Skip to Content

Soft Proof

Translations and com­pro­mises in im­age mak­ing or; the Image Cult Society.

There’s an in­ter­est­ing thing that hap­pens when a new idea or tech­nol­ogy gets in­tro­duced then quickly as­sim­i­lated into the back­ground hum of our daily lives. It starts out with a dis­creet name — a clear iden­ti­fier of what this thing is and means. Than this name just sort of … slips away. It be­comes so nor­mal that to name it would seem strange. Its orig­i­nal name does­n’t seem to fit any more, as the name ex­isted in the first place to de­mar­cate the new thought from the or­di­nary. And now the new thing is just or­di­nary. Think about Google Maps. It’s just … a map. In 2005, when Google Maps was first re­leased, it’s par­tic­u­lar ap­proach to the in­ter­face of a dig­i­tal map was called a slippy map’. Weird, right?

This is an in­ter­est­ing phe­nom­e­non around cul­tural ap­proaches to tech­nol­ogy, but not ac­tu­ally what I want to talk about. I want to talk about soft proofs. Soft proofs are an ex­am­ple of this taken to an ex­treme — you use them every day but you have prob­a­bly never heard of them. There is no need for the soft proof to be some­thing other than nor­mal, the soft proof just is nor­mal. But what is a soft proof, and why is it so nor­mal? And why do I want to ex­plore a topic so quo­tid­ian that the word used to mark it as in­ter­est­ing is so faded and worn?

A soft proof is a way of view­ing an im­age be­fore the im­age has been re­pro­duced me­chan­i­cally. In con­trast to the soft proof is the hard proof: a way of view­ing an im­age im­me­di­ately af­ter it’s been re­pro­duced me­chan­i­cally. Basically, a soft proof is an im­age on a screen that will be sent to a printer. Otherwise known as an im­age. It’s need for a dis­creet name seems so un­nec­es­sary that it seems bizarre to re­fer to all im­ages - even this text as I write it — as soft proofs. But that is, in essence, what they are. We see im­ages on our screens that an be re­li­ably turned into im­ages on other peo­ples screens, and even into phys­i­cal im­ages on pa­per.

The rea­son why this needed a name to de­mar­cate it as spe­cial — dur­ing the ad­vent of the dig­i­tal — is that this is a re­ally hard prob­lem to solve. There are a range of math­e­mat­i­cal mod­els for ap­proach­ing a rel­a­tively uni­fied the­ory of color and vi­sion, and a wide range of phys­i­cal pieces of ma­chin­ery that are tasked with pro­duc­ing those im­ages — from print­ing presses to mon­i­tors. The act of en­sur­ing an im­age can be pre­dictably re­pro­duced is nec­es­sar­ily an act of trans­la­tion. Translating from this color space to that; from an ad­di­tive color model of a screen to the sub­trac­tive color model of ink and pa­per; ap­prox­i­mat­ing the color of a pa­per stock to be printed on.

This trans­lat­ing process is done us­ing some­thing called a Color Profile. A Color Profile is a set of rules for en­sur­ing that an im­age cre­ated with red, green, and blue light can be repli­cated on off-white pa­per us­ing cyan, ma­genta, yel­low, or­ange, and green inks. The cur­rent work­flow of dig­i­tal to print is so smooth, so ubiq­ui­tous and mun­dane, as to oc­clude the mas­sive tech­no­log­i­cal feat that it sup­ports it.

This feat was un­der­taken by a small group of tech­nol­ogy com­pa­nies I the early 90s, and they col­lab­o­rated to de­fine a uni­ver­sal stan­dard of how this would work.

The International Color Consortium was formed in 1993 by eight in­dus­try ven­dors in or­der to cre­ate an open, ven­dor-neu­tral color man­age­ment sys­tem which would func­tion trans­par­ently across all op­er­at­ing sys­tems and soft­ware pack­ages. . . . The eight found­ing mem­bers of the ICC were Adobe, Agfa, Apple, Kodak, Microsoft, Silicon Graphics, Sun Microsystems, and Taligent.

- Color Science History and the ICC Profile Specifications, Elle Stone

The cur­rent, base­line pro­files built around RGB and CMYK came about with the rise of dig­i­tal im­age-mak­ing, which is the ba­sis of the the cur­rent world around us, a world built on and pred­i­cated by im­ages.

