Diving Deep in Image Management

Our workflow begins with client supplied images and proceeds along separate parallel tracks : one for CMYK and one for RGB. For the sake of argument, let’s assume you have several unaltered digital photos to manage. And, just to narrow the focus a little more, let’s confine the purpose of this entry to print production, not web or internet. Therefore, the beginning image management section of the RGB workflow is the appropriate place to start.

As you may have noticed, most RGB images have an embedded profile of sRGB IEC61966-2.1. Unfortunately, this is not a trustworthy production color space. It’s fine for web, but if you’re going to design for print you should convert all your images to Adobe RGB (1998). Why? Some RIPs have an out-dated version of this profile and will image poorly if you print directly from sRGB IEC61966-2.1. Furthermore, many higher-end digital cameras have a color space that is described as sRGB IEC61966-2.1 (it’s what shows up in the profile information in Photoshop) but is actually larger and closer too Adobe RGB (1998). Predictable color is dependent on having a common reference for conversion and output. Some service providers recommend the profile sRGB, so far, we have more predictable and reliable results with Adobe RGB (1998).

To convert your images in Photoshop, go to the top menu bar and select Edit > Convert to Profile. Make sure to check the “Use Black Point Compensation” and “Use Dither” boxes in the “Conversion Options” of the dialog box. This will mitigate any posterizing of your image. And finally, don’t hit “Okay” without clicking the preview button a few times. There is one caveat concerning “Black Point Compensation”; if your image has little to no shadow content you might want to uncheck this option. That’s what the Preview option is for. Preview is your chance to see if you are going to have any issues with your color conversion. Finally, save your converted files as new Photoshop files. You never know when you might want to go back to your original customer provided files. Sometimes the client needs them because they’ve lost track of their original images. Occasionally, there’s a concern over the appearance of the printed piece and your client will want to examine your work against what was provided. CYA, conserve your assets.

You can use Photoshop to convert all your files as a batch action. Photoshop batch processing, actions and droplets can save you a lot of time. When we employ these features, we always open the new files and compare the appearances to the original files. A blind process does not always yield the best results. We’ll talk more about the dangers of blind processing when we cover PDF creation.

Now that all your images are in a common color profile perform your edits (clone brush retouch, color range adjustments, addition of other elements, sizing and cropping). You’ll want to save these as multilayer Photoshop files. Do your alterations in a non-destructive manner with layer copies, masks and adjustment layers. You never know when you’ll have to go back and tweak something for a current or future project. Once all edits have been finished, save copies of your files as flattened TIF’s. DO NOT save over or discard your multilayer Photoshop files. That is the digital equivalent of painting yourself into a corner. Work that way and one of these days you’ll be sorry.

You may have heard about a loss-less version of jpeg called JPEG 2000. The problem with that format is not every RIP can process it. It’s no fun getting a call late at night from a disgruntled printer and having to perform an emergency fix. Don’t gamble if you can help it.

Now that all your images have been edited and saved as flattened tiffs, it’s time to pick which side of the street you’re going to drive on. If you’re dealing with art that must be printed on a conventional 4 to 6 color press it’s time to change over to the CMYK workflow. And we hope you have properly calibrated your monitor. The shift to CMYK will need to be managed. Save your CMYK art as a new file, don’t overwrite your RGB file. You may need it at some future date. BPI always adds the color space and profile to our file names (for example, FileName_SWOPv2, FileName_Gracol06, FileName_PrinterNameG7, etc…).

Now, you’re probably wondering what an appropriate CMYK profile is. In the United States, US Web Coated SWOP v2 is almost a default. However, a number of printers and large format providers have moved to Coated Gracol 2006. Some print providers have a custom G7 profile and can provide you with one to install in Photoshop. It’s always a good idea to call and ask.

