the first two inputs encode the x and y coordinates of a pixel as 8-bit integers. 0 bits are encoded as salt atoms and 1 bits are encoded as fire atoms.
the second two inputs are the constants 0 and 255, encoded as 8-bit integers in the same way.
each solution is simulated until it drops a single output, or to 50000 cycles, whichever comes first.
if a solution reaches 50000 cycles without an output, the pixel corresponding to the x and y inputs is colored black.
there are two color modes—palettized and true color.
the quintessence output is the palettized output. atoms dropped on this output color the pixel according to a palette that roughly matches the colors of the atoms themselves.
the 24-atom output is the true color output. compounds dropped on this output are interpreted as 24-bit RGB colors. a salt atom is interpreted as a 0 bit, and anything else is interpreted as a 1 bit.