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

Re: Color fringe hires behavior



Bryan Parkoff wrote:
"Michael J. Mahon" <mjmahon@aol.com> wrote in message puOdnSegHN1A_sfYnZ2dnUVZ_vqdnZ2d@comcast.com">news:puOdnSegHN1A_sfYnZ2dnUVZ_vqdnZ2d@comcast.com...

BLuRry wrote:

The independence of these two problems strongly suggests that any
emulation should layer the display rendering into a "system-dependent"
part and a "display-dependent" part, so that users could select the
emulated display that they would like to "attach" to the system
emulator, and so that new "displays" could be easily emulated.


I agree 100%.  The approach I'm taking is to generate everything in the
form of a monochrome 560x192 bit pattern.  From there it can be copied
to the screen as-is for monochrome, or via some form of translation
(simple 4-bit rgb conversion, ntsc approximation, etc)  The ntsc part
will wait, but I think I have a handle on the "convert to a 560x192 bit
pattern" part.

Sounds great!

IIRC, the MAME project is doing an NTSC emulation, or was...


I can tell that MAME and MESS are doing an attempt to emulate NTSC, but it does not look great. You want solid white pixel. Emulated NTSC looks like four vertical lines. First and third vertical lines are white and second and fourth vertical lines are less than white like little light gray. When you try to convert from NTSC signal to RGB signal using conventor device, you try to look at the RGB monitor closely, you may notice vertical lines.

Since the video bandwidth of *any* NTSC monitor displaying composite
color video will be limited to less than 4MHz in the luminance channel
(and much less in the chrominance channels), the 14MHz dot clock of the
Apple II will be completely unresolvable.

Further, the Apple video output for white is a high video level
*uninterrupted* by any transitions, so any visible fine structure
is an artifact of the conversion and not a property of the video
or of NTSC encoding.

If one wished to get *really* authentic, then a high-resolution display
might show the color dot clusters at the emulated CRT's shadow mask
pitch, but that would be, perhaps, more authentic than anyone would
choose to watch.  ;-)  (It would, of course, also be much darker than
the "host" display, since much of the emulated screen is actually dark.)

Up until now, I've never considered the color "fringing" that results
when even a monochrome (no burst) display is shown on a color CRT. Any
"beat" between the pixel pitch and the shadow mask pitch will result in
color fringing because of incomplete illumination of some color triads.
Typically, the dot pitch is much finer than the pixel pitch, and the
dot pitch is sufficiently fine that the eye tends not to see it at
normal viewing distance, but it *is* an effect that might be emulated.

-michael

NadaNet networking for Apple II computers!
Home page:  http://members.aol.com/MJMahon/

"The wastebasket is our most important design
tool--and it's seriously underused."