The magazine of the Melbourne PC User Group Why Are My Beautiful Scans Terrible When Offset Printed Stephen Davey

Many budding DIY publishers are disappointed when they get the final results of their first commercially printed project. The colour scans that looked wonderful on the computer screen and inkjet proofs, have turned out either dark and muddy, washed out, or worse yet... only in black and white. Why would this happen? 

To a certain extent it may be the quality of your scanner. There has to be a good reason why commercial publishers will pay $40,000 for a scanner when there are plenty around at prices as low as $100. However, it may be more a case of what you do with the scan (either deliberately or inadvertently) between the actual scanning process and placing the image in your page layout program ready for output to film and/or printing plates. 

Virtually all inexpensive to mid-range scanners scan into an RGB colour space. This means that for each scanned pixel they measure and record the amount of Red, Green and Blue. Generally they look good on your computer monitor because your monitor (and television screens) display everything you see using those same three colours, RGB. 

The problems arise when we wish to reproduce an RGB scanned image in a commercially printed colour publication. Although there are some exceptions, most colour printed items (such as this magazine) are produced by printing combination of four standard process inks - Cyan, Magenta, Yellow and blacK (referred to as CMYK). So we need to convert our RGB image into a CMYK image before the printer can create the CMYK film and/or printing plates required for the printing press. 

If a page layout file such as those created by PageMaker, InDesign or Quark contains CMYK images, the software can automatically create the required CMYK colour separations. However, if an RGB image is accidentally placed into one of these files, then the RGB image will only reproduce on the black (K) separation and not on the others. The resulting image, that was meant to be printed in full colour, would be produced in Black. 

To convert an RGB image to CMYK, most people simply use the built-in facility in an image editing program such as Adobe PhotoShop, Corel PHOTO-PAINT etc. and this is fine as long as the default settings within the program are suitable for your project. 

CMYK Images Are Not the Same as RGB 

Unfortunately a good looking RGB image will not necessarily convert to a good looking CMYK image. This is because the gamut (number of colours) that can be produced by an RGB image is quite a bit larger than the CMYK gamut. Some bright and vibrant colours in an RGB image will appear dull and dreary in the CMYK alternative. 

All CMYK Images Are Not Born The Same 

Halftone dots printed on rough and absorbent newsprint will bleed or swell and become larger than the image dot on the printing plate. This change in halftone dot size is called dot gain. Images printed on high quality gloss coated paper suffer much less dot gain than those printed on newsprint, and images printed on sheet fed offset presses have less dot gain than images printed on web offset presses. 

If we know that our image dots will gain size, then we can make the original image dots smaller, to compensate for this on press gain. Fortunately from many years of printing experience, the industry knows what happens to halftone dots on various papers printed through various printing processes and there are several standard compensation settings that can be applied to images. 

Converting RGB to CMYK 

Before converting an RGB image to CMYK in any image editing program, make sure you have carefully checked the conversion settings. To reproduce the same RGB image on different printing process, you will need to create different CMYK images using different colour compensation settings. 

In PhotoShop the settings are reached from File -> Colour Settings -> CMYK Setup. 

For the moment we will deal with just the built-in conversion facilities. ICC and table conversion can be the topic of another entire article. 

The colour settings include three main areas - Ink setting, dot gain and colour separation options. 

Ink Settings 

The adjacent images were all created from the same RGB image, with different CMYK settings. The differences on this page may be subtle as they have all been printed on a reasonable quality coated stock. The bottom images are lighter than the top images because allowance was made for expected dot gain (which in this case did not occur because we are not printing on newsprint). However, if we printed the bottom two images on newsprint, with all else being equal, they would appear as dark as the top images. 

It should be noted that dot gain affects not only colour images but also black and white images (ie. grayscale) and flat areas of colour made up of percentages (stipples) of various colours. 

A grayscale image that prints well on glossy coated stock will be very dark when printed on newsprint because the halftone dots swell too much. A good looking newsprint image will be faint and washed out if printed on glossy paper because the dots will not gain as much size as expected. 

The flat area of colour (shown above) is made up of 80% yellow and 40% magenta. It would be much darker on newsprint because with dot gain the 80% is likely to fill in to 90% or even 100% and the 40% would increase to 50 - 55%. 

Separation Options
 

To compensate for this deficiency in the color separation process, printers remove some of the cyan, magenta, and yellow in areas where the three colours exist in equal amounts, and they add black ink.In the extreme case - 100% Cyan + 100% Magenta + 100% Yellow in theory makes 100% black (and takes a 300% ink coverage to do it), while 100% black makes a much cleaner black and requires only 100% (1/3rd as much) ink coverage. 

A given color can be translated from RGB mode to CMYK mode in an endless number of ways. But prepress operators typically use one of two ways to generate black in print: undercolour removal (UCR) or gray component replacement (GCR): 

With UCR, black ink is used to replace cyan, magenta, and yellow ink in neutral areas only (that is, areas with equal amounts of cyan, magenta, and yellow). This results in less ink and greater depth in shadows. Generally UCR is used for newsprint and uncoated stock, because it uses less ink. 

With GCR, black ink is used to replace portions of cyan, magenta, and yellow ink in coloured areas as well as in neutral areas. GCR separations tend to reproduce dark, saturated colours somewhat better than do UCR separations, and better maintain gray balance on the press. 

For the moment we'll deal just with the GCR settings. You can choose from several Black Generation settings in the Separation Setup (see below). 

It is important to realise that any changes made to your CMYK settings don't affect an image until you next use those settings to convert an RGB image to CMYK. Existing CMYK images (previously converted) will not be changed by altering these settings in your conversion software. 

Becoming a colour separation expert will not happen overnight, but the most important step is to remember NOT to accept the default CMYK conversion settings, unless you have checked with your printing company that they suit the particular job you are printing, and unfortunately, different jobs may require converting a new set of differently configured CMYK images - so do not discard your original RGB scans.

Reprinted from the July 2001 issue of PC Update, the magazine of Melbourne PC User Group, Australia

[About Melbourne PC User Group]