Trimingham describes how to build incredible looking grayscale prints on dark garments
There are lots of ways to create separations for prints on dark shirts, but very few really make the graphic jump off of the garment with the least amount of work. The best prints on dark garments use the garment to a degree to darken the shadows in the image, and it always is a bonus if the printed area feels light to the touch.
Without printing discharge or water-based inks, the best way to create a softer feeling shirt is to change the printing order on some of the separations so that there is less ink going down and still a sense of brightness to the image. It is possible to achieve both of these objectives by using a style of printing and separation that many printers tend to avoid. This style is very useful particularly when printing and separating grayscale images on dark shirts because it can really help to create the deep shadows in an image with having the colors all be completely under printed with a white ink. The lower ink volume is what really gives the shirt its incredible look and feel.
The trick to building an unconventional grayscale print on a dark garment is to follow a simple three-step process. First, you extract the grayscale base part of the image. Second, you decide if you need a bump plate or color overlay. Finally, you pull out the highlighted white areas and review the work in a digital proof.
Before you start
The same process that is used in high-end, detailed separations in multiple colors when you begin the file preparations is also used for quick grayscale separations. It is critical to check the file to make certain it is the right size (actual size of the final output) and a decent resolution (at least 150-300 dpi) prior to starting the separation process. If you need to adjust other things about the file, it is always better to do so prior to separation—examples include typography, edge quality in the design, and the overall tonal range. Once everything is adjusted or clarified, then the separations can be started.
Probably the most common problem with submitted files is that they tend to be low in resolution and in edge clarity. The reason this is a concern for images that will be broken into halftones is that without a clear definition to the edges in an image, the ink tends to wander past where it should go and the whole graphic can begin to look fuzzy or indistinct. A lot of times this is something that won’t be caught until it is being printed as well, because the computer simulation will show a clear edge to graphics. This is because it doesn’t simulate the degrading of the halftones on the edges and the tendency for low end dots to become lost in the screen making process.
Remember, a 5% dot on the computer screen that creates an edge in a dark garment will likely not hold on the screen unless you use a very controlled process. Most printers are lucky to hold 10% dots in, especially with film. In the example image (Figure 1) the digital picture looks fine from a distance, but on closer inspection the edge quality is poor. The likely result of this is that the fur texture and eye would mash together in the final print and cause a very muddy appearance.
The solution for this issue is to carefully check the density of your edge defining areas in the image and to make adjustments for areas that are too low and will probably be squeezed out in the screen creation process. A filter in Photoshop can aid this process, so I took the wolf image and created a duplicate layer of the wolf, used the Curves menu to increase the contrast of the image slightly (deepened the blacks and brightened the whites), and then ran the Poster Edges filter on a light setting of 1 edge thickness, 0 edge intensity, and 6 on posterization.
The next step was to select just the fur areas of the wolf quickly using the lasso tool (this is called a pre-selection) and then use the Color Range tool and extract just the black areas (or the edges created by the poster edges effect). I could then create an extra layer on top of the original image and fill just these outlines with black and then edit them to create better edge quality where I needed it and erase extra shapes where I didn’t need them.
These steps create little outlines in the image that will aid in keeping the edge quality sharper and help to make the overall print lighter by allowing the black of the shirt to come through the image in those areas (Figure 2).
Now that the image had all the quality to a decent level, it was ready for separation for grayscale printing on dark shirts. If you were going to image on lighter shirts, you would have to reconstruct the background or go back and edit out areas to create a border for the design to blend properly.
Steps for separation on dark shirts
Extracting the grayscale underbase The first step of creating the grayscale underprint is deceptively easy when the image is created on a black background. I just created four extra alpha channels after the image channels and named the first one shirt channel and changed the Channel Options color to black. After this was completed, I copied the original file, pasted it into the next channel, and then inverted it and titled this channel my gray channel (Figure 3).
Initially, this pasted graphic will appear too heavy, but all it takes is a little bump of the contrast to make it right on target. The major concern is that the image may have quite a few areas that are very light in halftone density and may not hold onto the screen for this color. A simple solution for this is to ramp up the gray base to compensate for this possibility. The easiest way to do this is to use the Curves menu and push over the black point toward the middle of the graph while keeping it on the edge. You are just looking at increasing the lowest level of dot to at least 12% or so and then the heavier areas will need to be higher as well to compensate for the ink being gray instead of white.
Determining the need for a bump plate or overlay color A bump is most commonly used when the grayscale image has a strong color cast to it. For example, I added two quick color casts to the wolf image that required two color bumps (Figure 4). If there is not a color cast to it, then it really doesn’t apply. If there is a color cast to the image, then it is important at this point to create a color-overlay plate.
One of the best ways to do this is to find the most saturated area of the color and use either Color Range or pull a selection that you can convert into a channel using the Image Mode split method, where the image is split into a CMYK image from an RGB image and then the closest channel that has enough color information in it is used as a bump plate or color overlay on top of the gray color. Typically, the ink is reduced a lot so it stains the color underneath without creating too much opacity.
Creating the white plate As you can see for these styles of images, the gray screen becomes the underbase screen and the white can be printed on top of the gray (flashed, of course). The huge advantage to this is that both inks can then be run through much higher mesh counts than they would normally be. The gray will be brightened by the white on top of it, and the white will be underbased, thereby allowing the use of a reduced ink that can flow through a higher mesh count. The gray can be a 200-thread/in. mesh and the white can be a 190-thread/in. mesh using a 55-dpi dot.
These settings allow the highest quality print in this style with the least amount of ink buildup. To create the white, all you have to do at this point is duplicate the original image and then Curve out the lower end and save the highlights. Make sure to use a nice, smooth curve in the Curves menu so both colors blend well together (Figure 5). That way, the transition between the gray and white screens will appear very smooth and clean.
Practice with this style of separation can become second nature and extremely quick once you master the subtleties of it. Let’s look at a new image that will be created using the same method (Figure 6). The file had to be prepared properly, because again, in a close view, the edge quality was poor (Figure 7) and there were a few other issues. The leopard looked a little bored, so I copied the lower jaw, moved it up and closed its mouth and then duplicated the eyebrows and tilted them slightly in so the cat would look a little more intense.
The process of boosting the edge quality was almost the same as what I used for the wolf: I increased the contrast using Curves, then I sharpened the image using the Unsharp Mask dialog (amount 78 / radius 2.7 / threshold 1), and finally, I duplicated the layer and ran the Poster Edges filter (same settings) and then Color Range selected the black, created an extra layer, and filled in just the black and quickly edited it. This may seem like a lot of steps but it was only seven or eight minutes of work, which is a lot less than redrawing things or creating shapes or paths with the brush tool to define edges (Figure 8).
The separation followed the same method, except there were two overlay plates that were added to the grayscale image. These colors were reduced inks that were to be printed over the white and the gray at the end of the printing sequence. So the print was executed on press using this print order: gray, flash, white, flash, pale gold, gray blue.
The separations were produced in this order with first copying the image and placing it into the first alpha channel and then bumping it to make the gray underbase. This underbase was colored slightly to add to the color cast, so it wasn’t a pure gray, but more of a blue-gray underbase (Figure 9).
The next step was to again duplicate and then squeeze the original image using the Curves menu to produce the white screen, and this image was added to the next channel in the print order. The final colors were pulled using the CMYK method, where a duplicate of the image was split and then pieces of the resulting channels were merged back into the separations as a gray-blue and a pale-gold channel. This process was fully completed in less than 30 minutes, with the separations and final design all ready to print. The final product was bright, detailed, and still had a softer feel because it used less ink volume on the dark shirt (Figure 10).