faustus@ucbcad.berkeley.edu (Wayne A. Christopher) (01/13/87)
Maybe somebody who knows a lot about CRT technology can answer a few questions I've been wondering about... How close are the color CRT's used for suns, uVaxes, etc to the current state of the art? (I'd estimate they're about 80 dots per inch.) How long will it be before we start seeing 300 dpi color CRT's available? Or are they already available for very expensive machines? How about LCD displays? Is there any theoretical reason for LCD's to be more or less suitable than CRT's for such high-resolution displays? (Is it even possible to make multiple-color LCD's?) Thanks in advance for the enlightenment, Wayne
hutch@sdcsvax.UCSD.EDU (Jim Hutchison) (01/14/87)
In <1219@ucbcad.berkeley.edu> faustus@ucbcad.berkeley.edu writes: >[...] How long will it be before >we start seeing 300 dpi color CRT's available? Correct me if I am wrong, but is there any need to go past 100 dots / inch? Last time I browsed the literature, the claim was 24bits of color and 100 dots / inch. (sorry no bibliography). >[...] How about LCD displays? Is >there any theoretical reason for LCD's to be more or less suitable than >CRT's for such high-resolution displays? (Is it even possible to make >multiple-color LCD's?) [...] I have seen multi color flat screens, it could have been a plasma instead of an LCD. You can make color LCD. LCD technology can not currently be designed with low persistance comparable to most color monitors. If you are shooting still frames, I guess this would not be so bad. You would definitely get the blurs on real time animation. - -- Jim Hutchison UUCP: {dcdwest,ucbvax}!sdcsvax!hutch ARPA: Hutch@sdcsvax.ucsd.edu "We live in a global village. I am sys V rel 2. You are version 6."
bobr@zeus.UUCP (Robert Reed) (01/14/87)
In the new products section of the latest Computer Graphics and Applications is an announcement for a two plane monochrome system designed for publishing applications that claims something like 1600x1280 on a 12 inch monitor. It also suggests that the system looks as good as output from a 300 DPI laser printer. I don't have the article with me, so I can't recite company name, etc., but you should be able to find a copy easily. -- Robert Reed, Tektronix CAE Systems Division, bobr@zeus.TEK
cmcmanis@sun.uucp (Chuck McManis) (01/15/87)
The Sun 3/160C monitor (19") appears to be about 80 dpi. The 3/260 has a higher resolution monochrome monitor available. In CRT's the state of the art is probably about 100 - 110 DPI on color monitors and slightly more for monochrome. A really interesting developement has come out of a start up in Milpitas CA. called Greyhawk Systems. It uses a 6" square LCD and a semiconductor laser to make a projection onto a built-in screeen. The screen is "D" sized (22" X 34") and displays vectors at a resolution of 400 dpi (yup, four hundred) Unfortunately it is a bit slow (Vector writes at 2000 inches/second) and being marketed as a paperless plotter. It does support 4096 colors though. More information can be undoubtedly gleaned from Greyhawk. Their number is (408) 945-1776. (I am in no way affilliated with Greyhawk, just saw their stuff and thought it was neat) -- --Chuck McManis uucp: {anywhere}!sun!cmcmanis BIX: cmcmanis ARPAnet: cmcmanis@sun.com These opinions are my own and no one elses, but you knew that didn't you.
