garnett@cs.utexas.edu (John William Garnett) (09/21/90)
I've heard that the color system for the color NeXTstation is
16-bit color. Would anyone care to expound upon this? Is the
16-bit color divided into 4 Red bits, 4 Green bits, and 4 Blue
bits (and 4 bits for ?) or does the 16-bits refer to the size of
the lookup table (i.e. 2^16 colors selectable from a palette of ?)?
Thanks.
--
John Garnett
University of Texas at Austin
garnett@cs.utexas.edu Department of Computer Science
Austin, Texasasd@mace.cc.purdue.edu (Kareth) (09/22/90)
In <34@brazos.cs.utexas.edu> garnett@cs.utexas.edu (John William Garnett) writes: >I've heard that the color system for the color NeXTstation is >16-bit color. Would anyone care to expound upon this? Is the >16-bit color divided into 4 Red bits, 4 Green bits, and 4 Blue >bits (and 4 bits for ?) or does the 16-bits refer to the size of >the lookup table (i.e. 2^16 colors selectable from a palette of ?)? Well, let's pull out the ole NeXT brochure, and see what it says! "16-bits-per-pixel color: The NeXTstation Color computer can display 4096 colors simultaneously, letting you create images of near- photographic quality." Now let's check out the exact specs on it: Display Memory: o 16 bits/pixel color o Includes 4 bits/pixel alpha channel o 4096 simultaneously displayable colors There is nothing in the brochure that mentions if it's from a lookup table or not. As I understand it (forgive me for being stupid), there are 3 4-bit RGB fields, and one 4-bit for opacity (see-througable-ness), Which makes 2^12 colors or 4096. Likewise the NeXTDimesion uses 3 8-bit fields, for 24 bit color (16.7 mil), and 8 bits for opacity. The 'alpha' channel is the opacity factor I gather. BTW, has anybody noticed the weight for the color monitors (from my spec sheet) weighs in at a hefty *** 70 LBS! *** Are color monitors normally that heavy? That's twice the weight of the monochrome one! Heck, the Cube only weighs in at max of 37lbs (from spec sheet). -k
cyliao@hardy.u.washington.edu (Chun-Yao Liao) (09/22/90)
In article <34@brazos.cs.utexas.edu> garnett@cs.utexas.edu (John William Garnett) writes: >I've heard that the color system for the color NeXTstation is >16-bit color. Would anyone care to expound upon this? Is the >16-bit color divided into 4 Red bits, 4 Green bits, and 4 Blue >bits (and 4 bits for ?) or does the 16-bits refer to the size of >the lookup table (i.e. 2^16 colors selectable from a palette of ?)? > >Thanks. well, the definition of 16-bit color here on NeXTstation means 12 bits of colors and 4 bits of transparency. So you actually get only 4096 different colors, but with 4 bits transparency. cyliao@wam.umd.edu o NeXT : I put main frame power on two chips. @epsl.umd.edu o people: We put main flame power on two guys. @bagend.eng.umd.edu o :::::::::::::::::::::::::::::::::::::::::::: xxxxx@xxxxx.xxx.xxx (reserved) o RC + Apple // + Classic Music + NeXT = cyliao
kent@decwrl.dec.com (Christopher A. Kent) (09/24/90)
With Display PostSCript, you don't deal with lookup tables any more. You pick from a fixed pallete -- the RGB (or CMYK) color space, and the interpreter renders your chosen color as closely as the hardware can do so. On a 16-bit NeXT, you get 4096 distinct colors in the destination color space. On a 32-bit, you get 2^24. Chris Kent Western Software Laboratory Digital Equipment Corporation kent@decwrl.dec.com decwrl!kent (415) 853-6639
gessel@cs.swarthmore.edu (Daniel Mark Gessel) (09/24/90)
In <1990Sep24.053046.5143@wrl.dec.com> kent@decwrl.dec.com (Christopher A. Kent) writes: >With Display PostSCript, you don't deal with lookup tables any more. >You pick from a fixed pallete -- the RGB (or CMYK) color space, and >the interpreter renders your chosen color as closely as the hardware >can do so. On a 16-bit NeXT, you get 4096 distinct colors in the >destination color space. On a 32-bit, you get 2^24. >Chris Kent Western Software Laboratory Digital Equipment Corporation >kent@decwrl.dec.com decwrl!kent (415) 853-6639 And remember, postscript does some pretty nice dithering. Whenever you ask for a color on the 16 bit NeXT that isn't exactly an available color, it should dither for you, like on the ancient two bit cubes (although I haven't seen this, it'd be a shame if it wasn't true.) I would suspect that the 24 bit's o'color (+ 8 alpha) there's no dithering, since I think it's a rare few who can tell the difference between rgb = 100,20,150 and rgb = 101,20,150. If all this is true, if you squint, have bad eyesight, or stand 10-15 feet from the monitor, you probably can't tell the difference between the two, since the pixels are so nice and small. Somebody who's seen all this care to comment? BTW, anybody know about implementing a z-buffer on the NeXTdimension, since the i860 can do this at a blinding speed? It'd make the color cube a competitor with loaded IBM R6000, or personal Iris, (although I haven't heard about them for a while), both of which are, I think, significantly more expensive. Dan -- Internet: gessel@cs.swarthmore.edu UUCP: {bpa,cbmvax}!swatsun!gessel
