Photoshop Lab Color- P10 ppt

Notice first that below the color wheel onthe left, the program permits us to express these colors in xyYor even LUV, if we’d like to avoid LAB.. On the other, one color expert offers the

shot from a product of GretagMacbeth, a

leading vendor of color measurement

instru-mentation It has measured 50L(30)A40B

where the desired value is 50L(40)A30B, and

it wants to know how far different these two

greens really are from one another

Notice first that below the color wheel onthe left, the program permits us to express

these colors in xyYor even LUV, if we’d like

to avoid LAB But that’s a side issue The big

action is underneath the green circle, where

we find one large and three small numbers

that are introduced by the lowercase letter d.

It stands for the Greek delta, which

mathe-maticians often use to denote the quantity

of change The asterisks remind us that the

snooty name is L*a*b* In comparing the

two colors, the program notes that the L

values are identical, so dL*=0 da* and db*

are both 10

To Each His Dulcinea

“A knight errant who loses his lady,” Don

Quijote remarked to Sancho, “is like one who

loses the eyes that let him see, or the sun that

shines on him, or the food that maintains

him I have told you this many times before,

and now I say it again: a knight errant

with-out a lady is like a tree withwith-out leaves,bricks without mortar, and a shadowwithout the body that cast it.”

Right above the three readings we’vejust been discussing in Figure 13.3, muchlarger type informs us that dE=14.1 That’sthe key, the iconic number Pronounce it

“Delta-E,” and prostrate yourself before

it if you believe that machines see colorbetter than you do

Delta-Eis an attempt to quantify thethree readings on the line below it tocreate one comprehensive number thatdescribes how far the two colors are fromone another There are several formulas

to produce it; this one involves weightingthe Achannel more than the B, a lot ofsquare roots, and other fancy stuff Butwhatever the formula, a dE of zero is ideal;

the higher it is the worse the match; and theobjective of a good interchange between twodevices is to have the average dE be as close

to zero as possible

Opinions on the merit of dE are, shall wesay, varied On the one hand, certain cali-brationists are so smitten that they make aDulcinea out of dE: they believe that sheembodies perfection and it is only a matter oftime before we find, attain, and adore her

On the other, one color expert offers thefollowing succinct opinion: “The originalconclusion therefore remains, the CIELABisperceptually extremely non-linear and thedE* unit of color-difference is totally worth-less for all processes and uses in photo-graphic imaging.”

Personally, I am somewhere in the middle,but closer to the second view than the first

The real Dulcinea, like dE, was no thing ofbeauty She herded pigs, a dirty and unpleas-ant task that nevertheless needs to be done

dE is like that, too Machines can aid incalibration, and a formula like dE, howeverimperfect, is the only way they can come to adecision We simply have to understand that

Figure 13.4 The middle row holds the reference colors

Are the colors at the top closer to the middle ones than the

bottom ones are? A machine would think they are each

equally far off.

sometimes it gives misleading results, and

sometimes we deliberately want to disregard

it because we get a closer visual match with a

worse literal match

Let’s assume that we are trying to make

output from a desktop printer match that of a

commercial print shop so that we can make

inexpensive proofs at home A conventional

idea is to have the print shop produce a

variety of patches, which are then measured

spectrophotometrically The desktop printer

spits out the same swatches, which are

mea-sured and compared to the press’s

Adjust-ments are then made to try to get dE lower

by forcing the desktop printer to get closer

to the print shop’s results

There’s nothing inherently wrong with

such an approach, but the two greens that

we’ve been discussing—50L(30)A40Bon thepress and 50L(40)A30Bwhen we try it on ourprinter—demonstrate the deficiencies of anysuch formula, in at least three different ways

First, humans expect greens to be greenerthan a machine does If the desktop printerproduces a better-looking green than thepress does, we may declare that the twomatch, even though a spectrophotometerfinds a big dE Also, the machine thinks thatmisses of equal magnitude make for equal

dE Not so If one of the misses is greenerthan the desired value, and the other is lessgreen by an equal amount, humans willinvariably consider the greener one a closermatch to the original

Second, how far one color channel appears

to be off from the desired value depends onwhat goes on in the other, a fact lost on aspectrophotometer If the desktop printer’spatch had measured at 50L(50)A40B, the ma-chine would compute the same dE, becausethere is still a difference of 0L10A10Bbetweenthat and what’s wanted However, we wouldperceive that color as being closer to theoriginal than 50L(30)A40Bwas, because, whileboth the Aand Bwere each inaccurate by 10points, both were further away from 0A0B

than the desired value was Were they thesame 10 points off, but one channel movedtoward neutrality and the other away from

it, a spectrophotometer sees less differencethan we do

If you don’t believe it, check out Figure13.4 The “desired” colors are in the middlerow In the top row, bothAand Bwere moved

