OvationPro2.75(08-November-05)djpu*z  xd6HSVnddB  L L   @0 8  tn L  GenericBlack!fWhitefTransparent fRedfGreenfBluefCyan!fMagenta!fYellow !fRegistration Af `.HSVnff. @0_a;F`U. ^Trinity.Medium.ItalicTrinity.MediumTrinity.Bold `,Bodytext#`;.. "40P33`* Header 1)^66)LA0P33L3 Header 2 (^..(#70P33+ Top Header `;PFPF03tO0P33?  h$ Beg88`;.. "40P33`* eg8^..(#70P33+ eg8 `;.. "40P33`* eg8 ^..(#70P33+ eg8& a.."60P33+ eg8 a.."60P33+ NV ^66)LA0P33L3 NV^66)LA0P33L3 Va.."60P33+ V`;.. "40P33`* d0$`;.. "40P33`* V ^..(#70P33+ eg8"`;.. "40P33`* eg8#`;.. "40P33`* eg8K$`;.. "40P33`* eg8%a.."60P33+ eg84a.."60P33+ V5^..(#70P33+ V7^66)LA0P33L3 NV8^66)LA0P33L3 eg89a.."60P33+ W;`;.. "40P33`* -%$$=^PFPF4S0P33A NV>^66)LA0P33L3 eg8 ?`;.. "40P33`* eg8@a.."60P33+ eg8Aa.."60P33+ z =T!1&G?5x$Y Y >5ȮT!FÓ479 ">*@"Y Y_4oA55$ >5ȮT!1&Z'FÓ479 ">*@"Y Y_455$ NV`VhWpd0$x-%$$eg8 4LL  !MainDict !IgnoreX  {filename} P   {filename} P{pagenumber} {datetime}-iKZ  L A4 4tt L    nn7. nn7. p2`yy! ^\ccMR r" X]7 D@ PP Z4 H$H  ! l"  L A4JL t L   nn7. nn7. p2`yy! ^\ccMR r" X]7! @x@ P ZTh4 H@4"TramTMM7!7!oBJ,117!VZ $$ Hnn76 On`Td. z_ N: 0lcIB{AE  t  X]7W# o3Zpo *>*g*g*g*;$This document contains a brief outline of a colour management system and denitions of the more common terms associated with it. Those looking for more authoritative documentation can nd it at the International Color Consortium's website, currently at http://www.color.org$.  >Why Colour Manage?7 $If accurate colour processing is to be done there must be some method to match the way different devices see colours. A scanner may see red better than blue so an image coming from this scanner would appear abnormally red on a monitor which wasn't able to compensate. If the printer's green ink was a brighter than its blue then this would contribute to more colour inaccuracies. Whilst it is quite possible to balance these factors out and make good prints on a single system it is a difcult calibration process to carry out and one which needs constant supervision. 9Colour Management$ provides a method by which this calibration process can be largely ignored by the user whilst also providing the opportunity to move images between different computer systems with a reasonable expectation that the image can be rendered the same on each.  $The conversions outlined are carried out by a 9Colour Management System $(CMS). This is a set of routines made available to an image processing application. Examples of CMSs are the Windows ICM, Adobe's ACE and LittleCMS, an open source CMS. $The CMS maintains an image within the 9Prole Connection Space$ (PCS). This is an internal device-independent space which can represent all colours. The CMS also handles the movement of an image into the PCS by using$ 9input9 proles$ to convert an image's colours to the chosen $colour space$ on introduction, and 9output profiles$ to convert the image from the chosen 9colour9 9space$ $into the$ gamut$ of the output device, normally screen or printer.$  $A 9device input prole$ is not absolutely necessary to produce a satisfying image. However, it is possible for an input device to be able to reproduce colours which fall outside the boundaries of the common 9colour spaces$. $This is especially true of top-end digital cameras.$ A 9device input prole$ will control how these colours are treated. Many input devices$ do not have 9device input proles$.$ As i$$t is necessary for an image to be assigned a specic$ 4colour prole$$, if an image does not have one, one should be allocated prior to any image manipulation. Many digital cameras will assign a9 9colour space "name$, normally sRGB, to an image in the EXIF data. This is not a prole and there is no mapping of colours but the information allows an image manipulation application to assign the correct prole on loading.$ $A 9device output prole$ is absolutely necessary for each output device to allow the correct rendering of an image $in a colour managed environment$. Without them an image can not be viewed correctly on-screen nor can it be correctly printed.