reid@cpswh.uucp (Dr Richard J. Reid) (09/24/90)
In article <7949@milton.u.washington.edu> cyliao@hardy.acs.washington.edu (Chun-Yao Liao) writes: >In article <34@brazos.cs.utexas.edu> garnett@cs.utexas.edu (John William Garnett) writes: >>I've heard that the color system for the color NeXTstation is >>16-bit color. Would anyone care to expound upon this? Is the >>16-bit color divided into 4 Red bits, 4 Green bits, and 4 Blue >>bits (and 4 bits for ?) or does the 16-bits refer to the size of >>the lookup table (i.e. 2^16 colors selectable from a palette of ?)? >> >>Thanks. > > well, the definition of 16-bit color here on NeXTstation means > 12 bits of colors and 4 bits of transparency. So you actually > get only 4096 different colors, but with 4 bits transparency. > > How does this transparency work? If I set a pixel to fully transparent, what do I see through to? Isn't this 16 bits all there can be for this position, with nothing behind it? Thanks, Dick
flynn@cse.nd.edu (Patrick J. Flynn) (09/24/90)
In article <1990Sep24.053046.5143@wrl.dec.com> kent@decwrl.dec.com (Christopher A. Kent) writes: >With Display PostSCript, you don't deal with lookup tables any more. >You pick from a fixed pallete -- the RGB (or CMYK) color space, and >the interpreter renders your chosen color as closely as the hardware >can do so. On a 16-bit NeXT, you get 4096 distinct colors in the >destination color space. On a 32-bit, you get 2^24. I attended *two* presentations before roll-out and was UNABLE to get a straight answer to the question: `Does the 16-bit color system have a 12-bit, WRITABLE color map, or are there only 4 bits of intensity on each gun?' In X-windows talk, is the color model `12-bit pseudo-color', or `12-bit true-color'? Either model gives you 4096 simultaneous colors on-screen, but the true-color model (for example) only allows 16 gray values. As an image-processing weenie, I would much rather have a 12-bit writable color map than a 12-bit static color palette. Hell, I'd rather have an *8*-bit color map than a 12-bit static color palette. Obligatory Rumor for Visualization people: I hear that OSC is porting apE to NextStep. Oughta look pretty sharp on the NextDim cube. TTFN-Pat -- Patrick J. Flynn, Computer Science & Engineering, University of Notre Dame flynn@cse.nd.edu <--- best flynn@cps.msu.edu <--- still works
rca@cs.brown.edu (Ronald C.F. Antony) (09/25/90)
In article <34@brazos.cs.utexas.edu> garnett@cs.utexas.edu (John William Garnett) writes: >I've heard that the color system for the color NeXTstation is >16-bit color. Would anyone care to expound upon this? Is the >16-bit color divided into 4 Red bits, 4 Green bits, and 4 Blue >bits (and 4 bits for ?) and 4 bits for alpha (transparency) This is 2^12 colors, i.e. 4096. Ronald ------------------------------------------------------------------------------ "The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man." Bernhard Shaw | rca@cs.brown.edu or antony@browncog.bitnet
wiml@milton.u.washington.edu (William Lewis) (09/25/90)
In article <444@news.nd.edu> flynn@cse.nd.edu (Patrick J. Flynn) writes: >I attended *two* presentations before roll-out and was UNABLE to get a >straight answer to the question: `Does the 16-bit color system have a >12-bit, WRITABLE color map, or are there only 4 bits of intensity on >each gun?' ... >I'd rather have an *8*-bit color map than a 12-bit static color >palette. Well, my understanding is that there are only 4 bits per gun. Dithered, no doubt. This seems sort of restrictive until you think about the trouble you'd have coordinating color palettes across umpteen different color windows, each with possibly completely non-overlapping palettes, with new drawing going on all the time and realtime video thrown in to boot. Look up palettes are a nice optimization but only if you know ahead of time what's going to be on the screen -- which, in a multi-window multi-process environment like a NeXT, you don't. -- wiml@milton.acs.washington.edu Seattle, Washington (William Lewis) | 47 41' 15" N 122 42' 58" W "These 2 cents will cost the net thousands upon thousands of dollars to send everywhere. Are you sure you want to do this?"
