Lord Myn wrote on Yesterday, 08:52:

I have a few games which are known to compatible EGA or VGA only, but there are CGA compatible versions of them: […]
Show full quote

I have a few games which are known to compatible EGA or VGA only, but there are CGA compatible versions of them:

Rise of the Dragon (16 color version works on CGA - 640x200 mono)
Heart of China (16 color version works on CGA - 640x200 mono)
Police Quest III (16 color version works on CGA with compatible driver)
Castle of Dr. Brain (16 color version works on CGA with compatible driver)
King's Quest V (16 color version works on CGA with compatible driver)
Space Quest I SCI (16 color version works on CGA with compatible driver)
Space Quest IV (16 color version works on CGA with compatible driver)
Leisure Suit Larry 1 SCI (16 color version works on CGA with compatible driver)
Leisure Suit Larry 5 (16 color version works on CGA with compatible driver)
Operation Stealth (early version works on CGA)
Vette! (there is a separate CGA version)

Rise of the Dragon is interesting, because the CGA mode just doesn't look right in emulation, the dot spacing is off for some reason.
I've taken a photo of how it looks on an PC/XT class PC with on-board CGA.
The PC was a Siemens Nixdorf 8810 M35, from ca. 1984.
Re: MCGA Games (PC/DOS) - LCD vs CRT

What I wonder, though, is why 640x200 mono and not 320x200 4c was used.
The dithering used is ordered dithering (?) and it's hard to make things out against the dithered background.
Someone should think that 320x200 16c to 320x200 4c conversion
would have made more sense than 320x200 16c to 640x200 mono.

Because the readability of the text didn't improve either by using 640x200 mono, I think.
That was different with the old Sierra games.
In higher resolution mode, such as Hercules, the text font was usually of higher quality, at least.

I think the CGA mono mode was used because the pseudo color dithering is more functional in double resolution than in 4 colors. The same is true for the rest of the Sierra SCI1 games, they only work with CGA320M.DRV (640x200x2) driver from QFG2 or KQ1SCI (they don't support 4-color CGA mode either), every prior driver (e.g. CGA320C or CGA320BW) will freeze, because screen scrolling wasn't implemented in them.

Jo22 wrote on Yesterday, 11:15:

Rise of the Dragon is interesting, because the CGA mode just doesn't look right in emulation, the dot spacing is off for some re […]
Show full quote

Lord Myn wrote on Yesterday, 08:52:

I have a few games which are known to compatible EGA or VGA only, but there are CGA compatible versions of them: […]
Show full quote

I have a few games which are known to compatible EGA or VGA only, but there are CGA compatible versions of them:

Rise of the Dragon (16 color version works on CGA - 640x200 mono)
Heart of China (16 color version works on CGA - 640x200 mono)
Police Quest III (16 color version works on CGA with compatible driver)
Castle of Dr. Brain (16 color version works on CGA with compatible driver)
King's Quest V (16 color version works on CGA with compatible driver)
Space Quest I SCI (16 color version works on CGA with compatible driver)
Space Quest IV (16 color version works on CGA with compatible driver)
Leisure Suit Larry 1 SCI (16 color version works on CGA with compatible driver)
Leisure Suit Larry 5 (16 color version works on CGA with compatible driver)
Operation Stealth (early version works on CGA)
Vette! (there is a separate CGA version)

Rise of the Dragon is interesting, because the CGA mode just doesn't look right in emulation, the dot spacing is off for some reason.
I've taken a photo of how it looks on an PC/XT class PC with on-board CGA.
The PC was a Siemens Nixdorf 8810 M35, from ca. 1984.
Re: MCGA Games (PC/DOS) - LCD vs CRT

What I wonder, though, is why 640x200 mono and not 320x200 4c was used.
The dithering used is ordered dithering (?) and it's hard to make things out against the dithered background.
Someone should think that 320x200 16c to 320x200 4c conversion
would have made more sense than 320x200 16c to 640x200 mono.

Because the readability of the text didn't improve either by using 640x200 mono, I think.
That was different with the old Sierra games.
In higher resolution mode, such as Hercules, the text font was usually of higher quality, at least.

Lord Myn wrote on Yesterday, 11:35:

I think the CGA mono mode was used because the pseudo color dithering is more functional in double resolution than in 4 colors. The same is true for the rest of the Sierra SCI1 games, they only work with CGA320M.DRV (640x200x2) driver from QFG2 or KQ1SCI (they don't support 4-color CGA mode either), every prior driver (e.g. CGA320C or CGA320BW) will freeze, because screen scrolling wasn't implemented in them.

