Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Technology

Cleartype In Depth 142

spectecjr writes "Looks like Microsoft have at last released some detailed information on their ClearType technology. It involves a whole load of Fourier Analysis to come up with the optimal distribution of color energy and to reduce color fringing. You can read the paper and more indepth info in PDF and gZipped PostScript, as well as the paper submitted to the IEEE Signal Processing Letters journal PDF. Samples of ClearType vs. standard anti-aliasing are up online too."
This discussion has been archived. No new comments can be posted.

Cleartype In Depth

Comments Filter:
  • ...unless you have an LCD (which is not an iMac).

    From the site: "In order to view a sample of the optimal ClearType filters, you must view this page on an LCD with RGB stripes. You will not be able to see the full ClearType effects on a CRT. Some LCDs (such as in the iMac), use BGR stripes, which will not work with these samples."
    --
    Compaq dropping MAILWorks?
  • by brennanw ( 5761 ) on Tuesday June 13, 2000 @08:11AM (#1005028) Homepage Journal
    because I don't see any Internet Explorer integration ANYWHERE. :)
  • I copied the sample from the website and did a zoom in on it. Cool! You can see all the fill-in pixels
  • I copied the sample from the website and did a zoom in on it. Cool! You can see all the fill-in pixels.
  • until someone demonizes the technology because it was created by Microsoft? The under is 1 minute.
  • ...what, there seems to be something actually new in this implementation?

    Slashdot groupthink had convinced me that MS never creates anything original... hmm... must think... *smoke* brian overheating...

    *kaboom*
  • That's why I used the !@#$ smiley
  • They had this in the Apple 2, so they could squish ultra-small letters for 80 and 120 column terminals.

  • ...when viewed using a CRT (against their advice of using LCD for display), the samples appear as follows:

    a) cleartype samples appear a bit sharper/darker than "traditional" anti-alias techniques, when viewed from monitor-eyeball distance. Closer inspection revealed some aliasing effects on the cleartype sample.
    b) cleartype looks just a hair better at small font sizes.
    c) the difference between cleartype and traditional anti-aliasing, on a CRT rather than an LCD, is pretty insignificant.

    If I can get either of my laptops semi-reliable for more than a few minutes, I'll post some info on the difference. So far, my impression is that this is redundant technology that would mostly benefit Microsoft rather than the licensors of other anti-aliasing technology - perhaps the impetus for the creation of cleartype in the first place.

  • Another Microsoft Innovation. The Apple II did the same thing. Read all about it.

    http://grc.com/cleartype.htm [grc.com]

  • Don't mind, troll alerts are blossoming everywhere. Look at most of TRoLL's production [slashdot.org].

  • by RPoet ( 20693 )
    These are crazy days, when microsoft provide downloadables in GNU Zip! :)
    --
  • While I heartily applaud the steps being taken to make screen more legible, I question the truth of the Microsoft claim to have invented this technology. Steve Gibson of Gibson Research Corporation [grc.com], has a whole page dedicated to this very issue, located here [grc.com]. He makes some very straighforward points, including reminding us that Apple had something which worked the same way around 15 years ago.

    And of course this does absolutely nothing for screens other than LCD..
  • For some more information on the "jaggies [netmeg.net]" (the stuff that makes things look bad, that this technologies aimes at improving in fonts) in OpenGL and OS/2 check out:

    [edm2.com]
    http://www.edm2.com/0603/opengl.html
  • Since when does the iMac have an LCD screen?
  • Very interesting
    Yes, your still have to have an LCD to see that it works

    checkout:
    http://grc.com/ctwho.htm [grc.com]
  • Do they ahve some prototype we don't?
    Or do they perhaps mean the iBook?

    Certainly I have found that different Apple LCD screens have different pixel orders
  • Damn useful for all those black-and-white e-books...
  • The clear-type sample [microsoft.com] on that there page looks real good when ya' look at it compared to that teeny-weeny lil' web font they'all use.

    Now mebbe if Micr'Soft replaced all 'content' on tha' web with -1 font-sizes -ClearType'd look even better by comparison.

    -
  • That the anti-aliasing technique shown is 61-tap... fairly CPU intensive.

    If ClearType is computationally equivalent to an (example) 8-tap anti-aliasing method than it is a major breakthrough. (can't believe I'm saying something positive about M$...eww...)

  • All I have to do is iterpret symbols on the screen as letters, the more you dull them out the less information can be packed into the screen. (Of course that would be one wacky font.)

    Besides, this ClearType also requires a color display adding to the battery draw...

    --
    Eric is chisled like a Greek Godess

  • by Juggle ( 9908 ) on Tuesday June 13, 2000 @08:27AM (#1005048) Homepage
    All arguments over wether this is actually different than what Apple did way back with the II or not aside I'm not as impressed as I had expected to be.

    Back when this was first announced I thought it made sense and would work great. I even read a few pages which showed how Apple did the same thing and loaded the samples on there up on a few screens to see how they looked. Yeah there was a difference but nothing major.

    Now this comes across with what sounds like samples made using the exact algos that MS is touting as their great new innovation. So I pull out the old palmtop and laptop and check the page out.

    But this looks like hell to me! I can see major color fringing and in those text waterfall examples I see rainbows in the CT examples that are so prominent they make it harder for me to read than the AA example! Yuck!

    Does anyone else notice it as much as I do? Or are my eyes just abnormally color sensitive?
  • by tjwhaynes ( 114792 ) on Tuesday June 13, 2000 @08:31AM (#1005049)

    Cleartype is basically antialiasing which takes note of the way in which the LCD screens make up their pixels. From the comparative samples, it's difficult to compare the two samples given on a normal (i.e. CRT) monitor. I strongly suspect that this is not going to make such an impact on CRT-based techniques for several reasons. The conventional monitor scans a modulated electron beam across a mask before hitting the coloured phosphors which make up the display. Despite the regular display patterns of the mask, a pixel on a conventional CRT could line up with any combination of the coloured phosphors, as the start of each pixel could be on any of red/green or blue phosphors. The Cleartype technology relies on being able to make use of the arrangement of single-coloured pixels to enhance the imagery (and yes, I do believe it can make a difference) and must therefore be aware of the mapping from the resolution of the image on screen to the resolution of the actual screen matrix itself.