$  Colour Models. Colours are often described in a relatively abstract mathematical way . Normally this involves three (in the case of RGB, for instance) or four (in the case of CMYK) values or colour components. In the RGB  Colour Model one can describe a 3-D shape with (say) Red on the x-axis, Green on the y-axis and Blue on the z-axis. The 3-D shape so dened contains all theoretical colours. Examples of  colour models are RGB, CMYK and LAB.  Colour Spaces. Although all the colours in such a 3-D shape are theoretically possible, in practice this is not so. Hence it is necessary to apply a mapping function to the colour model to dene a subset of colours which are reproducible. This subset of colours is called a  colour space and it occupies a volume, and is completely contained, within our original 3-D shape. Examples of  colour spaces within the RGB  colour mode l are sRGB, AdobeRGB and ProPhotoRGB. In the CMYK  colour model, a  colour space may refer to any of a large number of printer's ink sets, examples are Euroscale, US Sheetfed and Japan Coloured.  Color space is thus a term for a certain combination of a  color model plus a mapping function. The term Acolor spacAe is often incorrectly used to also identify  color models since identifying a color space automatically identifies the associated  color model. Thus although several  color spaces are based on the RGB  colour model, it would be wrong to refer to the RGB  color spa c e.  Gamut The range of colours occupied by a  colour space within a  colour model is called the space's  Gamut. Any colour which dos not fall within this volume is referred to as  Out-of-Gamut.  Colour Proles A  Colour Prole is a piece of computer code which contains a mapping function. It can stand alone or be contained within an image le. There are a number of different types of prole but the two most common are: Document Proles This is often a small generic piece of code which denes the  colour space in which the colours in an image reside. Several such proles are normally available on a computer which is capable of colour management. When there is need to embed a prole into an image and no specic AInput Device Profile is available then a document prole should be embedded instead.   5Device Proles $A %device prole$ provides the data required to map $the" range of colours$ the device is capable of rendering, to a %colour space$. An input device, such as a scanner, 'sees' colours quite differently to a digital camera so its %Input Device Prole$ must take that into account. An output device, such as a monitor or printer, must have an %output device prole$ to allow the mapping of the image colours to the% gamut$ of the device.$ Such% proles$ are normally provided by the device manufacturer.   Rendering Intent $When a colour managing image processing application sends an image to the monitor it does so using the monitor %output% prole$ which relates the colours in the %colour spac$e$ assigned to the image to the capabilities of the monitor. Normally this works well because the monitor's% gamut$ is relatively wide. When the same image is sent to a printer this is not often the case. The inks used in the printer work in a way which reduces the saturation available as the image darkens and a printer has problems rendering a true black. In other words its% gamut$ is narrow. $In consequence the printer's %device prole$ must convert any of the image's colours which the printer$$ cannot render into colours which it can. This conversion can be handled in a number of ways according to the %Rendering Intent$. Relative Colorimetric ?One such method might be to simply clip all the @out-of-gamut? colours to the nearest available colour while leaving all the @in-gamut ?colours alone. This is the @Relative Colorimetric rendering intent?. The image's white point