The 9 colors of the CoCo’s high resolution screens…

I was today year’s old when I realized the CoCo’s high-resolution color values actually make sense.

Let me just get this right out in the open…

“I feel dumb I never realized the PMODE colors in the manual were correct.”

– Allen

You all probably knew this. If I did, I forgot I knew it. And I really don’t think I ever knew it. Because dumb.

My first CoCo was a grey TRS-80 Color Computer that I got around 1984. Today we call it a “CoCo 1” but back then it was the only CoCo.

The CoCo had eight colors (nine if you count black) that it could display:

Color Adjustment Test Display, Radio Shack TRS-80 Color Computer Quick Reference Guide, page 55.

The colors are listed as:

  • 0 – Black
  • 1- Green
  • 2 – Yellow
  • 3 – Blue
  • 4 – Red
  • 5 – Buff (I always called it white)
  • 6 – Cyan
  • 7 – Magenta
  • 8 – Orange

When you use the high resolution graphics (PMODEs), you have a choice of a higher high-resolution with 2 colors, or lower high-resolution with 4 colors. Each of these modes has two color sets, which select which 4 or 2 colors you can use on the screen. Here is how the Extended Color BASIC manual describe the PMODEs:

There are some high resolution graphics commands that let you specify a color — such as the COLOR, CIRCLE and DRAW — let you specify a color value to use. I found it odd that the color value was listed as “0-8” — as if you could use nine colors on a high resolution screen. You can’t! Only four, or two.

But the manual listed nine colors, anyway:

The first dumb thing I never realized until today was those high-resolution color numbers are the same as the ones for the text screen, as used by text commands CLS and SET. As I previously wrote about, I also did not realize that the colors available on the high resolution screens were actually the same as the ones on the text screen. I’d never seen them at the same time, and they always “looked different” to me.

But why did Microsoft let you specify color values 0-8 for a screen that can only have four colors? When I started playing with the CoCo again as a grown-up, I assumed Extended Color BASIC must have been ported from a system with more colors and they just left it as-is.

But the manual was weird about it. For example, it would tell you use use colors 5, 6, 7 and 8!

“If you want to try changing the dots’ colors, use buff (5), cyan (6), or magenta (7). Then change the color back to orange (8) before proceeding…”

– Extended Color BASIC manual, page 86

But I knew back then that this was just silly. If you used colors 1, 2, 3 and 4 in one mode, you could also use colors 1, 2, 3 and 4 in the other color set. You’d get the four unique colors in each mode, just different colors based on the mode. (Or you could use 0, 1, 2 and 3 just shifting the order over.)

The important part of the manual was listing which four colors you got when you used SCREEN to specify the color set to use:

SCREEN 1,0 gave you the green/yellow/blue/red colors on a green screen with green border, and SCREEN 1,1 gave you the buff/cyan/magenta/orange colors on a buff screen with a buff border.

Then, there were the higher high-resolution modes that only had two colors. The manual mentioned them (and the four color modes) later:

I guess they were introducing things slowly as you progressed through the chapters.

In a 2-color mode, SCREEN 1,0 gives you a green border with black background and you can draw in green. SCREEN 1,1 gives you a black screen with a white background and you can draw in white.

And my mind was blown when I realized that, though you could use 1-4 on any 4-color screen, or 1-2 on any 2-color screen, and get different colors, the way the manual tried to teach us was basically: “In this mode, use colors A to B. In that mode, use colors C to D” and so on.

Behold! PMODE 3,1:SCREEN 1,0 5-color mode showing what it draws for all 9 colors (0-8):

If you use the color chart, you can see you should be using colors 1-4:

  • 1- Green
  • 2 – Yellow
  • 3 – Blue
  • 4 – Red

If you go to PMODE 3,1:SCREEN 1,1 you get these colors:

And here is why the manual told you to use colors 5-8:

  • 5 – Buff
  • 6 – Cyan
  • 7 – Magenta
  • 8 – Orange

And in the 2-color mode, if you went PMODE 4,1:SCREEN 1,0 you would get the same two colors repeated over and over:

And that corresponds to using colors 0 and 1 in the chart:

  • 0 – Black
  • 1- Green

And PMODE 4,1:SCREEN 1,1 looks like this:

Here, you would want to use colors 0 and 5:

  • 0 – Black
  • 5 – Buff

I don’t know if ANY of us ever programmed using the numbers like that — mostly because the manual really did not make it clear which colors were part of which mode. You’d need a fancier chart that showed what you could do, like:

0 or 20XX
0 or 21XX
1 or 30XXXX
1 or 31XXXX
The “correct” color values to use in high-resolution graphics ;-)

I would have needed to keep this chart nearby for reference until I memorized it.

What color values did you learn to use when you learned Extended Color BASIC? Leave a comment and let me know.

Until next time…

Bonus Code

Here’s my sample program.

20 GOSUB 500
40 GOSUB 500
60 GOSUB 500
80 GOSUB 500

499 END

505 DRAW"BM10,3 D8R8U8NL8 BR20 ND8 BR28 R8D4L8D4R8 BR20 R8U4NL8U4NL8 BR20 D4R8NU4D4 BR20 R8U4L8U4R8 BR20 NR8D8R8U4NL8BU4 BR20 R8D8 BR20 R8U8L8D4R8"
510 FOR C=0 TO 8
530 LINE (C*28,16)-(C*28+27,191),PSET,BF
540 NEXT
550 IF INKEY$="" THEN 550

8 thoughts on “The 9 colors of the CoCo’s high resolution screens…

  1. Erico

    One very interesting use of these colors would be with ppoint for collisions if they are registered differently. Similar can be done with semgraphics4 with the various blacks

    1. Allen Huffman Post author

      Have you actually tried this? I would expect it would read the data from the byte on the screen —- so a four color would have 00, 01, 10 or 11 bits. It would only be able to give a result back, so I would expect 0 or 4 to appear as the same color and PPOINT(x,y) would give you back just 0, even if you put it there using 4.

      Great. Time to test.

    2. Allen Huffman Post author

      Doing a test, where I drew dots using colors 0-7 on PMODE 3, I find that PPOINT returns the “correct” values — 5, 6, 7, 8 — over and over. Interesting. I need to write another part!

        1. Allen Huffman Post author

          Oooh and you could use that for ‘paths’ on a solid black background, where one indicates where you could travel and the other (looking the same) would not allow travel.

  2. Sean Patrick Conner

    Extended Color Basic made … odd choices. The 6847 in the Coco only supports 9 colors in text mode, but only a max of four in any of its 8 graphics modes, of which ECB only supports 5. ECB also uses a “logical” resolution of 256×192 for each mode, even though most the modes are half the height or width (or both). It even allows one to specify any of the 8 colors, but will map them to either 2 or 4, depending upon the mode selected.

    On the four color graphics screens, you can select from two sets of four colors, one set being green, yellow, blue and red, the other set being buff (white), cyan, magenta and orange. What’s happening in ECB as far as color goes, the color you select is “ACTUAL_COLOR = (COLOR – 1) & 3” (i.e. the range from 1-8 is being adjusted to 0-7, then masked to be in the range of 0-3). When you call PPOINT(), the pixel value is incremented, and if you used SCREEN1,1, then 4 is being added to get the range 5-8. If you used SCREEN1,0, then this +4 adjustment isn’t done.

    Personally, I think this way of doing things is confusing, and would prefer for the width and height to be the actual pixel width and height of the mode selected, and the colors being returned be 0-3 (or 0-1 for two-color mode).

    1. Allen Huffman Post author

      The only advantage I can see of using the same coordinates is you could draw something and it would render on any of the modes. But, I agree, I wish it just called it what it was. It makes things more complicated when you have to /2 or *2 to scale.


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.