    Something does strike me as odd in the samples though. One of the things that ideal anti-aliasing should do is give a completely even weight to every letter/symbol in a font (assuming that the base font definition is designed with this in mind). However, scrutiny of the samples seems to suggest that either the font they are using is subtly broken in this respect, or that the anti-aliasing and Cleartype render used here is not ideal - take a look at the weighting of the 'x' character, and slightly less of a problem is the 'k'. This could be hinting gone wrong, or a bad font definition - I'd be interested to know whether it looks better on an LCD screen.

    Cheers,

    Toby Haynes

  • by drivers ( 45076 ) on Tuesday June 13, 2000 @08:32AM (#1005050)
    This is an interesting side of Microsoft I hadn't realized existed: Microsoft Research. It looks like they are working on a lot of very interesting stuff. I was searching through their stuff I found that one of the things they were working with was IPv6 related networking, and you can download a web server called "Fnord!". I downloaded it, and checked out license.txt and it is the GPL! Apparently they used Fnord! written by someone else and used it as a basis for some research software, and you can download it, source and everything. Cool.
  • Untrue. My Sony Trinitron monitor has the color cells aligned much as they would be on an LCD. I did see an improvement in readability when looking at the right images (gamma = 1.8 on my display)
    --
    Ski-U-Mah!
    Stop the MPAA [opendvd.org]
  • Where the heck did all that karma come from? damn +1 bonus, I want to earn my place.

    --
    Eric is chisled like a Greek Godess

  • by konstant ( 63560 ) on Tuesday June 13, 2000 @08:34AM (#1005053)
    Here is the software of the guy who claims Apple figure all this out years ago. He has a demo written (idiosyncratically enough) in pure i386 assembly:

    Free and Clear [grc.com]

    However, his discussion doesn't seem nearly as complex as the one we have linked in this article. My feeling is that the idea of sub-pixel manipulation is one of those "floating revelations" that recur to more than one clever mind, but that MS Cleartype is the first practical application.

    Oh, and here's the cleartype site at MS. Be sure to view it with an LCD screen - otherwise you won't get the benefit due to the triangular distribution of color spots on a CRT:
    MS Reader [microsoft.com]

    -konstant
    Yes! We are all individuals! I'm not!
  • You'll notice it like crazy if you're not on a LCD monitor with RGB stripes. You'll also notice it in low-light conditions. Doesn't work for white text on black background either, fringes like crazy. CT isn't likely to replace AA, but MS does use it in their PocketPC book reader, which is exactly the sort of thing it's aimed at.
  • Anyone remember Steve Gibson's ClearType [grc.com] demo? Steve says that Apple invented this tech looong time ago (so long that patent already expired). I'm hoping that he will re-check facts and tech again and see if MS really did innovate something this time. I still think that this is old Apple's tech that Woz discovered some time ago. If anyone has a LCD pannel handy I recommend you check MS samples and Steve's ClearType app and see that the results really are similar.

    Someone should ask Steve Gibson to opensource his app so that we can incorporate this tech into X display server (it's written in ASM so it's very fast to).


    --
    GroundAndPound.com [groundandpound.com] News and info for martial artists of all styles.
  • Well, the new bit is the calculations that go into making sure that the `color' is distributed properly. If you select the wrong gamma for your display, you get all sorts of colored fringes on the letters.
    --
    Ski-U-Mah!
    Stop the MPAA [opendvd.org]

  • I'm sure the analysis to get these results is much more sophisticated. Without a doubt it will be or is patented, but wouldn't a poor-man's approximation of this be to anti-alias (pixel interpolate) with blue tints on the right and orange tints on the left (instead of grey as is more standard)?

    I'll bet the algorithm is much simpler if you do it that way, and you'd probably get 90% of the results with 10% of the effort.
  • by konstant ( 63560 ) on Tuesday June 13, 2000 @08:37AM (#1005058)
    The Apple II did the same thing

    Since LCD monitors were not exactly prevalent at the time of the Apple II, and since MS Cleartype depends strictly upon the existence of a tri-bar distribution of colors that is found only in LCD's...

    Let's just say that having a bright idea and actually implementing that idea are two very different things. How many people in the /. audience have clever pet inventions that they just don't have the time, or the skill, to make a reality? Some day someone else will have the same idea, except that they will do the grunt work of actually producing a workable copy. Guess who will deserve the credit for bringing an innovative idea into the world?


    -konstant
    Yes! We are all individuals! I'm not!
  • It's possible that your display has a different ordering of the color cells, or perhaps you selected the wrong gamma..
    --
    Ski-U-Mah!
    Stop the MPAA [opendvd.org]
  • The anti-aliasing algorithm is really simple and typically works really well. Imagine a grid of squares (pixels), and you want to draw a thick diagonal line through this grid.

    Now, we'll encounter some problems. In a traditional pixel-drawing scheme, we would color all squares completely covered by this diagonal line, and any other square, even those partially covered by the line will not be colored.

    Zoom back from this drawing, and you have a very jagged, aliased line.

    Now the anti-aliasing scheme solves the jaggedness by a simple yet clever solution; the squares are shaded depending on what percentage of their area is covered. That is, if a square is 95% covered by the line, it will be very dark, and if it 5% covered, it will be very light. Zoom back from this, and you have a nice smooth anti-aliased line.

    Note that other anti-aliasing methods exist for differrent things, but the above is mainly for simple line drawing. For example, when anti-aliasing is done on a 3d rendered scene, a typical solution is to slightly shake the camera and re-render the scene, then blend the different renderings (this is known as oversampling).