jpm@calmasd.CALMA.UUCP (John McNally) (01/15/87)
In article <1219@ucbcad.berkeley.edu>, faustus@ucbcad.berkeley.edu (Wayne A. Christopher) writes: > Maybe somebody who knows a lot about CRT technology can answer a few > questions I've been wondering about... > How about LCD displays? Is > there any theoretical reason for LCD's to be more or less suitable than > CRT's for such high-resolution displays? (Is it even possible to make > multiple-color LCD's?) Thanks in advance for the enlightenment, Liquid Crystal Displays are just one of many forms of flat panel displays - others include plasma and electroluminesence. Each has advantages and disadvantages. For example, plasma screens are capable of high resolution and fast update speeds but require high voltage and are limited (currently) to an orange color. LCD displays have very low power requirements but (currently) have poor contrast, viewing angles, and update speeds. All flat panels obviously have the advantage of reduced space and simpler packaging than CRTs. However, there are even prototypes of "flat" CRTs (the electron gun is parallel to the tube and the deflection subsystem is radically different). The "winner" in this competition is yet to be determined... Currently LCDs appear to be the most promising flat panel technology and I expect that ultimately they will capture a major share of the display market. However, there are significant technical problems to be overcome (the Japanese are very active in this area). To me the most critical problem is the speed of update. Currently commercially available LCDs of any reasonable size do not have "pixel addressing" Writing to one pixel on a line or all pixels on a line requires the same amount of time - about 15 milliseconds. There are a lot of related problems in this area, but what it bolis down to is that LCDs are very slow in updating. Thin film transistor technology will solve this problem, but the technology has not yet reached the stage of mass manufacturing (several companies, mostly japanese are close). I refer you to an article that appeared in High Technology (May 1984, pp. 55 - 69), "LIQUID CRYSTALS BIG, BRIGHT, EVEN COLORFUL DISPLAYS". This is a good survey article on LCD technology. It discusses thin film transistor for active matrix addressing, the various methods used to make liquid crystals "turn on and off", and how color is introduced (using organic dyes). It also discussed futures and applications, and the inherent problems I have touched upon in much greater clarity. I think that it is premature to consider the use of LCD displays for CAD, especially areas requiring extremely good image quality (like solids modeling). We are still many years away from providing a cheap, reliable alternative to the 1024 x 1024 color CRT! (and the CRT manufacturers are not sitting still). -- John McNally GE/Calma 9805 Scranton Rd. San Diego CA 92121 ...{ucbvax | decvax}!sdcsvax!calmasd!jpm (619)-587-3211
garry@batcomputer.tn.cornell.edu (Garry Wiegand) (01/15/87)
In a recent article hutch@sdcsvax.UCSD.EDU (Jim Hutchison) wrote: >In <1219@ucbcad.berkeley.edu> faustus@ucbcad.berkeley.edu writes: >>[...] How long will it be before >>we start seeing 300 dpi color CRT's available? > >Correct me if I am wrong, but is there any need to go past 100 dots / inch? >... Sure - we'd always like the picture to look better! "Perfection" is reached only when adding more pixels makes no difference at all to my perceptions. As to where that point is - our 300 dpi laser printers produce lettering that looks ever-so-slightly lumpy to my eye. Commercial establishments which do serious work use 1000-dpi-or-better film recorders. I don't know whether they're perfectly happy with 1000 dpi, but that sounds like it might be getting warm. Only an order of magnitude away from where we are! But resolution isn't everything, I agree. As long as we're dreaming here, I'd also like to have monitors which: - don't jiggle and swim (GPX's make me seasick), - don't have intrinsically shiny faces, - are *flat*, - are able to reach full black (plasma screens are indeed nice if a bit snap-crackle-pop) and - can reach *all* of the hues my eyes can see. And no I haven't a clue how to do any of these. We build software. I note that what's already been done with frame buffers and CRTs is fairly amazing. - A 1024 by 1024 by 8 bit screen, noninterlaced, requires reading 1024*1024*8*60 = about 500 megabits of memory per second. And what it takes to push electron electron beams to that accuracy with just electric and magnetic fields I have no idea... are there any color hardware engineers in the crowd ? What are the real limits and state of the art? garry wiegand (garry%cadif-oak@cu-arpa.cs.cornell.edu)
wup@csv.RPI.EDU (Peter Y.F. Wu) (01/15/87)
