dmocsny@minerva.che.uc.edu (Daniel Mocsny) (11/17/90)
In article <16870003@hpfcdj.HP.COM> jayavant@hpfcdj.HP.COM (Rajeev Jayavant) writes: >/ hpfcdj:comp.arch / gillies@m.cs.uiuc.edu / 7:22 pm Nov 12, 1990 / >>> Why don't we see monitors that are 2048*1532, at 150 dpi? Would the >>> radiation from the high-frequency modulators fry your brains? > >I haven't priced 2048*1532 monitors recently, but they were in at >least the $20K ballpark not long ago. You also now have roughly 4x >the pixels as before, so you have to render 4x as fast just to have >the appearance of equivalent performance. Graphics hardware cost >starts going up pretty fast compared to today's megapixel displays. This is an issue of some interest to me, because I consider display performance to be one of the major limiting factors in the usefulness of computers. (Consider the longevity of paper as a medium for information display, despite its many disadvantages. Its primary advantage is that it can display much more information simultaneously than any computer display, and at a cost that can be very low.) I am aware that graphics display technology has not enjoyed the same performance/price gains that computer technology in general has. Would some informed person care to comment on the reasons for this, and speculate on the likely future costs of display technology? In particular, what is the scope for cost reductions through miniaturizing? In audio technology, a $100 pair of headphones can easily out-perform a $1000 pair of loudspeakers. Can "eyephones" such as the Virtual-Reality people are working on improve display hardware performance/price similarly? If display technology does not keep up with progress in processors, memory, and storage, then in time it will emerge as the limiting factor in most applications. An 80x25 character display is reasonable when you only have a few KB of storage space available. But when every PC is packing >1 GB of disk and >64 MB of RAM, the display "window" becomes relatively smaller and more constraining. >You also now have roughly 4x >the pixels as before, so you have to render 4x as fast just to have >the appearance of equivalent performance. Schemes exist to reduce the processing cost of pixel scale-up. These take advantage of the resolving characteristics of the human eye. The eye has much higher resolution near its focus, and lower resolution in the periphery of the field of view. Thus, with eye-tracking technology, a wide-field display can give the appearance of high quality by rendering in maximum resolution the portion of the display near the viewer's focus. The rest of the display can be in lower resolution, without loss of apparent quality. To an observer staring over the shoulder of the computer user, such an efficient display would seem somewhat blurry, save for a circular region that would stand out in crisp clarity. The circular region would appear to flit around rapidly, as the user's eyes aimed at different portions of the display. -- Dan Mocsny Snail: Internet: dmocsny@minerva.che.uc.edu Dept. of Chemical Engng. M.L. 171 dmocsny@uceng.uc.edu University of Cincinnati 513/751-6824 (home) 513/556-2007 (lab) Cincinnati, Ohio 45221-0171
spot@WOOZLE.GRAPHICS.CS.CMU.EDU (Scott Draves) (11/17/90)
In article <6726@uceng.UC.EDU>, dmocsny@minerva.che.uc.edu (Daniel Mocsny) writes: |> (Consider the longevity of paper as a medium for |> information display, despite its many disadvantages. Its primary |> advantage is that it can display much more information simultaneously |> than any computer display, and at a cost that can be very low.) I think there is an additional, and more important, barrier to paper's demise. copyright. If the ideal display were announced tomorrow, paper would still exist because you can't get information from publishers in digital/on-line form. This isn't being rectified as we speak, either. |> In audio technology, a $100 pair of headphones can |> easily out-perform a $1000 pair of loudspeakers. only for the deaf. headphones don't reproduce low frequencies well. Even good electrostatic ones. |> Can "eyephones" |> such as the Virtual-Reality people are working on improve display |> hardware performance/price similarly? I've read about an 80x25 character head mounted display that is small, light, and cheap. It used a single column of LEDs and a rotating (or vibrating?) mirror. Don't remember more than that... |> The eye has much higher resolution near its focus, and lower |> resolution in the periphery of the field of view. Thus, with eye-tracking |> technology, a wide-field display can give the appearance of high |> quality by rendering in maximum resolution the portion of the display |> near the viewer's focus. This technique is used in high-performance multi-million $ flight simulators, where the screen is being constantly recomputed. It doesn't really apply if you're using something like X. It's esoteric and expensive in any case. Consume Scott Draves Be Silent spot@cs.cmu.edu Die
lindsay@gandalf.cs.cmu.edu (Donald Lindsay) (11/17/90)
In article <6726@uceng.UC.EDU> dmocsny@minerva.che.uc.edu (Daniel Mocsny) writes: >I am aware that graphics display technology has not enjoyed the >same performance/price gains that computer technology in general >has. It has improved a lot. I remember a high-end graphics system of the early 70's: controlling the CRT beam strokes took about a cubic meter of analog circuits. (Gad, the number of screwdriver adjusts.) When RAM got cheap, everyone switched to raster graphics, a much more manufacturable technology. At first, hi-res raster systems put most of the $ into RAM: now that RAM is cheap, the $ are in things that aren't improving as fast. That doesn't mean that we're stuck, that means that it's time for another change. One possibility is liquid crystal. Some researchers are doing color via LC shutters over grayscale CRTs; this eliminates the fabrication/resolution issues of a color CRT's color mask. Some researchers use a small LC image, and use optics to make a large projection of it. There are fun projects writing luminous dots with lasers; the plasma panel people haven't given up; and in general there are a zoo of experiments and hopes that could still make it. In summary, I'm optimistic. All the labs know how valuable the right patent could be: they are spending the money to search for it. >Can "eyephones" >such as the Virtual-Reality people are working on improve display >hardware performance/price similarly? Yes. The 2.5-ounce product that's available now from Reflection Technology is of course PC-compatible: CGA monochrome. The designer claims that the resolution will improve, and he has hopes for color. From the technology he's using, that sounds like a fair assessment. The price-performance is also boosted by the fact that it's cheap. -- Don D.C.Lindsay
gd@geovision.uucp (Gord Deinstadt) (11/17/90)
>>>> Why don't we see monitors that are 2048*1532, at 150 dpi?