    Now, this cleartype technology seems interesting. I'm not sure how it works, but judging from the examples given, the fonts look a bit "fuller" when larger on the screen. Also, very small fonts are very easily read, whereas in an antialiased scheme the smaller fonts are difficult to identify (this is probably because the antialiasing algorithm has trouble dealing with sub-pixel drawing instructions). One downside of cleartype, however, is that I noticed a fair amount of green and red color fringing on the small-to-medium fonts. I suppose this type of distortion is a side-effect of taking advantage of sub-pixel drawing technology.
  • by Anonymous Coward
    Yes, but it still won't work correctly because there is no way for the software to accurately determine the exact pixel placement and size on your screen. Remember that LCDs have a fixed resolution, while CRTs don't.
  • You can download a demo of the technology used behind ClearType, at Steve Gibson's page:

    http://grc.com/cleartype.htm [grc.com]

    It is configurable for all sorts of LCD layouts, including BGR and RGB. Unfortunately, it runs only on Windows, so this probably won't help you if you have an iMac... but perhaps you could do screen captures then view the graphics on the Mac?

    Of course, this was all done over 20 years ago, with the Apple ][: it displayed color, but internally everything was black and white! Due to the way the NTSC television signal works, it could create color by placing thin white vertical lines very close together in precise positions. It did this, essentially creating more subpixel resolution. The application was different, using black and white pixels to create color (essentially opposite to what is done now, using color pixels to create more black and white resolution), but the underlying technology solution was the same.

    I just hope Microsoft doesn't try to patent this! Hopefully, the Patent Office will see the mountain of prior art that is the Apple ][.

  • it was demonized months ago when the original story was posted and everyone said "This isn't any different from anti-aliasing!"
  • Former research IBM scientist Ron Feigenblatt has some interesting comments [geocities.com] about Microsoft ClearType. Feigenblatt explains subpixel addressing, dynamic pixels, and color convergence problems on LCDs.
  • Apple had color LCD screens in 1985? I'm pretty sure what they had was whole pixel (monochrome) anti-aliasing as opposed to sub-pixel rendering. It's different.

    And of course this does absolutely nothing for screens other than LCD..

    This is akin to complaining that the latest rechargeable laptop battery technology isn't good for anything other than laptop computers. CRT's don't even need this technology-- it's designed specifically to enhance the text quality on portable color LCD devices.

  • Slashdot groupthink had convinced me that MS never creates anything original... hmm... must think... *smoke* brian overheating...

    brian is overheating? poor brian. maybe he should go jump in a lake.

  • This is new?

    Some people have too much time on their hands. Putting a shiny label on something to sell it better is kinda, ironic. It's sorta like changing the date on a carton of milk...

    Perhaps it is innovative because it suchs more of your processor into viewing text (whoopee!)
  • You can definitely see an improvement in the smaller font sizes. The 6, 7, and 8 point are much more readable with the cleartype. The larger ones are more crisp too. Not a huge difference but it does look better.

  • Several people have already pointed out that ClearType does not represent innovation on Microsoft's part. Here is a related point for your consideration.

    When the Free Software / Open Source community creates something innovative, anyone can use it in any way, improve upon it, incorporate it into any kind of tool. Everyone benefits. This is the paradigm of science.

    When a corporation such as Microsoft creates something innovative and keeps it proprietary, no one can use it in any way except as the corporation permits. The lion's share of the benefit goes to the corporation, and further innovation is stifled. This is the paradigm of intellectual property.

    Anyone, including a corporation, has the right to handle their creations as they see fit...but where do you want to go today?

    ------

  • I don't think apple did this for the 80-column cards on apple-2 series, but I remember reading about something very similar to this in nibble waaaay back when.
    It wasn't specific for putting text onto the screen, but allowed one to put finer, more detailed hires graphics onto the display just by doing some calculations and setting the colour of the pixel right; depending on what the colour was set to, a pixel would be shifted about 1/2 to the left or right on the monitor. The upside was that it almost doubled your x-resolution on the screen. Unfortunately, you really had to turn the colour on the monitor down, otherwise you'd get some horrible colour fringes around all lines and stuff.
    They'd even implemented some routines in machine language to do all this quickly and were accessible from applesoft basic using the good ol' ampersand. Anyone else remember that?
    Fun stuff...

    --C
  • Yea.. and the colour count (for the gamma 1.8 image) is 256.. lotsa colours .. browns and blues.. hmm.. very nice.. The regular antialiased version was just greyscale.

    Looks totally pointless on my CRT though but I guess that's the point as it 'needs' an LCD.

    Oh well, at least it will keep the baby-bills busy.
    --
  • > Let's just say that having a bright idea and actually implementing that idea are two very different things. How many people in the /. audience have clever pet inventions that they just don't have the time, or the skill, to make a eality?

    I've been thinking about this for a bit. A lot of good ideas are lost because of technical problems. Myself I have a big list of things I want to do "when know how to". How about creating an "idea database". Anyone with a good idea can enter it. Others can browse those idea's to find a solution to a problem they have. If they find someone they can contact the original author to discuss it further and/or help him implementing it.
  • [summary: Steve Gibson -- http://grc.com/cleartype.htm -- is wrong about Apple II having invented what MS calls "Cleartype".]

    OK, but you tell Steve he's wrong...I know him, and there's no chance I'll do it.

  • by GSearle ( 40628 ) on Tuesday June 13, 2000 @09:05AM (#1005074) Homepage
    This is another example of Microsoft taking someone else's idea, calling it their own innovation, and making a big fuss over it. Really this guy [grc.com] has been pushing this technology for quite a while before Microsoft picked it up.

    The Apple II doesn't really do this. It uses the properties of the NTSC colorburst signal to create color from a synchronized high-resolution monochrome signal. The physical "subpixels" on the CRT can't be aligned to this signal, and the end result is fuzziness, not clarity. The R,G,B phosphors on the screen are not directly addressed.

    CRT's and even analog LCD's don't gain anything from this, as this technology needs direct access to the R,G,B elements of the display to create antialiased text that is as sharp as possible. It even needs to know the order of the RGB elements. This done through wholly digital displays that directly address the color pixels on the display, such as an LCD on a laptop. The next step would be to make this independent of the display type, with tuning tools or profiles for individual display devices.