IEEE Computer Graphics & Applications (CG&A) has a section on "New Products". The laest (Jan'87) issue has an announcement for 1664x2000 resolution monitor and adaptor boards for PC's. Perceived resolution close to 300 dpi available from laser printers... peter wu wup@csv.rpi.edu
philm@astroatc.UUCP (Phil Mason) (01/15/87)
In article <1105@zeus.UUCP> bobr@zeus.UUCP (Robert Reed) writes: >In the new products section of the latest Computer Graphics and Applications >is an announcement for a two plane monochrome system designed for publishing >applications that claims something like 1600x1280 on a 12 inch monitor. It >also suggests that the system looks as good as output from a 300 DPI laser >printer. I don't have the article with me, so I can't recite company name, >etc., but you should be able to find a copy easily. >-- I used to work for the parent company of a startup in Pittsburgh, PA. The startup is called MegaScan Technology. They have recently exhibited a 4K by 3K Black and White 19 Inch monitor with video memory and driver. Essentially the resolution is 300 DPI. The display is quite spectacular. They also have a bit slice Raster-Op Engine and a 68020 CPU board set that can manipulate bit maps. If you are interested, the company's address is : MegaScan Technology 4005 Vista Vue Drive Gibsonia, PA 15044 (412) 443-5820 I am no longer with any related company and I am supplying this information not as an advertisement, but to let people know what is available. -- Kirk : Bones ? | Phil Mason, Astronautics Technology Center Bones : He's dead Jim. | Madison, Wisconsin - "Eat Cheese or Die!" ...seismo-uwvax-astroatc!philm | I would really like to believe that my ...ihnp4-nicmad/ | employer shares all my opinions, but . . .
kurtk@tekgen.UUCP (Kurt Krueger) (01/16/87)
One point that hasn't been touched upon are some inherent limitations of color CRT's. Most of the current high resolution color CRT's are built in much the same manner as TV tubes, that is they use three electron guns and use a shadow mask to insure that the red gun (for instance) only lights up the red phosphor dots. The trick comes with that shadow mask. A 1024x1024 display requires a mask with 1024x1024 holes that are big enough to pass enough electrons to get a reasonably bright display and yet is strong enough that it doesn't buckle under stress (it gots hot where the beam strikes it). It also has to be precisely aligned (and stay that way) otherwise the red beam may light up a little green. This is a rather tall order, and it is why color displays are typically little better than 1024. The phosphor dots presents a lesser problem, but remember for each pixel you need three dots (one each red, blue, green). Black and white displays don't have the problems with a shadow mask. The technology is a bit more advanced with monochrome displays as higher resolution displays are available. This same technology can be applied to color except for the fact that no one (that I know of) has been able to build a shadow mask with the required resolution and size (it is possible to build a small shadow mask at some high dots/inch but getting much more than 1024 on a terminal size display is where the hitch is). It is possible to build color CRT's without shadow masks. The techno- logy has been around a long time (projection systems - three monochrome tubes projected onto one screen, penetrons - three layers of color phosphors excited by varying the beam energy, liquid crystal devices, + ???) but they have their own sets of problems. The fact that every currently marketed high resolution color CRT that I'm aware of uses a shadow mask must indicate something. Another issue that I will only touch upon is grey scale. A typical color TV only has around 320 'pixels' of horizontal resolution. However, a continuous tone scene can render better on the TV than on a high (1024) resolution terminal. Why? The TV has a lot of grey scale while the terminal is limited. The better displays with at least 8 bits of color map information can do a respectable job but they are still lacking. Note that 8 bits only allows 3,3,2 bits for each red, green, blue. You only can get 8 levels for two of the colors and 4 levels for one. 256 colors sounds good until you look at it this way. If you divided your color space in this manner, a blue portion of your image can only have four levels (and one of those is black!). Techniques such as dithering can make improvements but the basic display is still limited. There is no technological limits that restrict a display to 8 bits of color information (indeed, 24bit frame buffers are available, the Tektronix 412x maps 256 colors into a 24 bit palette) but cost considerations enter - a 1024x1024 display needs a mega-bit of memory for each bit of color information. I expect this problem to be solved before the resolution problem is solved.