The dom­i­nant trans­la­tion is dom­i­nant be­cause it — to a large de­gree — works. Creating color pro­file is re­ally hard, mathy, ph­sysicsy stuff. It’s hard to do your­self. But CMYK/RGB cuts the cor­ners off the world to make it fit into a gamut that can be han­dled. But by ne­ces­sity it’s a com­pro­mise: what col­ors are we not han­dling in or­der to han­dle the max­i­mum num­ber of col­ors? What parts of the color space get left be­hind?

Many of these is­sues give me the feel­ing at times of re­luc­tant rather than open co-op­er­a­tion be­tween some of the com­pa­nies that cre­ated this stan­dard. Having said that, there does seem enough in­for­ma­tion in the pub­lic stan­dard (when com­bined with ex­am­in­ing avail­able ex­ist­ing pro­files) to ef­fec­tively and ac­cu­rately char­ac­ter­ize color pro­files of de­vices and color spaces. I could imag­ine there be­ing some poor re­sults at times though, due to some loose­ness in the spec.

— What’s wrong with the ICC pro­file for­mat any­way?, Graeme Gill

Its im­por­tant to un­der­stand this com­pro­mise; un­der­stand how it works and what ex­change we’re mak­ing in the process. What are we giv­ing up, and what are we get­ting in re­turn?

What are we leav­ing on the table? For ex­am­ple, pho­tog­ra­phy (up un­til the 80′s) cal­i­brated for white peo­ple. African Americans and other dark skinned peo­ple pho­tographed poorly. They were out­side the color space. The story goes that school pho­tos of in­ter­ra­cial class­rooms would have rows of per­fectly ex­posed white kids, and voids where the black chil­dren should have been (Adamn Broomberg). More than this, the stan­dards only changed with in­dus­try pres­sure from choco­late and fur­ni­ture man­u­fac­tur­ers — a realm of cap­i­tal where the browns and blacks mat­ter (Rosie Cima).

Film chem­istry, photo lab pro­ce­dures, video screen colour bal­anc­ing prac­tices, and dig­i­tal cam­eras in gen­eral were orig­i­nally de­vel­oped with a global as­sump­tion of Whiteness.’”

— Looking at Shirley, the Ultimate Norm: Colour Balance, Image Technologies, and Cognitive Equity, Lorna Roth

With the cre­ation of the RGB color space, with the cre­ation of the ICC, we ceded the vi­sual world to su­per­mas­sive tech in­ter­ests, much like we’ve ceded our pri­vacy and per­sonal data. In do­ing so we’ve in­her­ently made the cre­ation and dis­sem­i­na­tion of im­ages into a tool for cap­i­tal — one that sup­ports dom­i­nant power struc­tures.

How do we un­der­stand the im­plicit, in­vis­i­ble, baked in as­sump­tions of the soft proof? We can start by op­er­at­ing out­side the pa­ra­me­ters of the soft proof, rec­og­nize it as a tool to use or not use. The gap be­tween the soft proof & the hard copy is mea­sured in the gap be­tween the tools used to plan & pre­pare ver­sus the tools used to pro­duce, and we can move in to this gap and in­habit it. We can cre­ate work here, and in do­ing so re­claim some of the space that we’ve given away.

The Risograph, for ex­am­ple, has a tool­chain for soft proof­ing, but the ma­chine— through its high speed & low cost — also opens up the pos­si­bil­ity of de­sign­ing im­ages through it­er­a­tive hard proofs; blend­ing the tech­niques go the mod­ern dig­i­tal print process with the clas­si­cal ana­log ones.

The web is a strange medium — a blend­ing of soft and hard spaces. A plas­tique space, with plas­tic proofs and plas­tic copies. The same process of trans­la­tion is at work — be­tween the still & the in­ter­ac­tive, flow­ing im­age. This is why show­ing comps & wire­frames of web­sites to clients can be so tricky: our cul­ture of im­age cult and tech­no­log­i­cal process can elide the crit­i­cal dif­fer­ences we sense as agents of these sys­tems.

This is a call for a Marxism of im­age mak­ing — to seize the means of pro­duc­tion. To cre­ate rad­i­cal im­ages & tools that ex­ist in the cor­ners of the gamut and color spaces dis­carded by the soft proof. To un­der­stand that plan­ning is not do­ing, and take con­trol of our own vi­sual lan­guages.