    Television does this naturally, being a wholly analog system. Point a color camera at some text, and the edges of the text will fall on the color elements within the camera, irregardless of arbitrary pixel boundaries. If you magnify a still image on a TV set, you'll notice that any sharp edges are defined independantly of the positions of the color elements, and they are "smooth". In contrast, any computer-generated edges show a bias toward pixels, causing some jagginess in even the best anti-aliased graphic. Of course, if the source camera and the receiving TV set have different color element geometries, the result will be a little off.

    Whatever you call it, this is antialiasing taken to the max. I'm glad that someone is taking it seriously, even if it's Microsoft. CG for television should take notice, too, to try to simulate the natural look of purely analog signals.

  • Triangular distribution doesn't occur on aperture grille (Trinitron) monitors. However, even an aperture grille CRT will be highly unlikely to see the cleartype enhancements intended for LCDs.
  • by Anonymous Coward
    Um, here's a direct quote. This looks like sub-pixel rendering to me....
    The Apple II's highest resolution mode was 280 pixels horizontally by 192 vertically. However, this was really the 'sub-pixel' resolution. (Similar to the example above where an 800 pixel wide LCD is really 2400 sub-pixels wide.) The Apple II's display generated two sub-pixels per pixel. On an LCD display every third sub-pixel is Red, Green, or Blue and all three must be turned on to get white. On the Apple II, every other sub-pixel was green or purple and they both needed to be turned on in order to get white.

    here's more

    But thanks to Apple's built-in sub-pixel technology, white pixels were often composed from each half of adjacent whole pixels to yield a much smoother result.
  • Why don't you reference the linked pages before you flame someone?

    If you follow the links, you'll find Apple did in fact use sub-pixel rendering.
  • After a brief skim of the description of Apple II's sub-pixel rendering and the ClearType abstract I guess the best way to summarize the situation is Microsoft took anti-aliasing a step further by adding some math and taking into account the nature of LCD displays. Rather than linearly avergaing the RGB values of neighboring pixels (your basic anti-aliasing alogrithm) instead a larger area is sampled and FFT analysis applied to find the "optimal" RGB combination so the human eye detects the least amount of "image error". So I'm not so quick to doubt that Microsoft has come up with some new math for anti-aliasing, not that I think it's worthy of a Nobel Prize, but to call it "innovative", maybe. I'm not an anti-aliasing expert (INAAAE) so I can't say for sure.

    So why only LCDs? What about CRTs? I suspect what Microsoft is really after here is the handheld market. They'd like for PocketPCs leap ahead of Palm OS at least as far as display clarity goes. It may seem like a minor battle but Joe Consumer standing in Best Buy looks at chunky Palm OS two-color screen then sees bright and crisp PocketPC screen and buys that instead.

    What I find curious is ClearType has been presented at a few conferences and white papers submitted to IEEE yet I don't see any clear statement on how/if ClearType will be treated as a public standard, or released under some sort of licencse, or jealously protected by zealous IP lawyers. Could it be a case of adopt-now-and-we'll-bill-you-later? Maybe I just can't find a link to the license agreement. Whatever.

  • Let's just say that having a bright idea and actually implementing that idea are two very different things.

    Sure -- as Steve Jobs put it, "Real artists ship." But sub-pixel rendering was widely used in the Apple ][ days. Among other sources, it's documented in the MS Basic Interpreter refs.

    Since LCD monitors were not exactly prevalent at the time of the Apple II, and since MS Cleartype depends strictly upon the existence of a tri-bar distribution of colors that is found only in LCD's...

    True, the older method deals with two pixel colors. That's the basis for the "unprecedented" in LAS VEGAS -- Nov. 15, 1998 -- Tonight at COMDEX/Fall '98, Bill Gates, chairman and CEO of Microsoft Corp., unveiled an unprecedented innovation in font display technology during his keynote address. ?

  • On an LCD screen.. there are subpixels as follows
    RGBRGBRGB
    RGBRGBRGB
    RGBRGBRGB

    Traditionally, we treat any group of 'rgb' as a pixel, so.. to make a 'white' display, we need *any* 3 adjacent pixels.. they use this fact to effectively triple horizontal rezolution. ie: Insted of 50%r100%g50%b, on a diagonal line, the next one would be 100%g50%b50%r, etc...
  • You mean something like this [halfbakery.com]?

    engineers never lie; we just approximate the truth.
  • It seems to me that there are three questions that we must ask.

    1) Is the main purpose of this technology to provide a technically irrelevant selling point for the Pocket PCs?

    2) Is the main purpose of this technology to justify the cost of the processors to run the bloated software?

    3) Is the main purpose of this technology to show, when two PDAs are running side by side with ClearType, that the Pocket PC is the better machine(given the optimization)?

    When we know the answers to these questions, then we can decide whether this is innovative or just annoying.

  • They are actually taking advantage of the subpixel properties of an LCD.

    In an LCD, you have rgbrgbrgb... rather than saying that a pixel is 'rgb'.. a pixel can actually be any three adjacent subpixels, as the full color group will still be present. Effectively, they triple their horizontal resolution.
  • I remember a similar program for the Sinclair ZX81 that used a similar technique. It combined Apple's black and white pixel position tricks with flicker just slightly out of phase with the TV's refresh rate to create color. On an NTSC set, it was more like shades of gray with a slight tint. Word was it looked better on PAL. Either way, it needed big 16x16 px blocks to work, so the color bleed was far worse than a Spectrum's 8x8 px blocks. (IIRC. And by px, I mean real pixels, not the big 1/4-character pseudo-pixels the ZX81 used.)

    Wow, this is the second Slashdot story in the past week to remind me of those halcyon American-using-Sinclair days. Probably explains the Police kick I've been on lately. Ahh, the '80s! :-)

    Every day we're standing in a wind tunnel
    Facing down the future coming fast
    - Rush
  • with asm you /could/ say 'the app is the source'

    asm is moderately useless for putting into X, it has to be portable.
  • by YoJ ( 20860 )
    If you follow the previously posted links, you will see that the Apple II did in fact have subpixel rendering. The basic idea was that RB were on one side of the pixel and G was on the other side. So by drawing purple and green dots in the right places you could make lines anti-aliased. This is exactly the same fundamental idea as ClearType.