whelan@donald.wslab.Hawaii.Edu (Jerry Whelan) (09/25/90)
In article <8072@milton.u.washington.edu> wiml@milton.u.washington.edu (William Lewis) writes:
=> Well, my understanding is that there are only 4 bits per gun. Dithered,
=> no doubt. This seems sort of restrictive until you think about the
=> trouble you'd have coordinating color palettes across umpteen different
=> color windows, each with possibly completely non-overlapping palettes,
=> with new drawing going on all the time and realtime video thrown in to
=> boot. Look up palettes are a nice optimization but only if you know
=> ahead of time what's going to be on the screen -- which, in a
=> multi-window multi-process environment like a NeXT, you don't.
As I understand it, the color slab has one of those new fangled JPEG
DACs which is smart enough to give each window its own 4096 colors,
regardless of what the other windows are doing. I believe the JPEG chip
supports up to 256 windows, I don't know how many the NeXT implementation
supports, but 20 or so ought to be enough.
--
-------------------------------------------------------------------------------
whelan@ (uhunix.uhcc.hawaii.edu || uhccux.BITNET || nextsrv.wslab.hawaii.edu)madler@piglet.caltech.edu (Mark Adler) (09/25/90)
wiml@milton.u.washington.edu (William Lewis) conjectures: >> Well, my understanding is that there are only 4 bits per gun. Dithered, >> no doubt. This seems sort of restrictive until you think about the >> trouble you'd have coordinating color palettes across umpteen different >> color windows, each with possibly completely non-overlapping palettes, (etc.) I'd lay fairly good odds you are correct, and that a color map would have been too much the headache in NeXTStep, but I would point out for general edification that a color palette chip was announced about a month ago that had several palettes for separate windows and switched among them on the fly as the windows were scanned. Of course, the next question is "How many windows?", and as I recall, not terribly many. Still the possibility looms in future technology. Mark Adler madler@piglet.caltech.edu
clp@wjh12.harvard.edu (Charles L. Perkins) (09/26/90)
apE is already ported to NeXT and NeXTdimensio. They were at the intro and are selling apE on the NeXT for $75 including sources!! Charles
capslock@wet.UUCP (Allen Crider) (09/26/90)
In article <444@news.nd.edu> flynn@cse.nd.edu (Patrick J. Flynn) writes: > >... As an image-processing weenie, I would much rather have a >12-bit writable color map than a 12-bit static color palette. Hell, >I'd rather have an *8*-bit color map than a 12-bit static color >palette. > You should be able to use a settransfer proc in a window context to create a clut to do anything you want. You could make different settransfer procs for different windows in one program. Perhaps there is a way to also "settransfer" the DPS server globally. Maybe DPS looks for and executes a file called Sys/Start, like other PostScript interpreters with filesystems.
wisdom@pico.cs.umn.edu (Scott Wisdom) (09/27/90)
>Obligatory Rumor for Visualization people: I hear that OSC is porting >apE to NextStep. Oughta look pretty sharp on the NextDim cube. I was at the 68040 porting camp before the intro, and talked with the two students who were doing the port. When camp started, they did not have a running system. By the time intro rolled around, it seemed to be running very well. I was surprised at the speed they were able to complete the port. The guy who was doing demos brought up some cool looking images, and even did some animations. I asked him about getting this, and he said that it would probably go for about $75.00... WOW! And that includes complete docs! As soon as it's "official", I'm going to send away for it. Quite a deal! >Patrick J. Flynn, Computer Science & Engineering, University of Notre Dame - Scott Wisdom wisdom@heckle.cs.umn.edu <----- oh, PLEASE send any e-mail here! CODA MUSIC SOFTWARE HACKER (Who's trying to get those damned tuplets working!) Campus Consultant , U of Minnesota
thamer@gecko.cis.ohio-state.edu (M T ) (11/07/90)
What is the story on the NeXTstation Color? I've heard it's scheduled to be released 1'st quarter '91, has 12MB RAM, 16 bit color, and 105MB HD. Are they simply adding color to the monochrome slab, or are there any additional performance enhancements (why different simms speeds) ? How much of a disadvantage is it to have 16 bit color vs. 32 bit? Does it have any NeXTbus slots or isn't expandable like the mono-slab? Perhaps the best way to get a serious color system is to buy the NeXTcube at $8000 and NeXTdimension 32 bit graphics accelerator board at $4000 and MegaPixel Color Monitor at $???? We're definitely over $12,000 at this point and we still only have 8MB RAM and 105MB HD with the cube. If REAL color is going to cost this much, I'll take the $3300 (edu-discounted) mono-slab now. Any thoughts, thanks, -Mustafa -- "Two days ago I saw a vehicle that'd haul that tanker. You wanna get out of here; you talk to me." Max - The Road Warrior M. Thamer @cis.ohio-state.edu