is also remapped to coincide with the white point in the new@ gamut?. Perceptual ?Conversely the@ Perceptual rendering intent? converts an image by maintaining the relationship between all the colours in an image whilst pulling the@ out-of-gamut? colours into@ gamut?. Thus all the colours in the image are altered but the relationship between them is preserved. Absolute Colorimetric $The %Absolute Colorimetric rendering intent% $is very much like %relative colorimetric$ except for the handling of the white point. The% relative$ method remaps the white point from its position in the the original% gamut$ to the white point of the new% gamut$ whereas the% absolute$ method does not. The% absolute$ method can thus introduce colour shifts.  Saturation $The% %Saturation rendering inteAnAt$ aims to maintain saturation during conversions. It is more useful when moving from a more restricted% gamut$ to a wider one. Accurate colour relationships are not maintained during conversion.  The latter two intents are of little interest to image manipulators. Which of the former intents is chosen is largely down to the individual but many would recommend the use of  relative colorimetric rst.   Choosing a colour space The format of the image may determine what colour space is appropriate. If an image is for publication it is likely to be CMYK. Thus the colour space and the prole will be dictated by the requirements of the printing method, notably the inkset used. When using RGB images the choice of  colour space is largely personal$ as the differences between many RGB Acolour spaces$ is not great. A  colour space encompassing the colours present in theA gamuts of the devices used would be perfect but is often not possible so it is necessary to choose that which gives the nearest t. This is difcult to test in practice and many photographers have concluded, through experience, that AdobeRGB is a better choice than sRGB as the loss of saturation is more than compensated for by the increased range of colours available. sRGB The most common space is sRGB which was primarily designed to show well on a CRT monitor. It is the default space for viewing images on the web and, if that is the intended target for an image, then the choice of  colour space $should be sRGB. AdobeRGB The gamut of the sRGB space is quite limited and many devices can capture or display a wider range of colours than sRGB allows. AdobeRGB was dened to give a wider gamut but does suffer a drop in saturation level. It is more suited to colour printing and is currently the most popular choice for many photographers. ProPhotoRGB A newer  colour space, called #ProPhotoRGB, is gaining in popularity in some circles. It has a much wider gamut than even AdobeRGB but suffers few of the saturation problems normally associated with  wide gamut spaces. ProPhotoRGB encompasses better the wide range of colours detectable by modern high-end digital cameras but no output device is currently able to match this. Development in this area is rapid and it is likely that this will change but it may be a space to avoid until this occurs. web a ` .ʁz4\; . z4\$ .Jz4t$ .*z4$ .bz4$ . z4$ .JӁz4 $ . z7 .BzLA>l .jz4$ .z4$ .z4$ .*.z4$ .jfz40$ .z4$ .փz4x$ .*z4d$ .jGz4$ .z4$ .귄z4$ .z69 .j(z4$ .`z4$ .ꘅz4%$ .*хz4PF$ .j z4\$ .Az4 .Byz6$ .z6,9 .jz4ܼ$ ."z4$ .Zz4H$ .*z4|,$ .jˇz4 .z6$ .B;z6t9 .*tz4$ .jz4h$ .z6ԝ$ .z4W$ .Tz6$ .Œz6$< .ʼnz4$ .z4 .5z6$ .jnz4$ .z4ȯ$ .ފz4ܕ$ .z6P9 .Nz60$ .z4l.$|g7-"z6$g7- z4  .*Gz6l .jz6\ .z6, .z68Q .(z4 ._zLA 8 . z6ԑ .Jځz6  .2z4t .Jz6L .z4 .ֺzLAa \ .z6 .5z4 .nz4 .z6D .ރz4DF .z4 .zMz7ppd .Rz4ĕ .꾄z6 .*z6( .0z4 .Rhz4  .Ʂz6 .؅z41 .jz6\ .RIz4 .z7'5\ .*z6$( .z4" .)z6$ .bz4l$ .*z6$ .Ӈz4$ . z6($ .Bz6d$  .*{z6>% .z4T$L .Rz4 .v#zLA T .2fz4+$ .rz4$ . ։z6$ .Jz6$ .2Gz4L$ .~z6$ .z4$ .z40$ .2(z4$ .r`z4$ . z6 c$h7/z6 {$h7/Gz4d$h7/z4hF$h7/Rz4}$h7/z6Ё$h7/(z4h7/z_z7txh7/z6?h7/Ёz6@h7/* z6P?h7/jAz6@h7/Rzz4̙?h7/걂z6/?h7/z4h7/z!z7\h7/Zz6?h7/z4`?h7/*˃z6?h7/jz6(@h7/R