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

Re: 15Hz to VGA converter



Michael J. Mahon wrote:
Michael J. Mahon wrote:

drkenb wrote:

On Aug 13, 4:18 am, Michael J. Mahon <mjma...@aol.com> wrote:

Mark Frischknecht <appleii...@gmail.com> wrote:

Thanks to a tip James Littlejohn gave me I purchased a 15Khz RGB CGA
YUV Component Video YCbCr to VGA Converter Scaler from
http://www.ambery.com/rgbcgatovgac.html



  Wiring up the cabe was quite easy:



IIGS                          VGA 15 pin
Pin 1 Gnd -------->  Pin 5 Gnd
Pin 2 Red -------->  Pin 1 Red
Pin 3 CSYNC --->  Pin 13 HSYNC / CSYNC
Pin 5 Grn --------->  Pin 2 Grn
Pin 9 Blu --------->  Pin 3 Blu



I posted pictures of the video output on my Picasa albums:
http://picasaweb.google.com/AppleIIguy/CGAComponentToVGA#



It looks like 640x400 dithered doesn't come out to well...



It would be good to see the GSOS desktop.  That seems to be an "acid
test" of a converter's response to 16MHz dither--and a good indication
of what a user will see all the time.



Hi Michael - He does have one screenshot of the Finder desktop.  And
you're correct - that is the acid test for these devices!



When I "paged" through them yesterday, I didn't see the desktop.
Yep--pretty ugly.

There are only two choices for good conversion:  1) phase lock to
the *actual* dot clock of the source, or 2) oversample by a significant
factor (I'd say 3x or better) to avoid aliasing.  So far, no affordable
converter seems to do either of these things.

After resampling to VGA standards, if the video is presented to a
digital display (LCD), then the LCD's logic is responsible for further
resampling to its own dot pitch--which a good display should do without
further artifacts, assuming that the display's resolution exceeds the
resolution of the IIgs video (which for a VGA display is certain).

It is perfectly fine to display the fine lines of the dithering
without any need to "blur" them into a uniform background.  A fine
"crosshatch" is also visually pleasing.  A coarse aliased "beat"
between the alternating fine vertical lines and the sampling clock
is *not* pleasing.

To sum up:  the problem is *not* excessive display resolution.  The
problem is visible "beats" between the IIgs dot clock and the sampling
clock.

One way to prevent the beats would be to limit the bandwidth of IIgs
video signals to about 8-10MHz, but that would blur all the features
of the display, too.  I suspect most IIgs users would not like the
compromise, but it might be worth a try...


Whoops!  That bandwidth limit should have been *less* than 8MHz, since
8MHz is the fundamental frequency of the desktop background (alternating
signal toggling at 16MHz).

Since the impedance of the video lines is on the order of 75 ohms, it
would be fun to experiment with bridging the video lines to ground
with capacitors in the range of 47nF.  That should make the converter's
output more visually pleasing, but blurrier.  ;-)

Whoops squared!  ;-)

I blew the decimal point.  The shunt capacitors should be about
470pF.  (That's why I need an FPE.  ;-)

-michael

NadaNet and AppleCrate II: parallel computing for Apple II computers!
Home page: http://home.comcast.net/~mjmahon

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