sjrapaport@watcgl.UUCP (01/16/87)
In article <1219@ucbcad.berkeley.edu>, faustus@ucbcad.berkeley.edu (Wayne A. Christopher) writes: > How about LCD displays? Is > there any theoretical reason for LCD's to be more or less suitable than > [CRT's?] Just a personal opinion. Sorry to all the excellent CRT manufacturers out there, along with the electroluminescent and plasma guys, but I'm just tapping my fingers waiting until we get some decent fast LCD's. Glowing screens bug my eyes. Eyestrain is nothing compared to the horror of going to bed after a prolonged hacking session with visions of glowing dot-matrix characters dancing in your head.... -steve@watcomputer "If it glistens, gobble it!" -zippy the pinhead. "If it glows, smash it." -me
mmp@cuba.UUCP (01/17/87)
In article <1219@ucbcad.berkeley.edu>, faustus@ucbcad.berkeley.edu (Wayne A. Christopher) writes: > How close are the color CRT's > used for suns, uVaxes, etc to the current state of the art? (I'd > estimate they're about 80 dots per inch.) How long will it be before > we start seeing 300 dpi color CRT's available? Or are they already > available for very expensive machines? > You just asked the $65,000 question! And you are getting a LONG answer to it. A major difference between monochrome and color CRTs is that the latter has a mask interposed between the CRT's optics and its phosphor. CRT resolution is thus measured as a pitch, which is the center-to- center distance between the little holes in the mask. "Available" CRT pitch specs are: .31mm, .26mm, and .21mm. You may see instead: .30mm, .25mm, and .20mm. It all depends on where you measure it (the mask itself or through the glass in front of it). The .31mm CRT is the one most often found in Suns, uVaxen, etc 19-inch displays. The .26mm CRTs are becoming more popular, as their price has begun to drop now that they are in full production. The price of .31mm CRTs, however, is also dropping as that technology matures, and vendors are not likely to make the switch unless customers demand it (I'd imagine they rather enjoy the extra margins). The .21mm CRTs are now available in sample quantities and production quantities won't be available till later next year (remember, however, that this is vacuum tube technology and production quantities are not made available until there is a LOT of demand for it). I have seen one of these made by Hitachi and I am spoiled for life -- a .31mm and even a .26mm CRT will never look good again! This is definetely state of the art and it'll be a while yet. There's a technology/theory developed by Dr. Carlo Infante, now an independent consultant, that pretty much predicts that moving up in CRT pitch is more important to picture quality than increments in things like amplifier frequency, etc. His work, called Modulation Transfer Function Area (MTFA) actually quantizes the improvement to be expected and it's pretty close my own experience. So, if you believe that, you'd want the CRT (and monitor) vendors to get off their duff and bring us even lower pitch CRTs as soon as possible (which, by the way, also allows to get more picture on the screen!). The problem with fullfiling that wish, such as I have, is that even though all the CRT vendors are "working on it" (i.e. .20mm and .17mm pitch CRTs), it ain't easy to pull and still make them manufacturable. The main problem is easy to understand if you can visualize what happens to a sheet of metal the thickness of a 20lbs sheet of paper when you fill it with tiny holes .20mm or .17mm apart -- ok, now try to pick it up and roughhandle it into the back of a curved sheet of glass, etc. Now that you've managed to get in place, imagine what happens when you "heat" it with an electron beam, over and over (i.e. it tends to warp). SONY, for one, has tried avoiding those problems by using alternative technology to implement their masks. Conceptually, they start with columns of ultra thin "wires" stretched across a metal frame in a jail-bar arrangement. (This has the extra benefit that the raster beam is sampled/filtered in the horizontal orientation only, but not in the vertical orientation.) They then add two tension wires in the horizontal direction to keep the frame from collapsing (if you look closely at a SONY CRTs you will notice the tension wires -- they look like a "one pixel wide dark