I do agree in principle, dithering works better in 640x200.
But if source material is of same resolution as 4 colour CGA mode (320x200),
then couldn't have the dithering been omitted entirely?

I mean, on a monochrome monitor (say, green or amber) the designers basically had 4 levels of gray available. Like on a GameBoy.
This would have been enough to redraw the pictures in a way that objects would be easily recognizable.

Or was monochrome mode chosen because it was easier to auto-convert through a driver at runtime?
If the 320x200 16c source material was chosen for input it makes sense to me, since it's already heavily being dithered.

It just seems odd that 4 color mode wasn't being utilized from very start to make pretty images.
The outline of the characters and objects wouldn't have had needed to change,
the 2 two extra colours (besides black/white) could have been used to smooth out the edges (like anti-aliasing)..
The overall result would have been much closer to the 256c original than using ordered dithering.

Hm. I wonder if there's more to it.
a) Maybe the designers just didn't care about CGA
b) The Dynamix team at the time had no access to a software to easily down-convert to 4 grays (something equivalent to IrfanView of today)
c) Monochrome mode was chosen because it looked more digniful than "ugly" 4 colour mode in cyan/magenta/white/black (if a colour monitor was used)

Edit: I didn't mean to get on someone's nerves,
but I wondered about this for years, because the EGA version was said to be redrawn.

That's why I wondered why the CGA graphics are so hard to see, why they are the way they are.
If there was such an effort spent to optimize down to 16c EGA, then 4c CGA wasn't far away, either.

By my logic, I mean. 4c CGA would have had a larger audience than EGA, maybe, justifying a redraw by hand even more.
By 1990, the iron courtain just fell and many eastern countries had CGA compatible PCs, still.

What comes to mind is the Poisk-1 home computer which simulated CGA in software.
Or these East German PCs, such as Robotron PC 7150 or EC1834.
They had offered 640x400 in 16c (out of 4096) using same graphics chip as PC-98, but the PC compatibility was limited to CGA modes.

I mean, sure, these weren't PCs the ordinary citizens had access to.
But from a technological point of view, CGA still had some relevance as a lowest common denominator at the time.

The only good reason for redrawing in 320x200 16c besides EGA was Tandy 1000 support, maybe.
Because Tandy 1000 series was big in the US, the home land of many game developers.

But then, same time, I wonder why graphics hadn't been redrawn for 640x200 16c.
Because both EGA and Tandy (later models?) supported this resolution on an ordinary 15 KHz monitor.
This would also have fit resolution wise to CGA hi-res mode, all modes would have been 640x200 pixels I mean.

I think the color mode is dropped, because of the real-time dithering. I have a unique (early) KQ5 version with 4 different CGA drivers, there are two 4-color modes and two monochrome (LCD and CRT), the text in 4-color modes are barely readable or the graphics lose significant details:

Color Mode 1

Color Mode 2

Monochrome 1 (LCD)

Monochrome 2 (CRT)

Jo22 wrote on Yesterday, 12:36:

I do agree in principle, dithering works better in 640x200. But if source material is of same resolution as 4 colour CGA mode (3 […]
Show full quote

Lord Myn wrote on Yesterday, 11:35:

I think the CGA mono mode was used because the pseudo color dithering is more functional in double resolution than in 4 colors. The same is true for the rest of the Sierra SCI1 games, they only work with CGA320M.DRV (640x200x2) driver from QFG2 or KQ1SCI (they don't support 4-color CGA mode either), every prior driver (e.g. CGA320C or CGA320BW) will freeze, because screen scrolling wasn't implemented in them.

I do agree in principle, dithering works better in 640x200.
But if source material is of same resolution as 4 colour CGA mode (320x200),
then couldn't have the dithering been omitted entirely?

I mean, on a monochrome monitor (say, green or amber) the designers basically had 4 levels of gray available. Like on a GameBoy.
This would have been enough to redraw the pictures in a way that objects would be easily recognizable.

Or was monochrome mode chosen because it was easier to auto-convert through a driver at runtime?
If the 320x200 16c source material was chosen for input it makes sense to me, since it's already heavily being dithered.

It just seems odd that 4 color mode wasn't being utilized from very start to make pretty images.
The outline of the characters and objects wouldn't have had needed to change,
the 2 two extra colours (besides black/white) could have been used to smooth out the edges (like anti-aliasing)..
The overall result would have been much closer to the 256c original than using ordered dithering.

