A Minor Color-Theory Epiphany

I’ve spent the bet­ter part of the after­noon in the stu­dio work­ing on the sec­ond in my series of draw­ings of San Fran­cis­co land­marks (if I’ve start­ed the sec­ond it’s offi­cial­ly a series, right?) I’m in the ear­ly pen­cil phase, doing a lot of study of ref­er­ence pho­tos and cre­at­ing a lot of out­lines, not attend­ing to ques­tions about texture.

One of the fas­ci­nat­ing things about the human mind is its capac­i­ty to cre­ate con­nec­tions from total­ly dis­parate fields, and how these con­nec­tions can bridge the con­cep­tu­al voids that oth­er­wise would have been missed. 

A sim­ple exam­ple can be found in almost every episode of the pop­u­lar tele­vi­sion pro­gram House MD. Just as the first few min­utes for­mu­laical­ly teas­es the view­er to guess which ail­ment will be Dr House­’s case for the week, there are a num­ber of sim­i­lar plot devices which recur every episode. One such device is the epiphany trig­gered by an unre­lat­ed per­son­al con­ver­sa­tion. «You’re lying to keep from doing rounds in the clin­ic!» some­one says to House, which trig­gers a flash of insight about a virus dis­guis­ing itself as a bac­te­r­i­al infec­tion to move from one organ to anoth­er where the med­i­cine can’t get to it. Dr House, being a bril­liant and trou­bled, dys­func­tion­al mind, is there­fore well-suit­ed to diag­nos­ing very com­pli­cat­ed sys­temic fail­ures. It’s almost like the sto­ries of under­cov­er police who solve crimes by learn­ing to think like crim­i­nals. House detects com­pli­cat­ed ill­ness­es because he is a bril­liant and sick per­son himself.

Okay, enough pop-psy­chob­a­b­ble. The point is that this char­ac­ter invari­ably has to step away from the prob­lem in order to iden­ti­fy a pat­tern in anoth­er con­text. Fre­quent­ly my con­scious mind, when look­ing at a prob­lem, actu­al­ly pre­vents me from mak­ing these con­nec­tions, because I focus so much on the prob­lem itself and can­not let my implic­it knowl­edge fill in a pic­ture. It’s sort of a for­est-for-the-trees kind of prob­lem, but I think it goes far­ther than that. It’s more as though I can­not see a par­tic­u­lar tree until I look away and start to recall details I had­n’t con­sid­ered on first obser­va­tion, like the tree-like shape of the mys­tery objec­t’s shadow.

So it is not unusu­al at all that I’ll have the best insights about one kind of work while I’m doing some total­ly oth­er sort of work. Today it was pen­cil draw­ing in my stu­dio, and as I was try­ing to cre­ate rep­re­sen­ta­tions in two dimen­sions of shapes exist­ing in three dimen­sions, part of my mind was toy­ing with the rela­tion­ships between colors. 

One of the recent chal­lenges in my dig­i­tal Lux­o­graph­ic process has been to break images down into col­or com­bi­na­tions that Pho­to­shop does­n’t under­stand. There is no mag­i­cal col­or sep­a­ra­tion process to break an image down to non-pri­ma­ry col­ors, with the excep­tion of the tra­di­tion­al cyan, magen­ta, yel­low com­bi­na­tion. Even those should be thought of as pri­ma­ry col­ors; they are sub­trac­tive pri­ma­ry col­ors rather than the addi­tive pri­maries of red, green and blue.

My most recent pieces have employed a not-too-clever cheat: I’ve start­ed with an image with a sin­gle dom­i­nant col­or, tak­en two chan­nels and assigned their infor­ma­tion to col­ors I guessed would be a good match for the orig­i­nal, then adjust­ed the val­ues of those «spot» col­or chan­nels to cor­rect for the col­or-bal­ance shifts that come from los­ing one chan­nel and recom­bin­ing dif­fer­ent col­ors. This tech­nique has been suc­cess­ful, but it has tak­en quite a bit of tri­al and error and I’ve start­ed with images with very lim­it­ed col­or palettes to begin with.

The ques­tion was danc­ing around the edges of my con­scious­ness as I was putting pen­cil lines down on paper, and then the thought came sud­den­ly to front and cen­ter: how did they do it before com­put­ers?

It’s a sort of obvi­ous ques­tion for some­one attempt­ing to bring the human hand into the process to ask, but it’s some­thing I haven’t real­ly inves­ti­gat­ed in a long time. When I did, it was a ques­tion of most­ly idle curios­i­ty. Some of my assump­tions about the process of cre­at­ing line screens with­out com­put­ers went into my dig­i­tal recre­ation of the process, but that was lim­it­ed to the realm of cre­at­ing the halftone screens. How did col­ors get separated?

I had a crude guess that it was accom­plished by repho­tograph­ing pho­tographs mul­ti­ple times with fil­ters to exclude every­thing except the com­pli­ment of the col­or of the fil­ter. I went to my copy of the «Pock­et Pal» hand­book for graph­ic arts pro­duc­tion. Mine is a fif­teenth edi­tion, pub­lished in 1992, the year I first start­ed to work in pre­press. Dig­i­tal pre­press was still very new at that time and the bulk of the «Pock­et Pal» is devot­ed to tra­di­tion­al processes.

I’ve looked at the descrip­tion before, but today the pieces fell togeth­er. And yes, they did it just like I thought. They made four pho­tographs of a col­or pho­to­graph, one with a fil­ter for each of the addi­tive pri­maries and a fourth to cre­ate a mas­ter for the black ink.

The thing that makes this sig­nif­i­cant for me is that I can do that dig­i­tal­ly with­out rein­vent­ing any wheels. The col­or sep­a­ra­tion can be done with lay­ers in Pho­to­shop with­out much dif­fi­cul­ty, and it does not have to be done at obscene­ly high res­o­lu­tion, as does the halfton­ing of the images. While I’ll still have to do some tri­al and error to find the cor­rect mix of col­ors for the palette I want, I won’t have to start from a rep­re­sen­ta­tion of the wrong col­or and adjust. I can cre­ate chan­nels for the spe­cif­ic col­ors I want — for any spe­cif­ic col­or I want.

It seems sort of obvi­ous now that I’ve writ­ten it out, but for prac­ti­cal pur­pos­es I’m bet­ter off with a course of action that seems obvi­ous than one that seems arcane and obscure.