timothym@tekigm.UUCP (Timothy D Margeson) (09/26/85)
Just a short note about the way to do things in color.... In article <372@bbncc5.UUCP> sdyer@bbncc5.UUCP (Steve Dyer) writes: >> Hmmm... 256 out of 32K colors on a standard PC color display? Sounds >> interesting. I wonder if they're working on a higher resolution version >> that would run on the EGS (256 by 350 pixels, perhaps?) > >No, on a color composite video monitor or analog RGB display. It's >impossible to display more than 16 colors on an IBM-CGA-compatible >monitor or 64 colors on an Enhanced Color Display: the RGB signals >are digital, not analog, and there are only 4 separate inputs >(or 6, for the ECD). >-- >/Steve Dyer >{harvard,seismo}!bbnccv!bbncc5!sdyer >sdyer@bbncc5.ARPA Well buddy, you are wrong. For one thing, the RGB inputs, if driven strictly digital would yield only 8 colors, but with an I input, you can get 16 colors (both cases include black). The scheme, or method these others are using to obtain >16 colors is simple in concept, a little more difficult in scope. They are pulse width modulating the different digital lines during the normal pixel period (or slightly longer than normal, hence 256 bit res.). This technique will yield MANY more color options on your normal RGBI monitor. If it is still not understood how this works, here is an example. For normal operation (as an IBM color card does it), say we want to put one violet pixel displayed on the screen. We simply turn on the red and blue guns for 1 pixel time. Viola, violet is shown. For this enhanced method, say we want to have a color somewhere between the violet above and red. In this case we turn on the red gun for one pixel time as above, BUT, we only turn the blue gun on for 1/2 a pixel time. This results in our eyes perceiving red-violet in stead of what we saw above. IBM Color: __________ R _____| |______ G _______________________ __________ B _____| |______ __________ I _____| |______ To give light violet or dim violet. Improved: __________ R _____| |______ G _______________________ _____ B ________| |________ __________ I _____| |______ This line is probably modulated also. So, I hope you can understand this method better now. -- Tim Margeson (206)253-5240 tektronix!tekigm!timothym @@ 'Who said that?' PO Box 3500 d/s C1-465 Vancouver, WA. 98665
hr@uicsl.UUCP (09/30/85)
<>
"The scheme, or method these others are using to obtain >16 colors
is.. [to] pulse width modulat[e] the different digital lines during
the normal pixel period (or slightly longer than normal..."
How well does this work? It seems to me that halving a pulse to a pixel
would require higher bandwidth amps in the monitor. Perhaps the smaller
pulse width isn't so small that it's filtered out. As for "slightly longer
than normal", how much wider can the pulse be? If it were too long, wouldn't
the beam just strike the shadow mask? Or, don't the normal pulses paint
the entire phosphor dot?
harold ravlin
{ihnp4,pur-ee}!uiucdcs!uicsl!hrrupprech@bbnccv.UUCP (Wolfgang Rupprecht) (10/01/85)
In article <579@tekigm.UUCP> timothym@tekigm.UUCP (Timothy D Margeson) writes: >For this enhanced method, say we want to have a color somewhere between >the violet above and red. In this case we turn on the red gun for one >pixel time as above, BUT, we only turn the blue gun on for 1/2 a pixel >time. Initially, this is what I thought too. Unfortunately, this would require them to effectively run the monitor at TWICE its normal bandwidth. It could also cause bizarre shadow effects on higher res (.21 mm triad spacing) monitors. The red-violet of your example would most likely look like a small violet pixel with a faint red shadow to the left and right. I suggest that they are really modulating the colors by 'interlacing'. The red-violet in the above example would be created by having the pixel be red-and-blue (ie. violet) for one frame and only red for the next frame. This has the advantage of not raising the bandwidth requirements of the monitor. I wonder if that form of 30 hz color flicker is visible. Anybody know? -wr
glen@intelca.UUCP (Glen Shires) (10/03/85)
> The scheme, or method these others are using to obtain >16 colors is > simple in concept, a little more difficult in scope. They are pulse > width modulating the different digital lines during the normal pixel > period (or slightly longer than normal, hence 256 bit res.). > Who has seen these cards/monitors? Are the colors consistent between different units? ...between different monitors? I would think that some amount of tweaking is necessary to get the colors consistent. -- ^ ^ Glen Shires, Intel, Santa Clara, Ca. O O Usenet: {ucbvax!amd,pur-ee,hplabs}!intelca!glen > ARPA: "amd!intelca!glen"@BERKELEY \-/ --- stay mellow