God help me, I can't resist it...
We don't see them because the pixels are so small they're invisible????
--
Gord Deinstadt gdeinstadt@geovision.UUCP
Be gentle with me, I'm stupid.
gd@geovision.uucp (Gord Deinstadt) (11/18/90)
I wrote: >>>>> Why don't we see monitors that are 2048*1532, at 150 dpi? >God help me, I can't resist it... >We don't see them because the pixels are so small they're invisible???? and I'd better add :-) :-) lest anyone think I was serious... -- Gord Deinstadt gdeinstadt@geovision.UUCP
dmocsny@minerva.che.uc.edu (Daniel Mocsny) (11/18/90)
In article <11137@pt.cs.cmu.edu> lindsay@gandalf.cs.cmu.edu (Donald Lindsay) writes: >That doesn't mean that we're stuck, that means that it's time for >another change. One possibility is liquid crystal. Some researchers >are doing color via LC shutters over grayscale CRTs; this eliminates >the fabrication/resolution issues of a color CRT's color mask. Some >researchers use a small LC image, and use optics to make a large >projection of it. There are fun projects writing luminous dots with >lasers; the plasma panel people haven't given up; and in general >there are a zoo of experiments and hopes that could still make it. In >summary, I'm optimistic. All the labs know how valuable the right >patent could be: they are spending the money to search for it. I'd like to see a genetically-engineered animal with pixel-addressible skin pigment cells. We could start with an animal such as a squid or flatfish that already adjusts it pigmentation at will to camouflage itself. We could probably goose one of those babies up to run X-window. Then display size would be limited only by the size of the aquarium. It would be portable, provided you liked swimming. And best of all, it would be touch sensitive. I can see the project meeting already: "What do you mean, the code is late because a sperm whale ate your display?!?" -- Dan Mocsny Snail: Internet: dmocsny@minerva.che.uc.edu Dept. of Chemical Engng. M.L. 171 dmocsny@uceng.uc.edu University of Cincinnati 513/751-6824 (home) 513/556-2007 (lab) Cincinnati, Ohio 45221-0171
dmocsny@minerva.che.uc.edu (Daniel Mocsny) (11/18/90)
In article <1990Nov16.234227.3246@cs.cmu.edu> spot@WOOZLE.GRAPHICS.CS.CMU.EDU (Scott Draves) writes: >I think there is an additional, and more important, barrier to paper's >demise. copyright. If the ideal display were announced tomorrow, >paper would still exist because you can't get information from publishers >in digital/on-line form. This isn't being rectified as we speak, either. I imagine the appearance of reliable OCR will rectify it pretty fast. I do agree with your observation that paper is a necessary component in any commercial enterprise that relies on rationing information for its survival. Paper reduces the availability of information, making rationing practical. In very large quantities, information is cheaper on the computer. In small quantities it is cheaper on paper. On a plot of $/bit, the computer has a substantially high Y-intercept (because you have to buy at least one computer to read the first bit) and shallow slope. Paper has a Y-intercept of practically zero, but the slope is quite a bit higher, at least for text information. Photographics stomp the computer a little harder. >|> In audio technology, a $100 pair of headphones can >|> easily out-perform a $1000 pair of loudspeakers. > >only for the deaf. headphones don't reproduce low frequencies well. >Even good electrostatic ones. At the risk of ranging too far astray from the putative charter of comp.arch, I will note that the absence of low-frequencies is largely the result of the trend toward light, open air phones. The old Koss Pro-AA's had better bass than anything short of an 18" Cerwin-Vega... But they were uncomfortable, and the sealed cups (necessary for bass) made your ears sweat while you were riding your bike on rollers. -- Dan Mocsny Snail: Internet: dmocsny@minerva.che.uc.edu Dept. of Chemical Engng. M.L. 171 dmocsny@uceng.uc.edu University of Cincinnati 513/751-6824 (home) 513/556-2007 (lab) Cincinnati, Ohio 45221-0171
peter@ficc.ferranti.com (Peter da Silva) (11/19/90)
In article <6738@uceng.UC.EDU> dmocsny@minerva.che.uc.edu (Daniel Mocsny) writes: > In very large quantities, information is cheaper on the computer. In > small quantities it is cheaper on paper. On a plot of $/bit, the > computer has a substantially high Y-intercept (because you have to > buy at least one computer to read the first bit) and shallow slope. > Paper has a Y-intercept of practically zero, but the slope is > quite a bit higher, at least for text information. Photographics > stomp the computer a little harder. That's true for the first copy. In larger print runs paper gets the advantage again because of the large economy of scale of the printing press. For floppy disks, at least, the manufacturing costs of each floppy and the relatively slow reproduction rates shift the balance towards paper again. Of course CD-ROM has the potential of changing things again, but then the economy of scale is hurt because of the lower demand... -- Peter da Silva. `-_-' +1 713 274 5180. 'U` peter@ferranti.com
davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) (11/19/90)
I would expect that the rising cost of resolution for CRTs, and the dropping cost of flat panel display will result in a switch to high res analog solid state displays. Now would anyone like to guess when this will happen? -- bill davidsen (davidsen@crdos1.crd.GE.COM -or- uunet!crdgw1!crdos1!davidsen) VMS is a text-only adventure game. If you win you can use unix.