8

I was playing around with an emulated Apple /// and while looking at the color demo I noticed that unlike the Apple ][, the /// did not implement red as one of its supported colors.

Ignoring strange color systems like CGA or "close implemented" ones like the VIC (II) chips, I find it puzzling that a system wouldn't support red as a color: not only should it be relatively easy from the NTSC standard (given how every raster is made up of Red/Green/Blue), but also since color on the /// was mostly driven through an RGB interface which I presume was digital and would have supported red (e.g. EGA should be close..?)

Any idea why red wouldn't have been supported?

2
  • 1
    Are there any reddish colors listed? Some machines that derive color by dividing a chroma signal by four end up having a reddish brown, a reddish orange and a reddish purple, but not necessarily any color that one would simply call "red".
    – supercat
    Commented Feb 19, 2019 at 19:47
  • 1
    See also: retrocomputing.stackexchange.com/a/6274/75 Commented Feb 19, 2019 at 19:51

1 Answer 1

11

I was playing around with an emulated Apple /// and while looking at the color demo I noticed that unlike the Apple ][,

Not really, as the Apple II also didn't do red, just orange.

Apple II colours were black, green, purple and white plus blue and orange when shifted (*1),

the /// did not implement red as one of its supported colors.

While in Apple II mode, the AIII supported the same resolutions and colours as the AII - and as the A2 doing so only with NTSC encoding.

I find it puzzling that a system wouldn't support red as a color: not only should it be relatively easy from the NTSC standard (given how every raster is made up of Red/Green/Blue),

Erm, you seem to be mixing up several independent issues here. NTSC is an encoding scheme to transport colour information as a single channel using a phase shift, while RGB is an encoding using three different, independent channels. NTSC in turn is not an RGB signal (*2). Further the use of three (or more) colour sections (and guns) in a colour CRT is neither related to a signaling scheme nor the way these colours are generated by the source.

but also since color on the /// was mostly driven through an RGB interface which I presume was digital and would have supported red (e.g. EGA should be close..?)

Again no. The Apple III doesn't use a 3 colour RGB scheme, its designers even went a step further and implemented a clean 4 bit digital interface. The video logic loads one nibble per pixel into the output buffer, where an external colour encoder can use this (in addition the the sync signal) to generate any picture.

What colours an Apple III screen shows is encoded within the monitor, not the computer

To get the 4 bit digital signal into an RGB signal an encoder may be built from a simple resistor network as shown in the Apple III Owners Guide:

enter image description here

By using this encoder the colours should look much like the (BASIC) manual names them:

enter image description here

As with any system where the display interprets the data, alternate interpretations are possible - like this with the ability to tune colour levels - or one with just 4 resistors, that would producing eight different colours with two levels by simply using the lower 3 bits as colour and the top most as intensity. Now including a clear red ... or make up whatever table you like.

For the B&W output these 4 bits are added by a resistor array (*2) into a 16 level grey scale signal. Similar for NTSC, except here the colour comes from an A2 alike encoding where any 4 bit value unlike zero will produce the full intensity (*3).

Bottom line, in A2 mode it outputs an NTSC signal exactly like the A2, in A3 mode a 4 bit full digital signal is output and interpreted by the monitor ... much like with today's displays.


*1 - See this answer for a more detailed description.

*2 - And mixed with the sync signals as well.

*3 - Well, it's a bit more complex than that - or better it's simple, but rather complicated to explain. It all revolves of encoding the 4 bits into two NTSC 'clock' lines with by an 74LS153 mux controlled by the 7MHz colour signal and its 3.5MHz 'brother'. A look at the video encoding circuit may help.

5
  • 1
    So that particular encoder produces the same colours as Apple II low res colour mode — the 40x48 mode?
    – Tommy
    Commented Feb 19, 2019 at 23:56
  • @Tommy yes, that's the idea.
    – Raffzahn
    Commented Feb 20, 2019 at 3:23
  • I read this a few times, and I still don't understand why red wasn't one of the colors. Was the omission of red an arbitrary choice, or is it that once you choose a resistor network with four bits of input, red becomes impossible, or is it that if you make red one of your colors, then some other, more desirable colors become impossible with a resistor network? Commented Feb 20, 2019 at 20:44
  • 1
    @WayneConrad Point is that the computer didn't produce colour at all when outputing to the colour connector. Just TTL signals. And in Apple II mode colours where output like the Apple II did. Colour monitors sold for the Apple II (Not sure it there where many) had decoders decoding the 4 bit words according to (mostly) the same colours as the Apple 2 low res was.
    – Raffzahn
    Commented Feb 20, 2019 at 23:41
  • @WayneConrad The color number assignment for the Apple III have been chosen to imitate the colors available on the Apple II (low res, as commented by Tommy), they were not chosen from scratch with RGB in mind. The Apple II doesn't have an RGB output, but creates colors as "artifact colors" from high-frequency pixel patterns that get interpreted as colors by color NTSC video displays. The way the artifact color generation works on the Apple II was unable to generate red, and that's why the Apple III colors also don't have red. Commented Sep 4, 2021 at 19:47

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .