Subversion Repositories planix.SVN

.TH COLOR 6

.SH NAME

color \- representation of pixels and colors

.SH DESCRIPTION

To address problems of consistency and portability among applications,

Plan 9 uses a fixed color map, called

.BR rgbv ,

on 8-bit-per-pixel displays.

Although this avoids problems caused by multiplexing color maps between

applications, it requires that the color map chosen be suitable for most purposes

and usable for all.

Other systems that use fixed color maps tend to sample the color cube

uniformly, which has advantages\(emmapping from a (red, green, blue) triple

to the color map and back again is easy\(embut ignores an important property

of the human visual system: eyes are

much more sensitive to small changes in intensity than

to changes in hue.

Sampling the color cube uniformly gives a color map with many different

hues, but only a few shades of each.

Continuous tone images converted into such maps demonstrate conspicuous

artifacts.

.PP

Rather than dice the color cube into subregions of

size 6\(mu6\(mu6 (as in Netscape Navigator) or 8\(mu8\(mu4

(as in previous releases of Plan 9), picking 1 color in each,

the

.B rgbv

color map uses a 4\(mu4\(mu4 subdivision, with

4 shades in each subcube.

The idea is to reduce the color resolution by dicing

the color cube into fewer cells, and to use the extra space to increase the intensity

resolution.

This results in 16 grey shades (4 grey subcubes with

4 samples in each), 13 shades of each primary and secondary color (3 subcubes

with 4 samples plus black) and a reasonable selection of colors covering the

rest of the color cube.

The advantage is better representation of

continuous tones.

.PP

The following function computes the 256 3-byte entries in the color map:

.IP

.EX

.ta 6n +6n +6n +6n

void

setmaprgbv(uchar cmap[256][3])

{

    uchar *c;

    int r, g, b, v;

    int num, den;

    int i, j;

    for(r=0,i=0; r!=4; r++)

      for(v=0; v!=4; v++,i+=16)

        for(g=0,j=v-r; g!=4; g++)

          for(b=0; b!=4; b++,j++){

            c = cmap[i+(j&15)];

            den = r;

            if(g > den)

                den = g;

            if(b > den)

                den = b;

            if(den == 0) /* would divide check; pick grey shades */

                c[0] = c[1] = c[2] = 17*v;

            else{

                num = 17*(4*den+v);

                c[0] = r*num/den;

                c[1] = g*num/den;

                c[2] = b*num/den;

            }

          }

}

.EE

.PP

There are 4 nested loops to pick the (red,green,blue) coordinates of the subcube,

and the value (intensity) within the subcube, indexed by

.BR r ,

.BR g ,

.BR b ,

and

.BR v ,

whence

the name

.IR rgbv .

The peculiar order in which the color map is indexed is designed to distribute the

grey shades uniformly through the map\(emthe

.IR i 'th

grey shade,

.RI 0<= i <=15

has index

.IR i ×17,

with black going to 0 and white to 255.

Therefore, when a call to

.B draw

converts a 1, 2 or 4 bit-per-pixel picture to 8 bits per pixel (which it does

by replicating the pixels' bits), the converted pixel values are the appropriate

grey shades.

.PP

The

.B rgbv

map is not gamma-corrected, for two reasons.  First, photographic

film and television are both normally under-corrected, the former by an

accident of physics and the latter by NTSC's design.

Second, we require extra color resolution at low intensities because of the

non-linear response and adaptation of the human visual system.

Properly

gamma-corrected displays with adequate low-intensity resolution pack the

high-intensity parts of the color cube with colors whose differences are

almost imperceptible.

Either reason suggests concentrating

the available intensities at the low end of the range.

.PP

On `true-color' displays with separate values for the red, green, and blue

components of a pixel, the values are chosen so 0 represents no intensity (black) and the

maximum value (255 for an 8-bit-per-color display) represents full intensity (e.g., full red).

Common display depths are 24 bits per pixel, with 8 bits per color in order

red, green, blue, and 16 bits per pixel, with 5 bits of red, 6 bits of green, and 5 bits of blue.

.PP

Colors may also be created with an opacity factor called

.BR alpha ,

which is scaled so 0 represents fully transparent and 255 represents opaque color.

The alpha is

.I premultiplied

into the other channels, as described in the paper by Porter and Duff cited in

.IR draw (2).

The function

.B setalpha

(see

.IR allocimage (2))

aids the initialization of color values with non-trivial alpha.

.PP

The packing of pixels into bytes and words is odd.

For compatibility with VGA frame buffers, the bits within a

pixel byte are in big-endian order (leftmost pixel is most

significant bits in byte), while bytes within a pixel are packed in little-endian

order.  Pixels are stored in contiguous bytes.  This results

in unintuitive pixel formats. For example, for the RGB24 format,

the byte ordering is blue, green, red.

.PP

To maintain a constant external representation,

the

.IR draw (3)

interface

as well as the 

various graphics libraries represent colors 

by 32-bit numbers, as described in 

.IR color (2).

.SH "SEE ALSO"

.IR color (2),

.IR graphics (2),

.IR draw (2)