    I seem to remember some C64 demos that programmed the video hardware to get subpixel accuracy, but I think that was more of a hack of the hardware.

    nojw

  • by Misagon ( 1135 ) on Tuesday June 13, 2000 @09:30AM (#1005087)
    If you run xmag on your Netscape window, you will see that the Cleartype images are dithered somehow. The background color is also not solid white, but there are pink diagonal stripes. They are probably a by-product of converting the images to GIF.

    Another stupid thing about the images is that in neither of the images (Cleartype nor antialiased) is font hinting used. IMHO they should have compared hinted antialiased with hinted cleartype (with three times the horizontal resolution). Windows' font engine has hinting and antialiasing.

    Hinting, for you who don't know the term, is a way of snapping the font outline to the centers of pixels before rendering. This minimizes stupid solid grey lines in the edges for a well hinted font. The problem with hinting though, is that the quality of the hinting is dependent on the skill of the font designer.

  • Thats the first time I have heard of RGB being layed out in stripes. Traditionally it has been in triangles so that the RGB pattern is repeated every where. Like the example below, except they are circles and the touch,

    R G B R G B
    B R G B R G
    R G B R G B

    There is a shadow mask used to control the 3 electron beams. Everywhere you look there is always a triangle of RGB. I have never heard of them called stripes, but I guess you could. This tends to show the real reason the sceme works though.
  • especially if you're sick of the "gray fuzzies" that Acrobat's antialiasing produces... (I usually prefer ghostviewing PostScript versions when available.)

    Off-topic, but have you tried viewing PDFs in GhostView? They work nicely, since they're basically PostScript anyway.
    --
  • When Microsoft mentioned this around two years ago, my advisor posted his application of standard engineering practice (i.e., non-patentable) technique for making use of LCD subpixels. He posted it on the WWW simply to ensure that no company can patent these obvious, but perhaps useful, methods.

    See Color LCD Panel Subpixel Rendering [purdue.edu] by Prof. Hank Dietz, December 15, 1998.
    --

  • Since LCD monitors were not exactly prevalent at the time of the Apple II, and since MS Cleartype depends strictly upon the existence of a tri-bar distribution of colors that is found only in LCD's...

    Except for the small fact that you don't need to use a modern LCD panel to use this technology. I used to program quite heavily for my old Apple //c, and on the color output (plain old NTSC) this was made use of in several applications. Televisions *do* use 3-color "pixels", and at least DeluxeDraw used this technique to draw "straight" lines on a page. Also, the rendering engine that Sierra came out with (the name escapes me...) would use this technology to draw vector graphics on the screen.

    The wheel is turning but the hamster is dead.

  • by Anonymous Coward
    Ironically, this paper couldn't be released sooner because it was rejected the first time it was submitted for publication. Why, you ask? Because during the anonymous review process the reviewers felt that the technology was too similar to Microsoft's Cleartype!
  • The samples are disappointing. The improvements are marginal (only very visible at diagonal strokes such as in the digits 7 and 2).

    On the other hand, both examples look much better than the non-antialiased text I'm used to looking at under Linux... :-/

  • The hardware very is different, but both modern LCDs and old Apple displays have the concept of "subpixels". They exist for different reasons, but the software concepts used to take advantage of them is remarkably similar.

    --

  • I don't think it's just that the colors are in a triangular distribution. It's also that they're not in a predefined, addressable grid. (Which is why you can run your monitor at all sorts of random resolutions without resorting to the ugly "stretching" used to get 640x480 on 1024x768 LCDs.)

    --

  • When Microsoft mentioned this around two years ago, my advisor posted his application of standard engineering practice (i.e., non-patentable) technique for making use of LCD subpixels. He posted it on the WWW simply to ensure that no company can patent these obvious, but perhaps useful, methods.

    The thing is... his method appears to use a simple box-filter; not the complex filter used in ClearType.

    The difference may be marginal; ClearType certainly looks better on my HP Jornada than your advisor's solution though. And the difference is what counts; it's the type of filtering algorithm invented that is patentable - and that algorithm is an order of magnitude more complex than the simple one your advisor came up with (similar to Steve Gibson's one).

    Not to mention the other gears and cogs involved in getting it up onscreen; they're a siderunner.

    Simon
  • by Anonymous Coward
    It is true that the sony trinitron has RGB stripes...but unfortunately each group of 3 stripes does not correspond to a single pixel in display memory, like they do on a LCD monitor where each pixel in display memory is fixed to a specific pixel on the monitor. (which is why you can't change resolutions on a LCD without resulting to lossy scaling techniques)
  • Except for the small fact that you don't need to use a modern LCD panel to use this technology. I used to program quite heavily for my old Apple //c, and on the color output (plain old NTSC) this was made use of in several applications. Televisions *do* use 3-color "pixels", and at least DeluxeDraw used this technique to draw "straight" lines on a page. Also, the rendering engine that Sierra came out with (the name escapes me...) would use this technology to draw vector graphics on the screen.

    As any television engineer will tell you though, there's no way of addressing the INDIVIDUAL phosphors on the screen; so no, Apple didn't have "subpixel rendering".

    What it does have though is the ability to push its monochrome signal (in composite video/NTSC form) at a high enough frequency (ie. high resolution) that it went out of the defined NTSC bandwidth for a luma signal, and poked its way into the chroma. This is why two pixels together ended up as white (or black); because the width of the on/off pulse was wide enough to be encompassed in the luma part of the signal.

    When you half that width, you end up with the pixels poking into the chroma part of the signal; depending on the phase (ie if it's the left or right hand part of the pixel pair that's set), you end up with green or magenta.

    So it's not quite subpixel resolution. And it's useless for antialiasing (the luminosity distribution between white, black, green and magenta has obviously not been chosen to give antialiasing a helping hand). Well, you can get a half-level, I guess, if you use Magenta. But that's about it.