15 points further from zero In the bottomrow, one channel went 15 points in one direc-tion and one 15 points in the other Therefore,the top and bottom rows both have the same

dE with respect to the middle (None of theoriginal LABsquares were out of the CMYKgamut, so the conversion for printing didn’taffect the relation of these colors.) Humanswould see the bottom row as being much

Further Reading on Colorspaces

For those interested in more detailed discussion and

comparison of the colorspaces described in this

chapter, the enclosed CDincludes six papers by Prof

Gernot Hoffmann, a color expert at the University of

Applied Sciences in Emden, Germany They include a

general introduction to graphics for color science; one

paper each on the structure of XYZandLAB; a study

of the question of gamut generally; and one that is

specific to the question of CMYKgamut Finally, there’s

one paper that has nothing to do with the topic of this

chapter but does bear on Chapters 5 and 11, namely a

discussion of how gamma can create certain problems

when working in RGB

I suggest starting here because Gernot does a good job

of reducing technical concepts into somewhat

compre-hensible language There is a fair amount of

mathe-matics involved, but no calculus If you don’t

remember what went on in your 12th-grade algebra

course, however, you may find it heavy sledding

Thanks to Google and its ilk, much more information,

including the actual formulas for converting between

the spaces, is available by typing in search criteria such

asCIELUV,CIExyY, and so on

TheCIEis alive, well, and living in Austria Its URLis

www.cie.co.at

further from the middle than the top is,

partic-ularly in the context of a real image

And third, the age of a spectrophotometerdoesn’t affect its color perception The age of

a human does The measured green is more

yellow than the desired one A younger person

is likely to find this fact more objectionable

than an older one would As we age, our

corneas become yellower, lessening our

sensi-tivity to that color If you’re over 40, you

defi-nitely are seeing less difference between these

two greens than you would have in your youth

Matching Unmatchable Pantone Colors

If it sounds like the machine may be right and

humanity wrong, ask yourself who decides

whether the match is a good one If you want

your monitor, or your desktop printer, to predict

what output will look like on some other device,

are you going to accept something that a

machine says matches, when your own eyes tell

you it doesn’t?

Furthermore, many conversions are betterdone by ignoring numerical matching al-

together The prime example, and one that

features LABin a big way, is the handling of

Pantone Matching System colors

Authentic PMScolors are created by mixingspecial inks Such custom-mixed inks can

achieve certain colors, particularly pastels and

blues, that aren’t otherwise available on any

current output device

Unfortunately, printing with an extra ink isexpensive Clients often request that the PMS

color be emulated with standard inks, no

com-bination of which can match it This traditional

problem has a relatively modern solution

The traditional workaround was a set of

Figure 13.5 Three attempts to emulate PMS colors in

CMYK Top, Photoshop 7 and later versions use

Pantone-supplied LAB values, converted to CMYK here by the

default settings of Photoshop 6 and later Middle, the

same LAB values converted to CMYK using Photoshop 5’s

default Bottom, CMYK values were inserted directly

using Pantone-supplied tables (Photoshop 6 and earlier)

CMYK“equivalents” that Pantone issued for

each of its custom inks Most graphic arts

applications contained a library of these

“equivalents”; some still do

By the turn of the century, the problems

with this one-size-fits-all approach had

become apparent The same CMYKvalues

produce different results in different settings

A newspaper, for example, would get darker,

muddier colors than this book, which is

printed on much higher-quality paper

Fur-thermore, desktop printers that didn’t want

aCMYK file at all, but rather an RGBone,

were starting to make a dent in the market

Pantone responded by issuing real

equiva-lents to its inks—using LAB They were

intro-duced in Photoshop 7, in 2002 In previous

versions, specifying a PMS color in

Photo-shop’s Color Picker got us the

Pantone-supplied CMYKvalue, plus aLABvalue of

dubious origin Since then, we get only the

new Pantone LABvalue AllCMYKand RGB

values are computed from it, using whatever

our current color settings are

Today’s method is an improvement, but

no magic elixir When a color can’t be

matched, it can’t be matched

Take a look at emulations of nine PMS

col-ors, all of which are out of the CMYKgamut

Figure 13.5A uses the post-2002 LABvalues,

converted to CMYKusing the default

separa-tion setting of Photoshop 6 and later: the U.S

Web Coated (SWOP) v.2 profile, which was

derived from machine measurements of

actual printed samples Figure 13.5B uses

the same values, but the default separation

method of Photoshop 5, which was put

to-gether by human observation and tweaking

And Figure 13.5C was never separated at

all The CMYKvalues were inserted directly,

using the Pantone-supplied numbers of

Photoshop 6 and earlier

Unless you have access to a Pantone

swatch book, you won’t know which methods

worked best I’ve got one, and in my opinion,

Figure 13.5C is non-competitive in all ninecolors The current separation method, Fig-ure 13.5A, has a better match in five of thenine colors, but a serious problem, too

In their custom-ink incarnations, the threereddish colors are all more intense than any-thing you see here The pink is lighter andpurer The color that prints as orange in theupper right is more like an angry pink, andthe red at bottom center should simply beredder I rate Figure 13.5A as the closestmatch in all three

It also wins in two of the three greens, butnot the one at middle right Photoshop 5 did abetter job, because the machine-generatedmethod fell victim to the same problem thatmade all its blues purple, and made thisgreen too yellow

Ah, that problem with the blues If a PMSblue is a key color, converting with the v.2SWOPprofile is likely to get the job rejected

The blues—and that’s what they originallyare, blues, not purples—are a disaster inFigure 13.5A In the center swatch (PMS2728), the cyan ink is 16 points higher thanthe magenta in Figure 13.5A, but 30 pointshigher in Figure 13.5B

The atrocity occurred, I suspect, whensoftware decided that Figure 13.5A’s “blue”

had a lower dE with respect to the Pantoneoriginal than Figure 13.5B’s does And whynot? Any sensible dE algorithm is going togive much greater weight to fidelity to the Avalue than that of the B The desired numbersare 33L20A(69)B That (69)B is out of thequestion in CMYK

Photoshop claims that 33L20A(53)B isachievable A machine will think that that’sthe best we can do, since it tries to get as close

as it can to all three LABnumbers Loweringthe 20Awould, in its small mind, increase dEand therefore be undesirable

Human graphic artists instinctively knowthat the A must nevertheless be lowered,because blues that are overly cyan are apt to

be much more acceptable to a client

than blues that are too purple So dE goes

out the window, and the artist is true to

the spirit of the original blue, rather than

to its numbers

Maintaining the Distinction

The problems of converting out of LAB

offer instructive points about converting

out of anything else

LAB’s problems are more severe cause it is capable of constructing colors

be-that are wildly out of the gamut of the next

space However, the solutions are entirely

applicable to less onerous conversions

RGBto CMYKis the case that most people

think of, but it isn’t the only one Some

original LABvalues of Figure 13.5 were out

of the RGBgamut as well as the CMYK

Consider what happens when we proach the edge of the gamut Yellows are

ap-a strong point for CMYKand a weak one

for RGB, so they’re a good example A

value of 94L0A90Bis fairly extreme It

con-verts to 0C4M81Y, or 255R235G21B (The

LAB“yellow” is actually slightly orange.)

Raise the original value to 91B, and onepoint of yellow is added in CMYK The RGB

value also rids itself of some of the

conta-minating blue, dropping to 255R235G10B

Raise it to 92B, and CMYKadds anotherpoint of yellow The RGBhits 255R235G0B

It can’t get any less blue than that, so when

we raise the stakes to 93B,RGBhas no way to

call the bet

As we continue to increase the B,CMYKcontinues to add yellow, for as long as it can

By the time we get to 94L0A99B, it’s 0C4M100Y,

and now there’s no more yellow ink to add

The RGB “equivalent,” meanwhile, is still

dis-denotes an extremely intense yellow To saveroom for 37 even more vivid flavors of yellowwould be outlandish Every yellow in theimage would have to be drastically toneddown so that these hypothetical brighteryellows could be distinguished from them

Figure 13.6 All colors in the LAB original that produced these images were within the CMYK gamut The top version was sepa- rated using the default settings of Photoshop CS 2, which employ Relative Colorimetric rendering intent The bottom image’s colors are more muted because Perceptual intent, which was previously the default, was used to separate into CMYK

A

B

But what if a picture comes along that

somehow requires that such distinctions

be maintained? Couldn’t such a picture

exist, and if so, how can we possibly handle

it if all the yellows smoosh together during

conversion?

If such pictures exist, they’re rare And if

the distinctions would be obliterated during

the conversion, then if we need them we have

to act while still in LAB

And that, it turns out, is the generalized

solution to how to treat out-of-gamut colors

during any kind of conversion Namely, forget

them Just match everything else, and let the

weird colors worry about themselves Unless,

of course, distinguishing the weird colors

from the rest is a priority Then, attack them

before making the conversion

This principle may seem obvious when theoffending colors are so clearly out of gamut asthese are, but it didn’t seem that way in thelate 1990s, when new capabilities were beingengineered into Photoshop’s separation algo-rithm The theory then was that colors thatwere barely in gamut should be intentionallytoned down, so that any out-of-gamut inter-lopers would seem brilliant by comparison

This was called perceptual rendering, and

effective with Photoshop 6, it became thedefault way of doing things

In 2005, the error was corrected in shop CS2 Figure 13.6A converts the LABfileusing today’s defaults These colors are allfairly bright but all were originally within the CMYKgamut Therefore, the perceptualmethod used in Figure 13.6B toned them

Photo-down, thinking to accommodateany brighter colors that mightshow up

Rendering intent is set in ColorSettings and can be overridden

in Edit: Convert to Profile Thecurrent default, Relative Colori-metric, takes the simple view thatall matchable colors should bematched and whatever happens tounmatchable ones is our problem

(An alternative, Absolute metric, should be avoided ManyRGBs have “white points” that aretheoretically not white in CMYK

Colori-RelCol remaps them to 0C0M0Y;AbsCol may turn them blue.)The perceptual rendering in-tent, in any case, is too mild to be

of any real use It also is able for conversions into RGB Inthe previous example, it wouldhave increased the yellowness

unavail-Figure 13.7 This image was prepared

for prominent use in an advertising campaign.

more slowly, maxing out at 104B rather

than the 99Bof RelCol When we deliberately

tone down areas that we could have matched

if we had wanted to, usually we want to tone

them down a lot more than the perceptual

intent does

LABis the universal interchange standard

We should use it to try to match what we

can, in most cases There are also times when

we can use L A Bas an insurance policy

against the possibility of a bad conversion

The Knight of the Unambiguous Transfer

“Señor,” inquired the goatherd, “who is thisman, who dresses in such a way and carries

on in such a fashion?”

“Who could it possibly be,” replied thebarber, “but the celebrated DonQuijote de la Mancha, cham-pion of the weak, redresser ofinjury, righter of wrongs, theshelter and refuge of damsels,the horror of giants and thevictor in battle?”

“That sounds to me,” musedthe goatherd, “like what youread in books about knights-errant, who do all the things thatyour grace is telling me this mandoes, but as far as I’m con-cerned either your grace is jok-ing, or this gentleman has holes

in every corner of his brain.”

While I was writing this bookdur ing a break in one of my

classes, one of the students, a

pro-fessional photographer, requested

help in my capacity as redresser of

injury and righter of wrongs, said

wrong being that his job was in

jeopardy He had just prepared Figure 13.7,full of rich browns and dark reds, for a veryprestigious placement for his most importantclient The printed results had been quiteunsatisfactory, he explained—all muddy and lifeless

I opened his RGBfile—noting an alert as

I did so—and said that it looked fine to me

“Wait until you see what the printer did toit,” he replied But before he could bring me

a printed sample, I said, “Let me guess It

looked a lot like this, right?” And I produced

Figure 13.8 on the screen Bingo

This sad story has been repeated hundreds

of times over the past several years, times with pictures as important as this one

some-is, sometimes not, always intensely ing, not just to the victims, but to all rational

frustrat-Figure 13.8 When the image appeared in

print, it looked like this, the result of

misinterpretation of the original RGB file.

observers of the graphic arts marketplace,

who are tired of watching it happen

The executive summary of what happened

is that the photographer handed off an RGB

file tagged as Adobe RGB The printer did not

honor it, and assumed that the file was sRGB,

wrecking the job I’m suggesting that

some-times the photographer should hand off an

LABfile, which cannot be misinterpreted

Inasmuch as repeated episodes like this

demonstrate that even many professionals

have difficulty grasping the topic, we will take

the scenic route through that last paragraph

In defining a colorspace, everything is a

matter of interpretation For example, when

LABneophytes first see an Lchannel in

isola-tion, they are often surprised that it appears

lighter than a grayscale conversion of the

document would be This happens because

the Lis interpreted as being darker than it

ap-pears, for purposes of making a screen

pre-view of the color image, or for converting into

CMYKor RGB You could create a different

kind of LABin which the Ldidn’t behave this

way, but you couldn’t use it in Photoshop

InRGB, though, we can use variants that

are darker than others, or more colorful

Beginning with Photoshop 5 in 1998, users

were encouraged to choose their own

defini-tions of RGB, rather than having a single

imposed standard Its supporters trumpeted

this concept as solving more of the world’s

problems than Don Quijote ever claimed

to even want to But there was one major

unforeseen drawback

Once we choose our own RGB, we needn’t

worry about the topic again—provided we

never send to or receive files from anyone

else We’ve told Photoshop what “RGB”

means, so our own work will be interpreted

correctly Somebody else’s work may not be, if

they’ve defined RGBdifferently and our copy

of Photoshop doesn’t know it

That’s why, before starting discussion of

numbers in Chapter 2, I told you what my

own color settings are If you didn’t changeyour own settings to match mine, thencertain of the numbers I talked about wouldvary somewhat in your system That would

not be a tragedy It would be a tragedy if you

took one of my RGBfiles and output it in aprofessional context without taking account

of how I had defined RGB Fortunately, there’slittle chance of that, since I would sooneropen a cage full of lions than I would put RGBfiles in the hands of strangers

The original concept was that each userembeds a tag into each RGBfile, identifyingwhat kind of RGB is in play The tag isrecognized by the next user’s system, the file

is handled properly, and the knight ridesRocinante off into the sunset

In practice, this works well among thosewho know what they’re doing Experiencedusers throw tagged RGBback and forth allthe time without a hitch Unfortunately, the world at large, and service providersespecially, are protagonists in a differentpicaresque novel Tagging an RGBfile andgiving it to a stranger on the assumption thatthe tag will be honored is a lot like walkinginto a busy intersection on the assumptionthat the traffic will stop, except the odds arenot nearly as good

To Run Where the Brave Dare Not Go

The question of which RGBto use is beyondthe scope of this book But how colorful thedefinition is makes a difference to the currentdiscussion The more colorful, the higher theprobability that we will run into gamut prob-lems such as the ones in Figure 13.5, wherethe RGB file calls for colors that can’t beachieved on output But if the definition isn’tcolorful enough, it may not contain colorsthat you might need It would be best if the RGB’s gamut matched the gamut of theoutput device exactly, but that can’t happenfor a variety of technical reasons

The most prominent of the less colorful

RGBs is sRGB, a definition promoted by

Microsoft and Hewlett-Packard in the late

1990s with considerable commercial success

For those who prefer a wider-gamut

color-space, the usual choice is Adobe RGB,

al-though some use even more colorful spaces

Adobe RGBhas become somewhat of a

stan-dard among professional photographers

Most consumer devices now prefer orrequire sRGB Most service providers now

assume that any kind of RGBfile coming in

is an sRGBone

A stalemate has been reached Adoption ofsRGBas a consumer standard has grown so

much since 1998 that there is now no chance

of dislodging it Furthermore, vendors are less

likely now even to consider the possibility

that an incoming RGBfile is something else

Yet a substantial group hates sRGBso much

that they will never agree to use it

Adobe RGBis quite workable, as is passing

an Adobe RGBfile to me or anyone else that

you know for a fact won’t misinterpret it

Throwing an Adobe RGBfile out to strangers

or to the world at large, tagged or not, is

ask-ing for Figure 13.8 to become your problem

instead of my student’s

For the record, let’s analyze how muchcolor got lost in this critical image I’ve mea-

sured four areas on the bearded elf,

fore-ground left, and will give equivalencies in

both LABand CMYK

• His purple cap measures 117R43G63B

When properly treated as Adobe RGB, it

us The misinterpretation has turned anattractive mix of subtle colors into somethingdishrag dull

Some people actually enjoy it when thishappens It gives them an opportunity tovent They can denounce all printers as stu-pid They can call up the CSRand scream

They can demand that the job be rerun andthreaten to sic their lawyer on them All thiscan be psychologically rewarding, and dis-tract attention from others of life’s worries

If you would just as soon have the job donecorrectly without all the histrionics, and youwould otherwise be sending out an AdobeRGBfile to a stranger, the options are:

• You can call up the people you expect

to handle your work and find out how theyintend to behave when confronted with atagged file Problem: you may not know whothey are

• You can convert your own file to sRGBfore sending it out This is a more attractiveoption than it used to be The chances ofsRGBbeing misinterpreted as something elsehave gotten less as sRGBhas become moreentrenched, but they still aren’t zero

be-• If the file is going to a commercial printer,you can convert it to CMYKyourself Thatwas why this particular photographer was

in my class He had, understandably, decidedthat Figure 13.7 was the very last RGBfile ofhis that any commercial printer was evergoing to encounter If you want to try some-thing similar, though, you’d best have confi-dence in your CMYKskills

• Finally, the one nearly foolproof method

There are many RGBs, but there’s only oneLAB Convert your file before handing it over,and it can’t be misinterpreted That’s whyPhotoshop uses it internally for most of itscomputations Getting anLABfile forces thenext person to convert to his own RGBor his

own CMYK, eliminating all ambiguity The

worst that can happen is that you get a call in

the middle of the night asking what to do

with this crazy messed-up file To that, I say

that anyone who can’t figure out what to do

with an LABfile certainly can’t be trusted to

figure out what to do with tagged RGB

Of Children and Colorspaces

This chapter opened with a problem that

could be seen as either one of Photoshop

technique or one of calibration The user had

two CMYKspot channels and couldn’t figure

out how to convert them into “real” channels,

something that Photoshop doesn’t make easy

Coming up with a solution requires a

knowledge both of Photoshop and of why

LABis used for interchange We first think

about how to accomplish the task while

staying in CMYK (it can be done, but it’s a

nuisance) Then, we consider whether LAB

can expedite the process, which it can

Hence, the answer: re-create the channels in

LAB, just the way Photoshop does when it

converts one colorspace to another

More such hybrid uses of LABwill nodoubt suggest themselves as the colorspacebecomes more mainstream Our next chap-ter, for example, is mostly RGB, but it’s heavilyLAB-flavored; the two exist side by side

Other image-processing products havetried to promote the use of LAB, and it isn’t arecent phenomenon Figure 13.9, a double-page advertisement for scanning software,appeared in 1996 The thrust is that LABissuperior to CMYKor RGB, which it is in a lot

of ways Note the claim on the right side:

“The human eye uses CIELAB.” There’s truth

in that, too

The company was so smitten by the LABconnection that it changed the name of itssoftware from LinoColor to VisuaLab And ifyou think that “Everything you need to knowabout color spaces” is a bit of overkill, youshould know that the children appear next

to the crayons because the theme of theadvertising campaign was “Color Is Child’sPlay.” Personally, I would like to lock theperson who thought that one up in a roomwith a copy of Chapters 14 and 15, and see if

Figure 13.9 This two-page advertisement, trumpeting the virtues of LAB , ran in 1996

he still felt that way after

reading them

Contrary to the tion of the advertising, all

implica-scanners operate in RGB

In fact, the software’s only

real connection with LAB

was that it converted its

raw scans internally to a

form of LABbefore saving

them, just as drum

scanners have alw ays con

-verted their RGB data to

CMYKbefore saving

Lino-Color and VisuaLab

soft-ware offered prescan color

correction, but in a relative of LAB, not the

real thing

LinoColor, which was transferred to the German printing giant Heidelberger

Druck maschinen when it purchased

Linotype-Hell, is no longer made But, as

it was once common and you may run

across some of its files, we should briefly

discuss two of its confusingly named

colorspaces

As we’ve just discussed in relation to

my student’s disaster, one great advantage

is that an LABfile is completely

unam-biguous; it is the same from one

Photo-shop user to another Once we leave

Photoshop, however, that is no longer true

There are at least half a dozen variant

versions of LABfloating around, one of

them in this software It’s called LAB(LH),

and is a smaller-gamut version of the LABwe

know and love Granted that Linotype-Hell’s

client base was heavily CMYK-oriented, the

decision made perfect sense Triple-digit

values in theAand Bare unprintable Making

an LAB that’s closer to the boundaries of

CMYKmeans that all objects fill longer ranges

in the Aand Band are thus easier to attack

with curves, Blending Options, and the like

LABis a relative of the colorspace known

as HSB(Hue, Saturation, Brightness; times known, with a slightly different third

some-or Lightness channel, as HSL) Like LAB,LUV,and xyY, color and contrast are kept in sepa-rate channels The color is defined differently,one channel defining the underlying hue andthe other its purity A piece of milk chocolate,

a brick, a human face, and a fire engine allhave about the same Hvalue, but each rates ahigher Sthan the one before it

Figure 13.10 Curvemeister, a Windows-only Photoshop plug-in, offers an LAB face for RGB files, including the ability to use LAB curves while in RGB

inter-LinoColor/VisuaLab’s version of this space

is called LCH(Luminance, Chroma, Hue)

There are no handling differences

To Dream the Impossible Dream

Getting back to the present, the most

inter-esting LAB-flavored idea is a Windows-only

Photoshop plug-in known as Curvemeister

The idea is to be able to write curves in

what-ever colorspace you like without necessarily

having the file in that colorspace

Figure 13.10 illustrates The file itself is in

RGB, but much information is available in

LABform The curves, neatly laid out, can be

applied as if the file were LAB,CMYK, or even

HSB We can even generate LAB-style

imagi-nary colors RGB is the strongest blending

space of the four, as we’re about to see in the

next chapter But it’s generally the weakest

for curves

Note that both of the products we’ve just

discussed have features that are regrettably

missing from Photoshop It would be nice if

there were an option for a narrower-gamutLAB, as in LinoColor; it would be very niceindeed if all curves could be displayedsimultaneously, as in Curvemeister

As uses for LABincrease, it’s likely that we’llsee improvements in how Photoshop hasbeen using it and the addition of capabilitiesthat will enable us to push the envelope evenfurther That is the practical side of LAB, theside that looks for results, not theory

As for the impractical, or quixotic side, theside of the academics and amateur theoreti-cians, the future is bright also, provided weunderstand that a hundred years from nowcolor scientists will think that we lived in

the Stone Age It is possible to program a

computer to analyze any image in the waythat a human does, and to figure out how ahuman would compute dE

Or, rather, it will be possible, because evenwith today’s colossal rate of improvement

in computing speeds, it will be a good 50years and probably more like a centurybefore a machine can calculate something

so complex on the fly

Like today’s color scientists, Don Quijotehad his heart in the right place; his error was

in so believing in his own infallibility that

he became ridiculous However, for all hismisadventures, he did do a great deal of goodfor some of the people he came across, and

he serves as an example to the rest of us ofthe power of trying to do the right thing

Before returning to matters of LABtechnique,it’s only fair to give him the last word:

“To expect the world to stay the same is awaste of the mental process Everything goes

in circles, I say, in circles; the spring yields tothe summer, the summer to the fall, the fall

to the winter, and the winter to the spring,and thus the endless circle continues; onlythe human life comes to an end, sooner thantime, without expecting any reprieve but inthe next world, which knows no limits.”

The Bottom Line

Photoshop permits an infinite number of definitions

ofRGBandCMYK, but only one of LAB Its status as

the one colorspace within Photoshop that is entirely

unambiguous gives LABa unique role in information

interchange Also, LABhas a stronger direct relation

with how humans perceive color than CMYKandRGB

do This offers important advantages in color

matching and in reconciling outputs from several

different devices

The extraordinarily large gamut causes certain

prob-lems when translating into or out of LAB Studying

how to do it properly offers lessons for other forms of

conversion, particularly the translation between RGB

andCMYK

The topic of this chapter does not mesh with the rest

of the book, which is less about theory and more

about technique Accordingly, it can be omitted by

those not interested in the subject

anguages, like color, like Photoshop itself, become mucheasier to learn once you’ve had even a little experience.

Certain concepts are impossibly difficult to fathom at first,but once you understand them, you find other situations towhich they apply As Chapter 13 indicated, LABis a languageall its own, and learning how to master it has a great deal incommon with learning how to master, say, the language that Don Quijotespoke to Sancho

Some linguistic changes make sense even when they are expressed in

an unfamiliar way I am the author, not the reader—or so the thought

is rendered in English Logically, though, the word I in the preceding sentence is redundant The word am can only apply to the person doing the speaking In many languages, Spanish being one, the I is optional.

Am the author is a complete sentence.

The idea that a channel doesn’t need to have color is no more difficult

to grasp than that a sentence doesn’t need to have a pronoun So, the Lchannel of LABis easy to understand, even for someone who’s never heard

of the colorspace before

On the other hand, certain concepts are very hard indeed When trying

to learn Spanish, I was confounded by what grammarians call the

reflex-ive verb The Spanish might say, for example, Me bought a book today.

As noted, the first word is optional But granted that they’re putting a

pronoun there, I could not understand why it should be me and not I.

The answer came not out of a textbook but from experience Spanishwas my first truly foreign language, but I happen to be fluent in certain

dialects that sound foreign to other English speakers I speak mid-century

Once for Color, Once for Contrast ,

Some LAB advantages are structural The A and B channels have

no counterparts in other colorspaces, so what they do is often impossible to duplicate But things that seem impossible outside

of LAB sometimes prove otherwise—provided you separate color from contrast in your mind.

14

rural Oklahoman; Standard Modern

Cana-dian; the weird noun-equals-verb jargon of

1970s American university students; and

the infamous New Jersey dialect If I were to

submit a manuscript in any one of these, the

publisher would take me for being,

respec-tively, illiterate, pretentious,

incomprehensi-ble, and obscene So, instead, I write in

Stan-dard American (more or less), permitting the

reader to take me for being all four at once

You Ain’t Seen Nothin’ Yet

How that Spanish sentence makes sense

dawned on me suddenly, when I translated it

into the only one of my dialects that regularly

employs reflexive verbs The technically

cor-rect translation of the sentence,

lexicogra-phers of Oklahoman would confirm, is Ah

bought me a book today, just as I might say,

Ah’m a-fixin’ to have me a good ol’ time in this

here chapter.

These sentences are considered

substan-dard usage in other parts of the United States,

meaning that a bunch of academics who

probably couldn’t put together a literate

sentence themselves have decided to brand

anyone who writes that way as a rube, a

phenomenon that has an analog in the

Photoshop world But in Spanish, the

struc-ture is quite correct And in Portuguese,

Italian, and French, too, which were of course

much easier to adjust to once I figured out

that Spanish and Oklahoman are similar

Returning to color, the grammar of LABis

also useful even when speaking a different

language Many LABtechniques work almost

unaltered in HSB, which is as close to LABas

Italian is to Spanish Between LABand both

RGBand CMYK, which are really two dialects

of the same language, not so much is

shared—except for the topic of this chapter

The thing that seems most foreign about

LABwas mentioned on the very first page of

this book It separates color from contrast If

you have mostly worked hitherto in RGBor

CMYK, you probably have rarely tried this

But you can, and it’s very powerful

Only those who are paid by the hour areinterested in spending more time than isnecessary correcting or retouching an image