line", and they are particularly easy to spot if you look at a "flat" area, with a homogenous color throughout -- if you want to irk TV salemen and the like, just stick your face up to the CRT and ask them, "what the heck is wrong with this TEEVEE? it's got a couple of them little lines missing!" :-) However, when, it gets past the .20mm pitch, SONYs technology poops out, too. That sounded like the end of the line for CRT technology, until Zenith changed things (yes, the _American_ company, and, no, they didn't do this research in Japan -- this is up and up American know-how). Zenith calls its technology the Flat Mask CRT, which conceptually is rather straight-forward (once they tell you about it): take the paper-thin mask, fill it with holes, and then glue it to the back of a sheet of glass! and that's it :-) Of course, the "sheet of glass" has to be optically flat, and that is no mean trick. Once you develop the technology to manufacture those things, however, the sky is the limit (easy for me to say): puch as many holes as you want in as thin a sheet of metal as you want (the thinner the better), then glue it to the glass sheet and you can let the American Tourister gorilla put it into the tube (i did preface this with the word "conceptually", didn't I?) Not only can you display a much higher resolution pictures that way, but you can also get a much higher contrast picture. The latter is a well known problem with spherical (or cylindrical for SONY) displays. Standard CRTs also have the problem that they catch reflections from around the room, while Zenith's Flat Mask eliminates that problem, too (ok, IF you stare into the CRT with your face levelled with it, AND you have a strong light shining on your face, THEN it will reflect on the display -- so, don't do that, which sounds unnatural anyway :-) Oh, yes. dots per inch, hey? /* Visible Lines Per Frame */ Horizontal_Sync_Rate = 78KHz; Vertical_Sync_Rate = 60Hz; 78KHz/60Hz => 1248 visible lines/frame + 52 lines/frame eaten up by the vertical retrace monster; Note also that if you want higher vertical refresh rates (an often heard goal/yearning/fad these days), you have to up your Horizontal Sync Rate just to retain your vertical resolution; Horizontal_Sync_Rate = 138KHz Vertical_Sync_Rate = 66Hz; /* a la Sun's */ 138KHz/66Hz => 2048 visible lines/frame + 47 lines/frame eaten up by the vertical retrace monster; moving that raster back and forth 138,000 times per second ain't easy; also pixel on-off time goes from 5.6 nsec/pixel to 2.9 nsec/pixel and theoretically that requires 350MHz DACs and video amps! /* "Holes" Per Line */ Line_Width (Active area) = 34.9 cm (13.75"); if(Pitch == .17mm) 34.9K/.17 => 2055 holes/line if(Pitch == .21mm) 34.9K/.21 => 1664 holes/line if(Pitch == .26mm) 34.9K/.26 => 1343 holes/line if(Pitch == .31mm) 34.9K/.31 => 1127 holes/line This if, of course, only partly true. To fully exploit the higher pitch CRT (and really get all those dots on the screen), you have to have an appropriately sized frame buffer behind it, AND the right Digital to Analog Converters (about 200MHz DACs for 1664X1248, and at least 350Mhz for a 2KX2K display, though 500MHz DACs are preferred!), AND the appropriate video amplifiers, etc. In addition, pitch is only one of the parameters: beam spot size is another. If you have a very small spot size (i.e. less than twice the pitch), you end up with lots of Moire patterns (look it up); if you have a very fat spot (i.e. more than three times the pitch), you end up with very soft images. Thus, the small spot size works better when displaying solids (i.e. sharper edges), while the fatter spots are better for displaying wire frames (i.e. less Moires). It gets worse (i.e. over my head) from there (beam optics, for one, are way beyond me). > > Are there any theoretical reason for LCD's to be more or less suitable than > CRT's for such high-resolution displays? (Is it even possible to make > multiple-color LCD's?) Thanks in advance for the enlightenment, > I don't know much about LCD masks. The idea here is to bypass the thin-mask-handling problem by encapsulating the mask "inside" two sheets of glass. You could conceivably change the mask dynamically to do windowing :-) All I know is that every time I've mentioned "LCD masks?" to people who know about these things, they make