[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Newbie Apple IIgs questions

Mitchell Spector wrote:

"Michael J. Mahon" <mjmahon@aol.com> wrote:



David Murray wrote:
Secondly, I get color once the OS starts loading, but for some reason the control panel wouldn't give color. Of course, I was surprised to see that same nasty color-phase-shift problem that older Apple's suffered from. I thought surely by the time the IIgs was out Apple would have fixed that problem. That is such a disappointment. I'll try building that adapter to connect to my Amiga monitor's RGB input and I can pray that maybe it will look better there. 

The color "aliasing" that you see on a composite video display is
not an "error", but the correct functioning of the method used to
create a color display.  It's just that in a text display, for example,
you'd rather not see the colors.

Intrinsic color bandwidth limitations of a composite video signal
prevent producing high-resolution color images.

RGB is totally different, and you will see exactly what you hope for.
While that's true of composite, it's worth noting the IIgs does not produce a standard composite signal. The end result is something 
actually sub-par and few steps below composite.


The deviations from NTSC are minor timing variances, particularly
in the vertical sync pulse.  One effect of which is the collapsing
of the the normal interlaced raster to a non-interlaced one.

In all other respects, the Apple composite video signal is
"acceptable" as an NTSC signal.


    Good example: Look at the fonts displayed inside a Finder window
on a TV set and the text actually appears in heavy bold (by no means subtle) bands of green, blue, purple and orange. The whole screen is quite literally a rainbow colored mess. Now switch over to output from a Video Overlay Card (which produces a corrected composite signal) and the black text appears greyish black. Any traces of color shifting 
is subtle and faint, you really need to stare closely at the screen.


The video overlay card does two things:  it creates standard sync and
blanking signals (correcting the minor timing differences) and it does
anti-aliasing of the chroma signal, so that there is less color artifact
on, for example, dot pairs.  If a single dot is displayed, and is not
shaped/stretched, then it must appear colored, since it represents
an appreciable amount of chroma energy at that position.

The "graying" of the black is a sign of anti-aliasing to reduce
chroma artifacts.  This is appropriate in a screen known to be
monochrome, but it is not appropriate if the screen is supposed
to be color.  Normally, the "intent" is reflected in the presence
or absence of a color burst on the "back porch" of the horizontal
blanking pulse, but most monitors require many scan lines to respond
to a change in color burst status, so entire fields are either color
or monochrome, not partial fields.
Another example. Look at a checkerboard pattern of black and white pixels (say the background pattern in PaintWorks Plus). From the IIgs's built-in composite port you get alternating patterns of blue and orange bars which blend to look purple. The Video Overlay Card shows it as a solid grey background (normal for composite to blur it that way). 

Actually, alternating black and white pixels at a frequency of 3.58MHz
should display as a solid color (whose hue is determined by phase--
whether the odd or even pixels are on or off).  Again, this is exactly
how the Apple II is _designed_ to create color displays.

At higher-than-HGR resolutions, the frequency produced by alternating
pixels is higher, and should not contain significant 3.58MHz components,
and so should appear to be about 50% gray.


A separate luminance-chrominance video signal (as used by some C=64
systems) can support better monochrome resolution and less color
artifacting.



    I've used my Commodore 64 with both the split chroma-luma (essentially
S-Video) and regular composite, and even the latter produces a superior
video signal compared with the IIgs's composite-out. The same is true with
my Amiga, Nintendo, Super Nintendo or any other composite device.


The C=64 and most video games use a technique of creating color video
that is more like the Apple II Lo-Res color, or Apple II DHR color,
based on 4-bit patterns shifted out at 14.3Mhz.


    Interesting, the color artificating appearing in Super-Hi-Res composite
should be green, blue, purple and orange (I never really looked closely at it
until now). Sound familiar? :)


These are the colors corresponding to in-phase, 90 degrees advanced, out
of phase, and 270 degrees advanced, the phase positions available to a
digital device with a 14.3MHz (4 x 3.58MHz) clock.

-michael

New Applesoft BASIC interface for NadaNet networking!
Home page:  http://members.aol.com/MJMahon/