Re: IIe double modes

[Jonathan Schattke <wizwom@ripco.com>]

Mike Pfaiffer wrote:
> 
> Jonathan Schattke wrote:
> >
> > I got a second-run //e to help me verify AppleWin's functionality, and
> > one of the first things I did was to check out the Double-Hires &
> > double-LoRes graphics.
> 
>         What ia AppleWin?

An emulator.

> > The following anomalies occurred:
> >
> > If I turn on AN3 with poke-16290,0 and have the machine in 40-column
> > mode, Lo-Res mode is 40x40, but the colors are as if the screen was
> > directly mapped to hi-res.
> 
>         As I recall it you should activate the 80 column card first. If the
> card is active you can switch back and forth between 40 and 80 columns.
> This will correspond to a 40 or 80 column grahics screen. If the 80
> column card is inactive you should only be able at access a 40x40 lores
> screen.

You should.  I was misbehaving, and knew it.  It is called "Hacking". 
The wonder came when I did the odd, though :)
 
> > If I then turn on 80-column mode, the lores is 80x40, but the colors for
> > the aux bank pixels are mixed up, 0,8,1,9,2,10,3,11,4,12,5,13,6,14,7,15
> > corresponding to 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,and 15.
> 
>         Without checking out everything it looks like the colour values are
> rotated right by one pixel. Since each byte controls two pixels try
> doing a direct poke into the text page and see what happens. This way
> you can see what's up.

Can you see why from the dicussion below?
 
> > Also, I seem to get slightly different colors in DHR mode: 32 of them (I
> > guess this is because a DHR pixel is half a pixel).

> For DHGR the process is similar in as much as the storage of pixels in
> monochrome mode are identical. The exception is the first set of pixels
> is stored in the AUX bank where as the next byte takes place in the MAIN
> bank.
> 
>         Colour DHGR is interesting. Each pixel is made up of four bits stored
> almost as you would expect. Here is what I mean with the first byte in
> the AUX bank and the second byte in the MAIN bank. P111 0000 P332 2221.
> Assuming the pallet bit remains the same the result should be sixteen
> colours. Here is the REALLY interesting bit... Are the colours the same
> if you change the pallet? If so, you could end up with 30 colours (black
> and white would remain the same for each pallet).

Not quite.  DHR starts a color as soon as it hits a pixel.  The pallette
bit is ignored.  The position of the pixel determines what color it
starts displaying, i.e. if the position would be even, it will be green,
if even +1/2, orange, if odd, purple, if odd+1/2 blue.

Now, if additional pixels are on, the system will modify the color burst
accordingly.  The lores colors are supposed to be evidenced when the
pixel pattern matches that of a lores color (which, conveniently, are
not mapped).
But, starting with a '1', there are 8 binary patterns less than 16.  And
when you consider that these patterns can start in any of 4 collumns,
you get the 32  displayed colors.  much of this is fringe effect: a
solid pattern will drop into one of the 15 lores colors.
1000 - 1/2 pixel, dim dot of color (dk grn, brown, purple, dk blu)
1100 - 1 pixel, hires color (green, orange, purple, blue)
1110 - 1 1/2 pixel, brighter hires color (Lt Grn, yellow, Pink, Lt Blue)
1111 - 2 pixels, white (fringed)
1010 - 1/2 pixel then 1/2 pixel -one of the greys (this gives us the 15
LoRes)
now the weirder stuff:
1001 - 1/2 pixel, then mix in 1/2 pixel 
1101 - pixel then 1/2 pixel
1011 - 1/2 pixel, then pixel
These will settle into being the above patterns, after the initial
burst.

> > Do these anomalies happen with a IIc or enhanced IIe?
> 
>         It should...

Could you check for me?
 
> > I'd really like AppleWin to be 100% accurate.
> 
>         What is AppleWin?

It emulates an Apple //e under Windows - hence the name, AppleWin.  It's
too bad that the IBM PC floppy drive has a layer between the bus and the
floppy drive, but I guess that's what you give up to have DMA floppies. 
As it is, I have to settle on disk images.

You can download AppleWin 1.11.B from
<http://pages.ripco.com:8080/~wizwom/applewin/applewin_dist.zip>

Just extract it into it's own directory, and run "Apple for Windows.exe"

I have not quite got the AN3 on lores mod done, although it's simple
enough.
Also there is a minor calibration bug - on extremely fast machines,
calibration takes a very long time.

-- 
\  /\  /                                       If you aren't happy,
 VVizVVom - Wandering Kernel of Happiness           it's not my fault.
 ~~~> homepage at http://www.ripco.com/~wizwom <~~~