a face. And Tektronix, the most vociferous proponent of that technology, has not delivered anything yet (that I know of). From my vantage point (i.e. ignorance of the gory details), I say that Zenith's technology, though analog, is going to work better _sooner_ than the LCD, "digital" approach. The big question is "when" is Zenith going to bring this technology to the 19-inch, professional market? All I've heard so far (up to three months ago) is that they have plans to make .31mm CRTs in 14-inch format for the consumer market, and .20mm CRTs for the professional markets, but only in 14-inch format, and not in the 19-inch format. My guess is that making optically flat glass in that size sheets is not quite a piece of cake/possible. Time will tell, but I can't wait. ____________________________________________________ * Matt Perez * sun!cuba!mmp (415) 691-7544 DISCLAIMER: beisbole has bean bery, bery guud too me
jon@eps2.UUCP (Jon Hue) (01/18/87)
In article <2029@batcomputer.tn.cornell.edu>, garry@batcomputer.tn.cornell.edu (Garry Wiegand) writes: > >>[...] How long will it be before > >>we start seeing 300 dpi color CRT's available? > > > >Correct me if I am wrong, but is there any need to go past 100 dots / inch? > > lettering that looks ever-so-slightly lumpy to my eye. Commercial > establishments which do serious work use 1000-dpi-or-better film recorders. > I don't know whether they're perfectly happy with 1000 dpi, but that > sounds like it might be getting warm. Only an order of magnitude away > from where we are! From what I've learned about the printing industry, I believe that continuous tone images are scanned at 12 dots/mm (300dpi) and line art (text) is scanned at 40 dots/mm (1000dpi). The input and output (laser) scanning are done at the same resolution. 300dpi pitch monitors might allow some type of soft proofing, but there are serious color problems that would need to be solved. As it is, no one using prepress equipment (Scitex, Hell, Crosfield) trusts the monitor, they always make a proof (Cromalin) and say things like "boost the cyan 5%". To give you some idea of what film is capable of (I use 35mm as an example, you can put 8 x 10 film backs on all these film recorders) I recall 35mm film falls apart somewhere between 2000 and 3000 pixels horizontally. As far as the high-end film recorders go (Celco, Dunn), they are something like 8K x 8K addressable, but with the pitch and beam and whatnot, there are something like around 5K distinct points. Dunn goes out of his way to explain this and be honest, and no one understands him ("But their brochure says 8K x 8K and his says 5K, so the other one must be better"). > magnetic fields I have no idea... are there any color hardware engineers > in the crowd ? What are the real limits and state of the art? I'm not one, but my office is right next to someone who is. I would say that with current off-the-shelf parts the state of the art frame buffer would be 1600 x 1280 x 24 bits. We figured that if you built a 1280 x 1024 x 24 bit frame buffer, you needed LUTs with 7ns access times. Fortunately, you can buy ECL static RAMs with 3ns access times. You can also get three TRW 8-bit 250MHz ECL video DACs on one chip. The machine I fool around with at work (a couple years old) has a video section capable of running at 100MHz, though there are only have 656 x 485 x 24 bits on the screen (NTSC). BTW, if you want to see a nice color display (not state of the art, but still nice), try to get a tour of a color separation house. The Scitex Response 350 is 1024 x 768 x 24 bits. It's pretty boring if the artist is touching up a scratch in a negative, but if they are doing some actual design, it is interesting to watch (though design is a bit expensive on these machines, at ~$700/hr). I saw something interesting in a magazine. It was a set of three boards that take up two slots in an AT (one is on standoffs). It has a 68020, a 68881, a custom VLSI graphics processor, a 640 x 485 x 24 bit frame buffer (NTSC), and a bunch of DRAM, probably 4MB. It runs some form of Sys V and has virtual memory. Anyone know anything about this board set? If it isn't too expensive, it sounds like a very nice low-priced platform. I imagine that we'll see some high-resolution (1600 x 1280, 1280 x 1024) full color (24 bits) systems for electronic prepress in the near future. "If we did it like everyone else, Jonathan Hue what would distinguish us from Via Visuals Inc. every other company in Silicon Valley?" sun!sunncal\ >!leadsv!eps2!jon "A profit?" amdcad!cae780/
