Newspapers & Technology spoke to Ken Elsman, chief color scientist for Global Graphics Software for a Q&A on converting RGB images to CMYK. Elsman can be reached at 781.392.1600 or via e-mail at kenneth_elsman@globalgraphics.com

There are many issues newspaper printing plants face when converting files from RGB to a CMYK format. Ken Elsman, chief color scientist at Global Graphic Software gave some advice on the issue as it relates to color quality management.

Elsman is the senior color scientist at Global Graphics. He has a masters degree in printing technology from the Rochester Institute of Technology. He also has worked as commercial photo lab manager, presentation graphics digital imaging center manager and prepress production manager for several travel industry magazines.

Elsman suggests that patience is necessary for anyone new to setting up color-managed workflows for digital proofing. He said that to build good profiles, newspapers must either hire or work with a color expert, and be prepared to a lot of testing.

What potential problems do newspaper face when images are converted from RGB to CMYK?

Severe gamut compression in some cases can introduce artifacts. Perceptual rendering is usually the approach to the RGB to CMYK conversion but it is a vendor specific “special sauce” and is usually driven by user preference, as some profiling packages perform better than others.

Proper selection of UCR requirements when CMYK profiles are made takes some expertise.

RGB working space must be part of the transform if any soft proofing or image editing is desired.

When converting RGB digital photographs to CMYK for process printing, are there any procedures to follow that give good results?

Define the workflow (where is the color managed conversion going to take place — Photoshop, page layout application, or in the raster image processor?) You must have appropriate profiles that define RGB color space for input and CMYK of press conditions used for printing for output You must select correct rendering intents to transform data, typically perceptual but there are exceptions

Do you lose any image quality when you convert from RGB to CMYK?

There are many ways in which to lose critical information due to gamut mismatches, but the eye is forgiving and therefore proper conversion can produce very good results. Issues include:

Changing the RGB workspace during image editing to a smaller gamut RGB space in Photoshop before conversion is potentially throwing away image data.

Profile Quality: Most digital cameras render images into some RGB colorspace (similar to film in relation to the original scene) that may or may not be ideal for reproduction. The most common is sRGB which is a somewhat gamut restrained colorspace but if the intention is to print on newsprint then the gamut is adequate. What is most critical is whether or not the CMYK profile really represents the press that is being used for printing. Achieving a press run that represents average printing conditions (within agreed upon process control aims) with color characterization targets on the sheets is a timely and costly endeavor but the only way to achieve good color characterization data for making press profiles.

Once converted to a press CMYK gamut, conversion to a different larger press gamut or, say, back to RGB for the web will likely diminish quality.

One profile for the conversion from RGB to CMYK might not work for all image types. High key images verses low key images often have different requirements. Thus some image editing of contrast and brightness may be required prior to the conversion.

Why do digital cameras transform their images to RGB instead of CMYK?

The charged couple device or complementary metal-oxide semi-conductor chips in cameras only capture light sensitivity. Filters are placed over these light capturing cells and RGB filter sets capture the scene similar to how the eye’s sensitivity works. Thus camera technology using RGB filters can capture large gamuts (most visible colors) and can easily be viewed on a monitor. CMYK conversion should be done based on the printing conditions down stream from image capture.

What do newspapers need to look out for when they are converting images to CMYK?

The CMYK profile must be made to represent the traditional settings used to set up the job for film or computer-to-plate for a given press, namely TAC (total ink) and UCR/GCR (gray component replacement) requirements. Typically the settings in a profile-making application to address these include TAC, black threshold, black generation, and UCR/GCR amounts. These variables work together and some understanding is needed when making the profile in order to produce the intended RGB-CMYK conversion. There is a fair amount of testing required during the profile-making process but once the CMYK profile is made, installed and working, the repeatability and overall quality is greatly improved.

Can you go into detail about the process color management companies developed to run newspaper stock on large-format inkjet printers and, along with proper color transforms, generating color-accurate proofs?

We first investigate the workflow — be it RGB to CMYK or CMYK to CMYK or RGB/CMYK to CMYK — and then build the color management solution around that.

If the workflow is strictly CMYK then, after a discussion on achieving processing aims, we work with the customer to add a color characterization target to a press run, take a sampling from that press run and build an input profile that captures the color capabilities of that press, substrate, and ink combination. We then build an output profile by running the same substrate through the inkjet printer, again building a profile that captures the color capabilities of the proofer, substrate and inkset. These profiles are then installed in the raster image processor, rendering intents are selected and the files are run producing very accurate color proofs.

For RGB workflows, the ability to produce color-accurate proofs requires an additional transform, which we call emulation. Emulation requires an additional profile, but this can be created from the same press data used to make an input profile. What this allows is the ability to achieve the best color possible on press but also make a proof that accurately predicts that color. What happens, in essence, is the RGB original image is perceptually rendered to the press gamut via an input RGB profile and a CMYK press emulation profile. Then we render this data to the proofer using colormetric rendering (match colors), which uses the CMYK emulation profile and the proofer’s CMYK output profile to achieve the color accurate proof. This process or technique from Global Graphics is unique in the industry as it can all happen in the RIP.

Finally, should you have a page with, say, RGB editorial images and CMYK advertisements with images embedded, you can still proof to an inkjet device running newsprint and achieve a color accurate proof by using input, emulation and output profiles.