declan@remus.rutgers.edu (Declan McCullagh/LZ) (11/08/90)
In article <85720@tut.cis.ohio-state.edu>, thamer@gecko.cis.ohio-state.edu (M T ) writes: > > What is the story on the NeXTstation Color? I've heard it's > scheduled to be released 1'st quarter '91, has 12MB RAM, 16 bit > color, and 105MB HD. Yes. Of course, you can add more RAM or order a larger HD. > How much of a disadvantage is it to have 16 bit color vs. > 32 bit? Not much. Believe it or not, PostScript drawing/moving windows, etc. is quite fast on the NeXTstation Color. PostScript also wins here, too: dithering a 24-bit image looks almost as good as it does on the NeXTdimension - it's hard to tell them apart unless you know what to look for. Even resizing a color image is quite fast. > Does it have any NeXTbus slots or isn't expandable like the > mono-slab? No slots here. > Perhaps the best way to get a serious color system is to > buy the NeXTcube at $8000 and NeXTdimension 32 bit graphics accelerator > board at $4000 and MegaPixel Color Monitor at $???? We're definitely > over $12,000 at this point and we still only have 8MB RAM and 105MB HD > with the cube. If REAL color is going to cost this much, I'll take > the $3300 (edu-discounted) mono-slab now. The list price of the MegaPixel Color Monitor is $3000. Educational discounts apply to that and the NeXTdimension board, naturally. "Serious color system?" Depends on how you define serious. "REAL color?" I have no idea what you mean... The NeXTstation Color is excellent for what it does, but the NeXTdimension is really in a different class. -------------------------------------------------------------------- Declan McCullagh / NeXT Campus Consultant \ declan@remus.rutgers.edu --------------------------------------------------------------------
portnoy@athena.mit.edu (Stephen L. Peters) (11/08/90)
In article <Nov.7.12.21.55.1990.11656@remus.rutgers.edu> declan@remus.rutgers.edu (Declan McCullagh/LZ) writes: >In article <85720@tut.cis.ohio-state.edu>, thamer@gecko.cis.ohio-state.edu (M T ) writes: >> How much of a disadvantage is it to have 16 bit color vs. >> 32 bit? > >Not much. Believe it or not, PostScript drawing/moving windows, etc. >is quite fast on the NeXTstation Color. PostScript also wins here, >too: dithering a 24-bit image looks almost as good as it does on the >NeXTdimension - it's hard to tell them apart unless you know what to >look for. Even resizing a color image is quite fast. I have another question about the color computer. I've heard that the 16-bit color is 12 bits of color, and four bits of alpha channel. I assume that this means 3 bits each or red, green, and blue, but I'm not sure what the "alpha" is -- some type of transparency effect? Stephen Peters
declan@remus.rutgers.edu (Declan McCullagh/LZ) (11/08/90)
In article <1990Nov7.203557.10442@athena.mit.edu>, portnoy@athena.mit.edu (Stephen L. Peters) writes: > In article <Nov.7.12.21.55.1990.11656@remus.rutgers.edu> declan@remus.rutgers.edu (Declan McCullagh) writes: > >Not much. Believe it or not, PostScript drawing/moving windows, etc. > >is quite fast on the NeXTstation Color. PostScript also wins here, > >too: dithering a 24-bit image looks almost as good as it does on the > >NeXTdimension - it's hard to tell them apart unless you know what to > >look for. Even resizing a color image is quite fast. > > I have another question about the color computer. I've heard that the > 16-bit color is 12 bits of color, and four bits of alpha channel. I > assume that this means 3 bits each or red, green, and blue, but I'm > not sure what the "alpha" is -- some type of transparency effect? Actually, there are four bits per RGB color, yielding 12 total. That gives you a fixed palette of 4,096 colors. The missing four bits are used for an alpha, or transparency channel ("monochrome" NeXT products also have a 2 bit/pixel alpha channel). While not incredibly exciting on a monochrome system, it's quite remarkable on a color system... -------------------------------------------------------------------- Declan McCullagh / NeXT Campus Consultant \ declan@remus.rutgers.edu --------------------------------------------------------------------