    In conclusion, Apple had a way of forcing color out of a monochrome signal (much like the ZX81, which gave you blue and yellow on a PAL screen if you futzed with the output display), but it wasn't anything to do with splitting pixels, or subpixel rendering, or any kind of attempt to increase the effective resolution of the display. Because believe it or not, if you magnify that television display, more than one phosphor will be lit-up green for the green pixels, and more than one phosphor will be lit-up magenta for the magenta pixels.

    Of course, getting Steve Gibson to acknowledge this (especially as he dissed Microsoft as copying Apple, and then implemented his "Free n Clear" with a several-orders-of-magnitude more complicated algorithm to reduce color fringing without batting an eyelid) is a task left to the reader.

    Simon
  • The idea behind Clear-Type is to make WinCE the dominant platform for e-books and web pads. Along with this, M$ is developing proprietary digital watermark/copy protection. Publishers will go for it - they hate the thought of us sharing our e-books without paying more royalties. If all the good content is only available in M$'s proprietary format, what kind of reader will you buy? When they have dominant market share in this space, look for them to hijack the standards process. Think Active Server Pages - Cool technology that requires M$ on the server end.
  • Thats the first time I have heard of RGB being layed out in stripes. ... There is a shadow mask used to control the 3 electron beams.

    Well, LCD screens don't use electron beams. And we are talking about LCD screens here... they use colored filters, with the colors laid out in stripes.

  • Where can I get more info about that? I always wanted to know how those worked.
  • No mistake. There's nothing new here - this type of display was used as far back as the Apple II's which only had 4 colors. Magenta and green or blue and orange were combined to create white. This white was then used to render the characters on the screen, with the two colors used to antialias the character.

    Don't be such a troll, especially when you don't know what you're talking about.

  • You've made all these judgments about ClearType, but none of you have looked at it on a PocketPC display! I mean, the webpage with the samples flat-out TELLS you that a CRT can't display it correctly, and you won't be able to appreciate it. Does ANYONE out there in have a PocketPC so they can tell us if it really works?
  • I seem to remember reading somewhere that for most of the current batch of e-books, the orientation of the LCD screen is 90 degrees off - that is, it's rotated so that the pixels go R-G-B vertically, not horizontally. This makes the technique pretty useless for text on these devices.
  • ...would be more like 50 pages, and it would actually contain equations. You can use harmonic analysis for lots of things, and this is a pretty simple application, that arguably doesn't really rely on it to work, rather Microsoft just uses it to give a hand-wavy justification of why it works.
  • A very concise and easy to understand explanation.
  • heeheehee

    The iMac does not have an lcd, silly. Maybe they meant iBook.

    Tom
  • Basically, you fake a 3x resolution increase to make edges smoother.

    My question is, will this technology be relevant once monitors at 3072x2304 are common? The extra resolution would surely be wasted on making your browser window or text editor larger -- because your eyes can only read so much. But if you used it to enchance the readability of everything on the screen, making all your edges sharper -- it could make things much easier to read, scale, and enjoy.

    1024x768 screens are fairly uncrowded. They give you lots of space to overlap windows when you need to work between a couple of applications, and provide a nice size for working in one application. But once you go past to resolutions of 1280x960 or 1600x1200, you begin to not see any useful gains. "Everything" begins to shrink, leaving more and more useless white space, while using an exponential increase in video ram and monitor bandwidth/scan time.

    Perhaps the real application of a technology like this is not for using the properties of LCDs to fake more pixels, but in rendering low-resolution 1024x768 screen on a high resolution 3072x2304 screen with anti-aliasing and proper kerning, and other smoothing techniques to make it as crisp as any piece of paper. This could make 2D cards require an extra bit of logic on the silicon, but the benefits could be many.

    I'm sure 3D graphics would benenfit too, considering how anti-aliasing could make the "low" resolution of 1024x768 look fine (the Voodoo5's FAA makes lower resolutions just as good looking as higher resolutions withouth FAA). Instead of worrying about larger memory usage by video cards, you could just work towards a fixed target, and let the monitor's increased resolution + the logic on the card do full screen anti-aliasing without as much special logic as the Voodoo5 employs, and it'd look crisper because you actually have those extra pixels to use (instead of just relying on the eye to average the shades of adjacent pixels).

    Just a thought :)
    ---
  • I obviously have nothing blatantly new to add to this topic since I have read all these posts which make me feel like I am lost in a nerd maze. But, from an idiots view...I expected more. I of course have your basic crt, so anything on these cleatypes is a loss for me, and I am sad about that. I would be so happy to have something that helps the headaches and blurred vision I get from spendig my life in front of this screen. I couldn't care less who's company developes, rips off, or copyrights it. Anything to help my poor failing eyesight would be welcomed.
  • ...that low-pass filters (exactly the same ones that are used in traditional antialiasing) work better on LCDs if sampling is done separately for three channels with taking phase shift between grids of red, green and blue pixels into account. While explanation of the reasons, measurement of difference between "most perfect" filtering and low-pass filters, and demonstration may be interesting, the idea itself is rather obvious and hardly can be considered a great triumph of human thought.
  • by be-fan ( 61476 ) on Tuesday June 13, 2000 @11:47AM (#1005115)
    Yea, I know it is kind of mean to harp on a system that is older than I am, but I think it is ridiculous that X is so far behind in modern display technology. (It is also ridiculous that we are still using it after so many years, but that's another story.) Coming from the all anti-aliased all the time world of BeOS, I notice that X has really poor font handling. If you read any of the FreeType docs, you'll see constant bitching about how X really doesn't provide any font support worth a damn. While X just recently got integrated truetype support, and is still far away from having anti-aliasing (which even Windows! has) Microsoft is introducing new technology to make fonts look even better! Aside from these small scope issues, this fact points to larger problems with the X architecture. It just wasn't designed to be extended cleanly. As such, X (at XFree86) still doesn't have genuinly usefull things like DPS and anti-aliasing. This is yet another reason that X needs
    A) A huge, major rewrite.
    B) To be replaced with something else.
    Berlin looks very promising. True it uses CORBA for its API services, which sacrifices speed, but even at the beta level that it is in, it already has nifty features like anti-aliasing support, and a totally cool imaging model. Also, it is designed to be extended cleanly.
  • Now the anti-aliasing scheme solves the jaggedness by a simple yet clever solution; the squares are shaded depending on what percentage of their area is covered. That is, if a square is 95% covered by the line, it will be very dark, and if it 5% covered, it will be very light. Zoom back from this, and you have a nice smooth anti-aliased line.


