least in my part, it's positional (red at top, green at bottom), and there's enough blue in the green light to let most people distinguish them.
Still doesn't address the broad cultural prevalence here of red/green status indicators, which have persisted even though ~5% of the population has trouble distinguishing them.
This tech will fail because simply sending two image streams isn't good enough -- it encodes assumptions about eye spacing, viewing distance and angle that are too restrictive. People aren't going to jump for a system that shows you a distorted and headache-inducing scene if you aren't sitting precisely in front of the center of the screen.
I've tried out a more sophisticated system that generates five points of view (from a 3D model) and fans them out with optics that don't require glasses. This greatly reduces the viewing-angle problem -- but it STILL sucks, because shifting between discrete views as you move your head is too disorienting, and because with current tech generating five views reduces your resolution by a factor of 5.
Real 3D won't dominate until it's being fed to a head-mounted display (or the equivalent), and/or we're shipping true 3D data (not just two fixed viewpoints).
That was what they called it at the time, and I think I got it in writing. They were just rolling out DSL, so I wouldn't be surprised if they were borrowing terminology from more familiar technologies. (I was working for a network-analysis software company at the time, so I was asking about "committed data rate".)
When I subscribed back in 1999, Verizon only offered 768k down/128k up, and the CIR was 16 Kbps bidirectional. That's right -- they promised that my connection would be at least almost half as fast as a 33.6K modem. Except, of course, when it wasn't working.
I don't think this will be the killer issue. They could use b-field resonant coupling, which loses a lot less energy into the environment, and use frequencies where body tissue is mostly transparent. I'd expect the effective power densities to be lower than what you get from a transmitting cell phone. (It looks like RF exposure limits are in the range of 1mW/cm^2, although it's a complicated issue.)
Anything like this will have to be pretty low-powered anyhow -- I don't know how much you can safely heat the eye, but we could start with existing regulations for IR exposure. This chart indicates that the maximum permissible exposure for longwave IR over an extended period is 100mW/cm^2; at those wavelengths, all that energy is converted into heat pretty much at the surface, so that's not a bad model for something that sits on the cornea. 100mW is actually a lot more than I expected we could get away with.
BTW, Indefensible Positions rocked. Thanks for making it.
Numbers, please. You're moving a significant mass of tissue down a centimeter, then back up, in the space of 100ms. It doesn't involve much energy -- but neither does a solar cell covering a few square mm, or an inductive loop less than a cm in diameter. And remember, as others have pointed out, you don't want this thing to use too much energy, because it's sitting on an eyeball, which closely resembles an egg in terms of "ease of cooking".
Even the much-derided "body heat" might be useful here, because you've reliably got a warm side (against the eyeball) and a cold side (facing the air and covered with a water-based film). Of course, you lose if it's too hot or humid, or if you close your eyes. And you'd be looking at just a few degrees of differential at best, which implies efficiencies below 1%.
Maybe, but that's an awfully complex problem, especially since you only want to correct for some of the imperfections -- the ones your brain isn't already dynamically mapping out. I don't know which will come first, the technology to do that, or the technology to go directly into the cortex.
Cool link -- thanks! It's easy to see why some people conclude it was all intelligently designed -- especially in light of the Gospel delivered by His Son, Rube Goldberg.
I think I see what you're missing. The contact lens is so close to the eye's lens that it doesn't really matter exactly where the lens sits. Yes, it drifts around, as you've seen on video; no, that doesn't make a significant difference in what you see.
Toric lenses are heavier at the bottom, so they naturally orient themselves properly. They'll be off by a few degrees more often than not, but for correcting astigmatism, that's completely unnoticeable.
Bifocal contact lenses can't work like bifocal glasses, because there's no concept of looking through the lower or upper half of the lens. You're looking through the whole lens, all the time. (More precisely, you're looking through the whole lens when it's dark and your pupil is dilated, and through only the center of the lens when it's bright and your pupil is small.)
For a lens to display an image, it's got to have optics that make the image appear distant. Yes, it'll have to keep track of its orientation and correct for it -- but that's just part of the localization problem you have to solve anyway for AR to work.
Fun trivia fact, BTW: In addition to moving left-right and up-down, your eye rotates within a limited but significant range to keep the same orientation as your head tilts left or right. I called BS the first time I read this, but I looked carefully in a mirror, and it's true.
Ugh. 100ms is nasty on all sorts of levels. Yes, the processing latency inside your head is worse than that in general -- but it's all finely tuned in wetware that's evolved over millions of years, and for the most part you can't just "adapt" to throwing another 100ms into the stovepipe. (Conversely, once we start tapping more directly into the visual cortex, taking away latency will be similarly fraught with risk.)
OTOH, we've already got consumer cameras doing 1000fps capture, and there's no reason in principle that we can't increase display refresh to arbitrarily high rates. If we can get down below 10ms total latency, I don't think that delay will affect anyone but elite athletes, who won't be allowed to use augments in competition anyhow. (Or if they are, visual augments will be only a minor component.)