Finally, let’s spend a little time talking about color conversion. Whether your converting from RGB to CMYK or CMYK to RGB you should treat conversion as a one-way street. Please take another look at the workflow diagram. Notice the arrow from RGB to CMYK. RGB is larger in color space then any CMYK profile in existence. BPI usually, about 90% of the time, has to perform color adjustments on a CMYK image to match (as close as possible) the appearance of the RGB counterpart. And, we do our CMYK color adjustments in a non-destructive manner with adjustment layers and masks. Once those edits are complete, we save a flattened tiff copy of the CMYK multi-layer Photoshop file.

You are now ready to construct your art.

Why What you See Is Not Always What You Get

Color management is probably one of the most confusing and difficult challenges facing the design staff of any firm.  Frankly, its one of those things that rarely gets adequately explained.  We think of it like roller skating.  It looks simple.  But when you strap the skates on for the first time and start barreling down the steepest hill you can find, things are about to get a little out of control.

So, let’s eliminate some of the mystery.  Just keep in mind, in many ways this will be like learning to skate or ride a bicycle.  So, we’ll start simple and build up speed gradually.

Way back in 1942, a gentleman by the name of Richard Hunter came up with a 3-dimensional coordinate system Called LAB.  LAB was a fairly successful attempt to describe and quantify the way the human eye perceives color.  The common model for LAB is represented by a sphere with the axis: L, a and b. The L axis has a range of 0 to 100 and defines the amount of white or black (depending on your point of view).  The a axis ranges (in Photoshop) from -128 to +127.  Positive numbers describe the amount of red and negative numbers describe the amount of green.  The b axis ranges (in Photoshop) from -128 to +127.  Positive numbers describe the amount of yellow and negative colors describe the amount of blue.

So what does this have to do with the real world?  To put it simply, LAB or more accurately CIE/LAB is the universal language for all print devices.  No matter what you build your file in (Photoshop, Illustrator, Corel Draw or Powerpoint) at some point all the digital color information is going to be converted to CIE/LAB for printing.

Now here’s one of the places color gets misleading.  Your computer monitor displays in a different color space called sRGB (to be precise your monitor probably displays in sRGB IEC61966-2.1).  This monitor profile was created by HP and Microsoft for color monitors.  The idea was to establish a global base line for display of color images over the web.  Color on your monitor is produced by exciting three different types of photosphere channels: red green and blue.  If all three channels are maxed out you get white light.  Conversely, if all three channels are completely suppressed you get black.

But wait there’s more!  Your computer display is governed by something called a CLUT. That’s an acronym for “Color Look Up Table”.  If your system is new, everything should be fine since all of the drivers are current and your hardware is new and running off of what amounts to factory presets.  However if your monitor has been around a few months, it might be a good idea to update your profile.  Unless you are doing critical Photoshop retouch, you can probably manage by using the operating system to update your display.  It’s also a good idea to keep your monitor updated because that also allows your system to keep the CLUT current.

Just for the sake of argument.  Let’s say you’re just working with a small desktop printer on your home system.  You build your art and your system displays the color values in sRGB IEC blah, blah, blah.  That display is referenced against the CLUT by your operating system.  You like what you see and you decide it’s time to hit print.  Now the hidden world of mad calculations kicks into gear.  Your operating system, the printer drivers, your graphics card and your application all interact to create and send a print file to your printer.  They work together to calculate the CIE/LAB equivalents of your RGB colors and then tell the printer what to print in the printers own language.  Now you look at your printed piece and most of the art looks right.  But some of the colors don’t appear quite the way you wanted.  What happened?

What has likely happened is some of your art has been created with colors that are outside of the gamut (the range of colors within the ability of a device to reproduce) of your printer.  Remember when we described the Lab color model as a sphere?  If the abilities of your printer were measured against Lab space it’s shape might better resemble an extra-large, jelly donut.  Lab (or at least the amount of Lab space available to your system and the software you’re employing) is an NBA basketball in comparison.  In essence, the creation of a printer file is an attempt to take that NBA basketball and stick it in the extra large jelly donut.  It is almost inevitable that something will get lost in the translation.

Is this the whole picture?  Afraid not.  Come back for the next installment and will take a more in-depth look at print file creation.