    This is one particular implementation of an antialiasing technique, which you might see listed in the appropriate literature as "reconstruction via box filtering" or "weighted area" sampling. It can be fairly simple and effective to implement, but it is far from the best possible results. Using box filters for reconstruction overly blurs the underlying "perfect" image that you are trying to reconstruct. In digital signal processing, you would typically use a sinc(x) filter (sinc(x) = sin(x)/x) to reconstruct, which is theoretically perfect, but expensive to evaluate. More moderate filters based upon cubic bsplines walk a pretty good line between sharpness and expense.


    Note that other anti-aliasing methods exist for differrent things, but the above is mainly for simple line drawing. For example, when anti-aliasing is done on a 3d rendered scene, a typical solution is to slightly shake the camera and re-render the scene, then blend the different renderings (this is known as oversampling).


    This is a type of oversampling that is currently getting alot of mileage because graphics hardware from SGI and now other graphics companies can do this in hardware, typically using an "accumulation buffer". This type of antialiasing isn't without its own share of artifacts, resulting from the correlations between subpixel locations. Oversampling in general has been around for a long time, it was part of Turner Whitted's original 1980 raytracing paper.

  • I looked at it on my primary surfing system, an Acer TravelMate 720 series laptop PC with a 1024x768 active matrix color LCD. Looks beautiful. It'd look even better on Nintendo's Game Boy Color, where the red, green, and blue vertical stripes are even more pronounced.
  • No, but colors in an NTSC signal are horizontally multiplexed in a similar way to colors on the LCD panel of the laptop computer I'm typing this on.
  • Yes, Microsoft Research has a lot of good people doing some interesting stuff. Keep in mind, though, that many of them have only been there for a couple of years, though, and the excellent work they have been doing at Microsoft has just been a continuation of work done elsewhere (ClearType may be an exception).

    Whether Microsoft Research will be a long term fixture among top industrial research labs remains to be seen. It's also unclear whether Microsoft Research will be any more successful at technology transfer into products than other labs. Apple, for comparison, used to spend a lot of money on research (Gates was bragging publically he didn't have to spend any, he'd just take Apple's results) but disintegrated within a couple of years when the company wasn't doing so well.

    With their wealth and promise of stock options, Microsoft has had an easy time luring away people from other research labs until now. Losing so many good people has been pretty bad for those other labs, as you can imagine, and that has given Microsoft Research the reputation of begin a raider (in line with the generally aggressive stance of Microsoft).

  • When looking at the demo on a Thinkpad laptop, I don't actually see much of an improvement. I'm somewhat surprised; after all, I thought having roughly three times the number of addressable pixels ought to help significantly.

    Does it actually look like a significant improvement to you?

  • I hope you mean a FONT protocol, not another damn "extension".

    If anti-aliasing requires using different X calls (and thus does not cause existing programs to antialias) I will kill the first XF86 representative I see.

    (To be realistic, I expect antialiasing to work if a truecolor visual is used and copy transfer function is in use.)

    The only "extension" I want is to add a simple X call: XSetTheDamnFont("Name of font goes here", point_size); This should ALWAYS set the font no matter what garbage I put in the name, set it to a default font if not found. And I should be able to get italic if the string contains the word "italic" in it.

    Well, I could also use UTF-8 encoding.

  • by spitzak ( 4019 ) on Tuesday June 13, 2000 @01:33PM (#1005133) Homepage
    The normal antialiased example is not really fair. The "61 point lancos filter" sounds impressive, and it is (it is a 61-pixel wide filter with 31 positive numbers and 30 negatives and is a very accurate simulation of the infinitely-wide "sync" filter). Unfortunately it also strongly biased towards preserving frequency information, at an extreme it could consider representing a 1-pixel wide 1.0 line as n lines that sum to 1 with a 0 line between each of them. This can be very good for photographic information but fails a bit for high contrast fine detail, which text is probably an excellent example.

    A better filter for text would be a cheapo box filter. The worst it can turn a 1-pixel black line into is two .5 lines that are adjacent. Adjusting for the screen gamma (their examples *do* do this) will make a much more "even" result than this fancy filter (and would be dozens of times faster). In particular the thin lines would all appear to be much closer to the same weight (it is also necessary that the original binary image have lines of even thickness).

    The "cleartype" I think looks somewhat sharper because it uses this simpler filter. You can also see dithering artifacts all over the white area to the right, so I believe they are rendering at 3x width and then translating this to the rgb values of each pixel, but using something like error diffusion to keep the average total color rendered "gray". I think they should reset the error after some number of pixels of solid color because your eye will not see this:

    RGBRGB.....RGBR

    as more red than this:

    RGBRGB.....RGB

    This would get rid of the dithering artifacts on the right.

    In any case it actually is an innovative idea. The truth about innovative ideas is that they seem quite obvious once somebody thinks of them. It also means that it is almost impossible to stop real "innovation" from being used by everybody, since just seeing the output provides enough information for the innovation to be recreated. In fact I would expect this to appear in lots of commercial and free products quite soon.

  • look here [grc.com]. This will better help to understand of what we are speaking, what is interesting about it (and what is old news).
  • When talking about Microsoft Research [microsoft.com], don't give a pointer to Microsoft [microsoft.com]. This is almost like confusing, say, NASA [nasa.gov] and U.S. Government (here [whitehouse.gov] or here [congress.gov]).
    --
  • The effect may not be as pronounced on a vertically striped CRT as on an LCD, but I can actually see the difference on my Sony Trinitron (tm) monitor. The text is noticeably sharper in all the sizes.