For haptics (sensing touch/contact/pressure), even 1ms delay can be too long. For visual loops, though, you get a little more leeway.
Slashdot Mirror
30" monitor = 2560x1600 resolution in this story. Any smaller monitor, and most larger ones, will only go to 1920x1200 or 1920x1080.
IBM had 3840x2400 monitors, but I haven't seen much about them lately. ISTR that they required multiple specialized video cards to drive a single display. I think AMD was trying to make the point that, should displays with resolution this high become widely available, they'll have no trouble driving them.
Wrong direction. You need NINE displays.
Actually, seriously, it seems like it would be more useful to have a standard 30" display centered in your FOV, and a projected 90" display surrounding it at lower resolution. You still get the peripheral cues, but you're not wasting resolution (and expense) on parts of the display where you can't perceive it. The math and logic is fairly simple, but I've never heard of a card that supports it. (There were some esoteric simulators many years ago that did this, but it never caught on in the wider market.)
I'll bet I can't get more than two of them into my machine, which means I'm still stuck with a maximum of 12 monitors. Dammit.
I don't insist on creatures that can actually alter the half-life of radioisotopes. Just ones that ingest them, do isotopic separation, and excrete the separated isotopes into segregated containers. We can let the uranium and hydrogen/deuterium separators drive the economy, with all the other separators running as boutique suppliers.
I love the notion of writing off most of the thermodynamic inefficiencies of an IC engine by using it to replace an existing furnace. But an IC engine requires considerably more maintenance than the typical furnace. Deploy 100 000 of these, and you're going to need an awful lot of mechanics.
Give me a way to expand time, so I can experience an hour of clear, cogent subjective thought and action in five or ten minutes of "real" time, and I'll be all over it.
And I promise not to use it for first posts.
6. Repeat steps 1-5, producing a second, identical ball covered in magnets.
7. Roll the balls past one another.
8. Observe that there's no net repulsion between them.
9. Go back to the parts of the electromagnetism textbook you skipped previously.
Let's see if the third time's the charm. The linked article goes into detail on how "real" monopoles would be expected to catalyze proton (and neutron) decay. It's not deep detail, but I have a feeling that any explanation much deeper would be lost on anyone but theoretical physicists.
...eliminates the soul-sucking ennui of day-to-day life.
I think they're missing the point.
If they were real, physical, isolable monopoles, they might turn out to have some minor applications in energy production. (Yeah, I linked the same article upthread; it's interesting enough to repeat.) The claim is that they would make protons (and neutrons) decay promptly. Of course, if these folks were seeing that kind of monopole, they would have noticed side effects, starting with a sudden inability to keep their samples below 1 K.
I've been reading for decades about the search for subatomic-particle-type monopoles, and all the wondrous things one could do with them. This sounds more like some kind of group phenomenon, an emulation of a monopole, if you will. Sort of like holes in a semiconductor, which behave in some ways like positive "things", but are actually just the absence of an electron in a lattice.
I'm guessing that these aren't the kind of "real" monopoles that would let us build super-powerful motors, or compact proton disintegrators, or whatnot. On the other hand, even though the semiconductor folks can't isolate and sell bucketloads of holes, they do turn out to be quite useful.
If you could build a printer yourself, it would be more than ten times slower than a commercial printer, probably have ten times poorer resolution, and cost more than ten times as much.
The DIY crowd has wisely skipped over 2D printers, and moved directly to 3D fabricators.
No, not really. Photons have to come from somewhere, and they travel in straight lines unless they're diffracted, refracted or reflected. For holograms, you've got to have a panel either in front of or behind the virtual image. Same thing for "3D TV", either with or without glasses. You can focus enough laser power to make the air itself go nonlinear and emit or reflect light, but lasers at that power level entail, shall we say, certain risks. And at that point, you could argue that the air has become a "volatile gas".
So, sorry, no hologram projectors that shoot a three-foot 3D image into midair from a little lens.
Where? They certainly aren't different shapes in the US. Here, at least in my part, it's positional (red at top, green at bottom), and there's enough blue in the green light to let most people distinguish them.
Still doesn't address the broad cultural prevalence here of red/green status indicators, which have persisted even though ~5% of the population has trouble distinguishing them.
This tech will fail because simply sending two image streams isn't good enough -- it encodes assumptions about eye spacing, viewing distance and angle that are too restrictive. People aren't going to jump for a system that shows you a distorted and headache-inducing scene if you aren't sitting precisely in front of the center of the screen.
I've tried out a more sophisticated system that generates five points of view (from a 3D model) and fans them out with optics that don't require glasses. This greatly reduces the viewing-angle problem -- but it STILL sucks, because shifting between discrete views as you move your head is too disorienting, and because with current tech generating five views reduces your resolution by a factor of 5.