bzs@bu-cs.BU.EDU (Barry Shein) (01/19/87)
A few years ago when we got our first 300 DPI printer I remember proudly mentioning this fact to a friend who is a graphic artist (who knows nothing about computers.) She looked at me puzzled for a moment and said "three HUNDRED dpi, no, you must mean three THOUSAND dpi..." Well, a dug my toe in the dirt and said oh shucks, showed her some output which she only sneered at mildly ("hmmm, well, some genius *has* tuned these fonts rather well, but it still looks like junk...") Really made my day. Anyhow, back to the point about CRTs matching print technology. If you deal with matching images on the screen and the paper you develop very little doubt that we are going to obscene contortions to try to resolve the differences, it's a huge waste of time and effort (and doesn't even work very well.) Two different font sets (I still haven't seen any edge description fonts I can actually use on the screen and paper in *my* software, it's still all bitmap image files) and other issues, ugh! I suspect the same is true for color devices and the advent of color xerography (which I assume is just around the corner from becoming popularized, it certainly exists) is just going to make this all worse. This is one problem where the software crew has by and large hit a fundamental brick wall and, I suspect, will only go away when the appropriate hardware becomes available. The current situation is ridiculous, try and write a WYSIWYG editor with embedded graphics and integration into any common print engine and you'll see immediately how absurd the situation is. -Barry Shein, Boston University
news@cit-vax.UUCP (01/19/87)
Organization : California Institute of Technology Keywords: From: jon@oddhack.Caltech.Edu (Jon Leech) Path: oddhack!jon In article <54@eps2.UUCP> jon@eps2.UUCP (Jon Hue) writes: > >I'm not one, but my office is right next to someone who is. I would say >that with current off-the-shelf parts the state of the art frame buffer >would be 1600 x 1280 x 24 bits. We figured that if you built a 1280 x >1024 x 24 bit frame buffer, you needed LUTs with 7ns access times. >... >I imagine that we'll see some high-resolution (1600 x 1280, 1280 x 1024) >full color (24 bits) systems for electronic prepress in the near future. We just got some HP 98720/21 graphics boxes, which are 1280 h x 1024 v x 24 bits. I don't know how expensive they are (it's part of an HP grant program), but they are available now. One of the amusing features is the ability to switch between two sets of color maps every 133 ms. -- Jon Leech (jon@csvax.caltech.edu || ...seismo!cit-vax!jon) Caltech Computer Science Graphics Group __@/
rotheroe@convexs.UUCP (01/19/87)
A company called Sigma Designs, and located somewhere in CA has recently begun shipments of their latest graphics board, aimed at desktop publishing. The resolution is two planes of 1664x1280 (I think), monochrome. This gives four levels of gray, and an effective resolution of 3328x2560 - sounds just about right for a laser printer image. The card(s?) are supposed to plug into anything from a PC on up. The price: $2300 with 19" monitor, $1800 with 15" and $1200 without a monitor (I'm recalling prices from memory). No, I don't work for them - the only connection I have is that I've bought stuff from them in the past, and may buy some new toys from them in the next few months. Dave Rotheroe {allegra, ihnp4, uiucdcs, ctvax}!convex!rotheroe CONVEX Computer Corporation Richardson, TX "Good afternoon, gentlemen. I am a HAL 9000 computer. I became operational at the Hal plant in Urbana, Illinois, on the twelfth of January, 1992." 2001 & 2010 (book only for 2010)
thomas%spline.uucp@utah-gr.UUCP (Spencer W. Thomas) (01/21/87)
In article <54@eps2.UUCP> jon@eps2.UUCP (Jon Hue) writes: >I would say >that with current off-the-shelf parts the state of the art frame buffer >would be 1600 x 1280 x 24 bits. We have (and have had for a couple of years) a 24 bit frame buffer that has a resolution of 1536x1152. So it is definitely possible! =Spencer ({ihnp4,decvax}!utah-cs!thomas, thomas@utah-cs.ARPA)