Hm. I wonder if there's more to it.
a) Maybe the designers just didn't care about CGA
b) The Dynamix team at the time had no access to a software to easily down-convert to 4 grays (something equivalent to IrfanView of today)
c) Monochrome mode was chosen because it looked more digniful than "ugly" 4 colour mode in cyan/magenta/white/black (if a colour monitor was used)

Lord Myn wrote on Yesterday, 13:21:

I think the color mode is dropped, because of the real-time dithering. I have a unique (early) KQ5 version with 4 different CGA […]
Show full quote

I think the color mode is dropped, because of the real-time dithering. I have a unique (early) KQ5 version with 4 different CGA drivers, there are two 4-color modes and two monochrome (LCD and CRT), the text in 4-color modes are barely readable or the graphics lose significant details:

The attachment image0001.png is no longer available

Color Mode 1

The attachment image0002.png is no longer available

Color Mode 2

The attachment image0003.png is no longer available

Monochrome 1 (LCD)

The attachment image0004.png is no longer available

Monochrome 2 (CRT)

Jo22 wrote on Yesterday, 12:36:

I do agree in principle, dithering works better in 640x200. But if source material is of same resolution as 4 colour CGA mode (3 […]
Show full quote

Lord Myn wrote on Yesterday, 11:35:

I think the CGA mono mode was used because the pseudo color dithering is more functional in double resolution than in 4 colors. The same is true for the rest of the Sierra SCI1 games, they only work with CGA320M.DRV (640x200x2) driver from QFG2 or KQ1SCI (they don't support 4-color CGA mode either), every prior driver (e.g. CGA320C or CGA320BW) will freeze, because screen scrolling wasn't implemented in them.

I do agree in principle, dithering works better in 640x200.
But if source material is of same resolution as 4 colour CGA mode (320x200),
then couldn't have the dithering been omitted entirely?

I mean, on a monochrome monitor (say, green or amber) the designers basically had 4 levels of gray available. Like on a GameBoy.
This would have been enough to redraw the pictures in a way that objects would be easily recognizable.

Or was monochrome mode chosen because it was easier to auto-convert through a driver at runtime?
If the 320x200 16c source material was chosen for input it makes sense to me, since it's already heavily being dithered.

It just seems odd that 4 color mode wasn't being utilized from very start to make pretty images.
The outline of the characters and objects wouldn't have had needed to change,
the 2 two extra colours (besides black/white) could have been used to smooth out the edges (like anti-aliasing)..
The overall result would have been much closer to the 256c original than using ordered dithering.

Hm. I wonder if there's more to it.
a) Maybe the designers just didn't care about CGA
b) The Dynamix team at the time had no access to a software to easily down-convert to 4 grays (something equivalent to IrfanView of today)
c) Monochrome mode was chosen because it looked more digniful than "ugly" 4 colour mode in cyan/magenta/white/black (if a colour monitor was used)

Those look sort of lazy, i get it, auto dither but text boxes (one would hope) could be an exception to howvthe rest of the imagery is parsed.

My guess is that a minor change to the font graphics assets (outline/shape/width/etc) would have made the text much more readable in all modes (even if the same dithering jank touched the text)

Jo22 wrote on Yesterday, 11:15:

What I wonder, though, is why 640x200 mono and not 320x200 4c was used. The dithering used is ordered dithering (?) and it's har […]
Show full quote

What I wonder, though, is why 640x200 mono and not 320x200 4c was used.
The dithering used is ordered dithering (?) and it's hard to make things out against the dithered background.
Someone should think that 320x200 16c to 320x200 4c conversion
would have made more sense than 320x200 16c to 640x200 mono.

Possibly, this is not intended to be viewed at 640x200 mono, but as pseudo 160x200 composite artifact color, see https://en.wikipedia.org/wiki/Color_Graphics_ … artifact_colors .

Some Swedish guy developed a Sierra graphics driver for the Olivetti/AT&T 640x400 monochrome mode back in the day, initially as shareware, but I believe he later released it for free.

I wonder how well that third-party driver would work with the 16-color editions of those newer SCI Sierra games, such as KQ5.

I remember trying that driver with QfG1, but I didn't notice any visual improvement over over the included 640x200 monochrome CGA driver. Perhaps the later games would make better use of the potential improved dithering abilities at twice the vertical resolution.

Then again, I don't know if that third-party driver had any screen scrolling support.

Perhaps such a 640x400 driver could be redone with an improved dithering algorithm.

It would also be interesting if an alternative SCI engine could be developed (perhaps a fork of FreeSCI or ScummVM, but with real-mode 8086 compatibility), which would actually draw the vector graphics of those games at higher resolutions than 320x200.