Real 3D won't dominate until it's being fed to a head-mounted display (or the equivalent), and/or we're shipping true 3D data (not just two fixed viewpoints).
Probably there are enough people for whom this is true that "3D" display technology based on 2D devices will fail in the marketplace.
Just as red/green status and traffic lights have failed because of the wide prevalence of red/green colorblindness?
It's a binocular world out there, and I don't think the rate of anomalous depth perception is high enough to change that.
That was what they called it at the time, and I think I got it in writing. They were just rolling out DSL, so I wouldn't be surprised if they were borrowing terminology from more familiar technologies. (I was working for a network-analysis software company at the time, so I was asking about "committed data rate".)
Some DSL providers apparently still use the CIR terminology.
When I subscribed back in 1999, Verizon only offered 768k down/128k up, and the CIR was 16 Kbps bidirectional. That's right -- they promised that my connection would be at least almost half as fast as a 33.6K modem. Except, of course, when it wasn't working.
Yeah, it's perfectly calm, as long as you ignore the triangular striated prints in the snow. And the disappearing sled dogs.
I don't think this will be the killer issue. They could use b-field resonant coupling, which loses a lot less energy into the environment, and use frequencies where body tissue is mostly transparent. I'd expect the effective power densities to be lower than what you get from a transmitting cell phone. (It looks like RF exposure limits are in the range of 1mW/cm^2, although it's a complicated issue.)
Anything like this will have to be pretty low-powered anyhow -- I don't know how much you can safely heat the eye, but we could start with existing regulations for IR exposure. This chart indicates that the maximum permissible exposure for longwave IR over an extended period is 100mW/cm^2; at those wavelengths, all that energy is converted into heat pretty much at the surface, so that's not a bad model for something that sits on the cornea. 100mW is actually a lot more than I expected we could get away with.
BTW, Indefensible Positions rocked. Thanks for making it.
Numbers, please. You're moving a significant mass of tissue down a centimeter, then back up, in the space of 100ms. It doesn't involve much energy -- but neither does a solar cell covering a few square mm, or an inductive loop less than a cm in diameter. And remember, as others have pointed out, you don't want this thing to use too much energy, because it's sitting on an eyeball, which closely resembles an egg in terms of "ease of cooking".
Even the much-derided "body heat" might be useful here, because you've reliably got a warm side (against the eyeball) and a cold side (facing the air and covered with a water-based film). Of course, you lose if it's too hot or humid, or if you close your eyes. And you'd be looking at just a few degrees of differential at best, which implies efficiencies below 1%.
Maybe, but that's an awfully complex problem, especially since you only want to correct for some of the imperfections -- the ones your brain isn't already dynamically mapping out. I don't know which will come first, the technology to do that, or the technology to go directly into the cortex.
Cool link -- thanks! It's easy to see why some people conclude it was all intelligently designed -- especially in light of the Gospel delivered by His Son, Rube Goldberg.
I think I see what you're missing. The contact lens is so close to the eye's lens that it doesn't really matter exactly where the lens sits. Yes, it drifts around, as you've seen on video; no, that doesn't make a significant difference in what you see.
Toric lenses are heavier at the bottom, so they naturally orient themselves properly. They'll be off by a few degrees more often than not, but for correcting astigmatism, that's completely unnoticeable.
Bifocal contact lenses can't work like bifocal glasses, because there's no concept of looking through the lower or upper half of the lens. You're looking through the whole lens, all the time. (More precisely, you're looking through the whole lens when it's dark and your pupil is dilated, and through only the center of the lens when it's bright and your pupil is small.)
For a lens to display an image, it's got to have optics that make the image appear distant. Yes, it'll have to keep track of its orientation and correct for it -- but that's just part of the localization problem you have to solve anyway for AR to work.
Fun trivia fact, BTW: In addition to moving left-right and up-down, your eye rotates within a limited but significant range to keep the same orientation as your head tilts left or right. I called BS the first time I read this, but I looked carefully in a mirror, and it's true.
Ugh. 100ms is nasty on all sorts of levels. Yes, the processing latency inside your head is worse than that in general -- but it's all finely tuned in wetware that's evolved over millions of years, and for the most part you can't just "adapt" to throwing another 100ms into the stovepipe. (Conversely, once we start tapping more directly into the visual cortex, taking away latency will be similarly fraught with risk.)
OTOH, we've already got consumer cameras doing 1000fps capture, and there's no reason in principle that we can't increase display refresh to arbitrarily high rates. If we can get down below 10ms total latency, I don't think that delay will affect anyone but elite athletes, who won't be allowed to use augments in competition anyhow. (Or if they are, visual augments will be only a minor component.)
For haptics (sensing touch/contact/pressure), even 1ms delay can be too long. For visual loops, though, you get a little more leeway.