If color and contrast need to be treated rately, there’s no problem in LAB, which treatsthem as distinct entities from the very begin-ning If we decide to separate color from con-trast in RGBor CMYK, we have to do at leastsome work twice and then merge it together

sepa-This can happen in several ways Suppose,for example, that you are impressed by some

of the color variation that LABcurves created

in one of the examples in Chapter 12, but youhave a hunch that you could have gottenbetter detail by doing the work in RGB

No problem—start with two fresh RGBcopies of the image Convert one to LABand

do what was done in Chapter 12 Return tothe RGBcopy and do it the way that you thinkcreates extra detail When finished, convert

it to LAB, and use its Lchannel to replace theone in the other file Presto: color done inLAB, detail created in RGB The power of thismethod derives from being able to work inRGBwhile paying absolutely no attention tocolor You can make your skintones greenand your skies orange if you like, or evenmake the entire RGBimage black and white

If the color is going to be replaced later, itdoesn’t matter what it is right now

We can even speak a variant of the LABlanguage without ever being in LABat all InRGB, we can create one version for color andone for contrast, and then paste one onto the other, creating a new layer If the versionwith better detail is on top, we set the layer toLuminosity mode; if the version with goodcolor is on top, Color mode

This chapter looks for places where RGBhas such an advantage that it would pay totake the extra time to do a contrast-only orcolor-only version there rather than use LAB

As this could be a book-length topic by itself,

we will be heavy on theory and light on

step-by-step examples We will divide the subject

into four sections: when to use RGB for

contrast in preference to LABusing standard

methods such as curves; when to use RGB

curves for color only; followed by two more

variants using channel blending, a topic

that we haven’t addressed so far but will be

featured in the next two chapters as well as

this one We won’t be considering CMYK,

be-cause for technical reasons that we need not

get into, RGBhas decided advantages over

its linguistic playmate for this kind of work

We begin by discussing not a word, English

or foreign, but rather a single letter—the S.

Detail, Range, and a Letterform

Most images have one or more areas of

par-ticular interest, areas that are so important

that we are willing to sacrifice elsewhere if

we can get more detail

The theory is simple, the execution plex Find the lightest and darkest points of

com-the object, and spread com-them apart Curves

are the usual method, but anything that

increases the tonal range will work In curve

language, we say, the steeper the curve, the

more the contrast Objects found in relatively

steep areas of curves gain contrast; those

found in relatively flat ones lose out

That’s a considerably more complex nition than the one most manufacturers and

defi-many retouchers use Their idea of “adding

contrast” is adding midtone contrast at the

expense of the lightest and darkest areas This

is what happens when you increase

“con-trast” on your monitor’s controls, or on your

television set, or when you use Photoshop’s

primitive Image: Adjustments>Brightness/

Contrast command

Sometimes this blunderbuss method ofadding contrast obliterates all variation in

extreme lights and darks; more sophisticated

variants simply reduce it in the interest of

promoting more range in the midtones

When expressed as a curve, the top andbottom sections are relatively flat and theinterior part is steeper It resembles the letter

S , and is commonly known as an S curve.

S curves are not appropriate for everyimage, only those where we wish to addcontrast to the midrange I’ve illustrated thisconcept elsewhere with pictures of white,black, and gray cats Apply an Scurve to awhite or black cat, and it yowls Only a graycat will purr when we stroke it with an S

To add contrast to a white cat, we hold thelightest point constant (since we can’t make

it lighter than the white it already is) anddarken its darkest point, which is usuallyaround a quarter of the way up the curve, inthe area that retouchers, logically, call the

quartertone The procedure is reversed with

a black cat The darkest point is held, and

the three-quartertone point gets lightened.

We can’t control what kind of cat we’regiven, but ef ah had mah druthers (that’sOklahoman), ah druther that it was a graycat, because we’re more likely to get awaywith drastic moves If the cat is white, thecurve we need to use darkens the entireimage, which may or may not be acceptable

If the cat is black, the whole image has to getlighter, ditto But if it’s gray, and we apply an

Scurve, the overall darkness of the imagewill seem about the same—it’s just that moredetail will go to the cat

S curves aren’t always the answer, butthey’re great things when they are

In LAB, the only way to add contrast is inthe Lchannel, so we can temporarily forgetabout what’s going on in the Aand B And, tosave you the trouble of reviewing 13 chapters

of examples, I will tell you that not only havethe large majority of curves we’ve applied tothe Lchannel been shaped like an S, but thepercentage is much higher than if we hadbeen working in RGB

Let me get to the bottom line first andexplain why afterward: if your work features