markp@valid.UUCP (Mark P.) (01/22/87)
> > We figured that if you built a 1280 x > >1024 x 24 bit frame buffer, you needed LUTs with 7ns access times. > >... > >I imagine that we'll see some high-resolution (1600 x 1280, 1280 x 1024) > >full color (24 bits) systems for electronic prepress in the near future. Actually, it is now completely cookbook to build a 1280x1024 non-interlaced display, assuming that you use 64kx4 video RAMs and a really nifty RAMDAC part from Brooktree (soon to be 2nd-sourced by Fairchild), called the Bt458. This combination effectively allows you to almost completely eliminate the ECL portion of your design (except for a clock driver for the Bt458), even at video rates of 125MHz. For those not familiar with video RAM technology, imagine a conventional 64kx4 DRAM, organized as 256 rows of 1024 bits each. There is a special cycle, called a load or transfer, which transfers an ENTIRE ROW of 1024 bits into a 25MHz shift register (assuming a 120ns access time part). This shift register then outputs 4 bits at a time, for an aggregate single-chip bandwidth of 100Mbps! However, to implement a 1280x1024, 125MHz display, you need to use 5 chips per plane (at any given time) and some unorthodox addressing techniques, which are left as an exercise for the reader. The Bt458 then receives up to 25 bits at a time, and multiplexes them at the higher bit rate (either 4:1 or 5:1), pipelines them, passes through three 256x24 lookup tables with overlay, three video DACs, and out pops RGB video! The internal RAMs operate at an effective cycle time of 8ns, and are, believe it or not, implemented entirely in CMOS. For even higher resolutions, video RAMs may be interleaved (which isn't very hard at 25MHz) and Brooktree has independent RAMs and video DACs which operate at 200MHz, for an effective cycle time of 5ns (>1536x1280). Except for the nasty problems of PC routing in the output stage, such circuits can easily be designed by a garage-based logic hacker. The possibilities generated by a cookbook approach for very high resolution displays are truly frightening, but unfortunately monitor costs will probably continue to keep the cost of graphics in the nether regions expensive. Further details avaiable from your local Brooktree, Mitsubishi, TI, AMD, NEC, Fujitsu [etc. ad nauseum] distributors. Mark Papamarcos Valid Logic Systems hplabs!{ridge,pesnta}!valid!markp P.S. I have no financial interest in Brooktree, or in any of the various semiconductor vendors which produce video RAMs, but I do personally think that they are neat parts.
henry@utzoo.UUCP (Henry Spencer) (01/31/87)
> ...Anyhow, back to the point about CRTs matching print technology. If you > deal with matching images on the screen and the paper you develop very > little doubt that we are going to obscene contortions to try to > resolve the differences, it's a huge waste of time and effort (and > doesn't even work very well.) Two different font sets... Um, it's worse than that, Barry. Don't forget that screens and printers have different kinds of nonlinearities in their presentation of the image. Given that getting good-looking text requires different font sets depending on whether your printer uses a write-white or write-black process, what makes you think there is any possibility of getting fonts that will look the same on the screen and on the page? I don't see it happening soon. -- Legalize Henry Spencer @ U of Toronto Zoology freedom! {allegra,ihnp4,decvax,pyramid}!utzoo!henry
rw@beatnix.UUCP (02/12/87)
In article <54@eps2.UUCP> jon@eps2.UUCP (Jon Hue) writes: >To give you some idea of what film is capable of (I use 35mm as an example, >you can put 8 x 10 film backs on all these film recorders) >I recall 35mm film falls apart somewhere between 2000 and 3000 pixels >horizontally. As far as the high-end film recorders go (Celco, Dunn), they The standard number quoted is 18M pixels for a still frame of Kodacolor 100. Kodachrome 25 would be even higher. 35mm movie film has about half the image area of 35mm still; 70mm movie film has over twice the image area of 35mm still. The data rate of a 70mm movie is thus approx: 35M * 3 (bytes/pixel) * 24 (frames/second) or 2.52 GBytes/sec. ShowScan, which runs 70mm film at 60fps has a data rate of ~6.3 GB/sec. How about GaAs LUT RAMs? Russell Williams ..!{sun|styx}!elxsi!rw