    Interestingly, though, the basics for this technology has been around for a while. Display stripes are not the only source of color-biased (and hence reverse biasable) geometric effects. Image pickup devices such as single tube video camera of a couple decades ago, and current color CCD cameras, have these effects even more extreme because of the more coarse striping that exists.

    I would suggest that a correction for the effect of the display device could be applied in general to the whole video signal being displayed, and there would be improvement overall. Then type hinting would still have to take into consideration that the snap points aren't the same for each color plane (part of what is apparently going on in Cleartype, anyway).

  • Bitmaps captured from laptop screens with cleartype are going to be messed up if you view them on a different kind of monitor, but I guess this isn't as much of an issue for PocketPCs. They should, however, consider having cleartype rendering being optional so that people can get useful screen captures for web sites etc.
  • The Apple II doesn't really do this. It uses the properties of the NTSC colorburst signal to create color from a synchronized high-resolution monochrome signal. The physical "subpixels" on the CRT can't be aligned to this signal, and the end result is fuzziness, not clarity. The R,G,B phosphors on the screen are not directly addressed.

    No, but there is still a color shift effect which results from using pixeling to effect NTSC "modulation". I remember seeing at least one program on my old long-gone Apple ][ which attemtped to compensate for this effect. That was limited by the lack of ability to change pixel value. I did play around with time-coding it to get that effect anyway on long-duration color photographs made from the CRT screen.

    CRT's and even analog LCD's don't gain anything from this, as this technology needs direct access to the R,G,B elements of the display to create antialiased text that is as sharp as possible. It even needs to know the order of the RGB elements. This done through wholly digital displays that directly address the color pixels on the display, such as an LCD on a laptop. The next step would be to make this independent of the display type, with tuning tools or profiles for individual display devices.

    Yes they do. It may not be as much, and certainly more complex to compensate for since you don't get to syncronize the effect at the pixel level. But a CRT, especially a vertically striped one, has the effect of shifting the color planes relative to each other at sub-pixel levels, and the effect used in ClearType partially compensates for this. The correction is a matter of degree and probably could be optimized for a specific CRT tube at a specific scan geometry. But it is very real and I can visually see it in the images from the Microsoft Research page. What ClearType is probably also doing is adjusting the snap points as well so that hinting applies differently in different color planes (and thus not necessarily shifting the color planes by the same amount for each character cell).

    Television does this naturally, being a wholly analog system. Point a color camera at some text, and the edges of the text will fall on the color elements within the camera, irregardless of arbitrary pixel boundaries. If you magnify a still image on a TV set, you'll notice that any sharp edges are defined independantly of the positions of the color elements, and they are "smooth". In contrast, any computer-generated edges show a bias toward pixels, causing some jagginess in even the best anti-aliased graphic. Of course, if the source camera and the receiving TV set have different color element geometries, the result will be a little off.

    When it comes to single tube (vidicon) and color CCD image pickup devices, then television actually has more of a problem with just the color plane shifting effect. I remember reading an article in one of the broadcasting tech journals at the time regarding this, and how some circuitry was used to correct this in the YUV color space, as opposed to the RGB color space. It happened to mention that CRT displays introduced a muted form of the same effect, but could not be corrected for because of all the variations that existed. I remember thinking at the time that it would be relatively simple to correct for that for a Sony Trinitron display, if I could first split the video into RGB signals, then apply a time-base corrector. It would not have been cost effective for a consumer device. But the sharp, crisp, image I saw on Tektronix studio monitors back then (late 1970's, when I worked in broadcasting) made me wonder if they were employing some of this. Perhaps we should inquire of Tek as to what research they have done with this, as well as Kodak, which I know has done research to correct things like this in their CCD cameras.

  • I was definitely wondering about that. I suspect a simpler approach (and certainly not patentable in the context of the clear thinking minds of /. readers) is to simply perform the type rendering, and hinting, in separate color spaces shifted by an amount intended to correct for the particular display in use, be it discrete LCD, analog LCD, or striped CRT. Still, that alone is not the whole effect. There is an effect of the fuzziness of the edges varying by horizontal position with respect to pixel alignment, and this, too, varies by color plane (absolute for discrete LCD). The Fourier Analysis could come into play to correct for this, though I suspect some pre-generated masks would really be all that would be needed.

  • You're correct that the Apple ][ didn't strictly do this, but Woz's moneysaving homebrew video hardware did do something that made a very similar technique possible. It essentially allowed a type of subpixel addressing, if you consider the black and white 140x192 display to be the pixels. Coloured pixels were actually the left and right halves of a white pixel (blue and orange or green and magenta depending on the half-pixel shift of the set). Since a lot of people (including me) took sneaky advantage of this technique to antialias their black and white text and images (before most of us knew what "antialias" meant), it's a bit facetious for Micros~1 to use the term "unprecedented" at the top of their Cleartype page, since they really are doing the same thing, right down to the fact that they're turning a slight display weakness into a strength for one particular purpose.

    The reason this comes to so many people's minds (and the real proof of the similarity, really), is probably the fact that the colour fringing on the Cleartype samples looks a lot like the colour fringing on an Apple ][ display. Try running an emulator with the colour burst simulated if you've never seen it firsthand. Quite the nostalgia trip actually.

    As for whether it's really any good, well, it's merely ok. I'm looking at it on a powerbook 2400 screen (RGB striped 96dpi display pretty similar to what's in colour PDAs), and the difference with the standard grey antialiasing isn't all that great; arguably the colour fringing makes it six of one, half a dozen of the other (ObPrisoner). It seems a little clearer if you look at a gamma lower than your real screen, or maybe mine really is at 1.6 these days. I wouldn't pay extra for it, anyway.

  • I tried it on both a Dell Inspirion 7000 and my Sharp mobilon HC-4500. On both screens I noticed very distracting color fringing. Like I said it may just be the way I perceive colors. But at this point if I get anything (software or hardware) that included this technology it had better have a way to turn it off! I'd rather read plain non anti-aliased chunk type than this rainbow colored eye fest :)

One person's error is another person's data.

Working...