It is the most important consideration in preparing a grey scale image for offset lithography. If you are not going to print your image on a press and it is only for display on your computer monitor (or desktop printer) then you do not need to worry about dot gain. However, printing your image on an offset press is an entirely different matter.
When ink hits paper the ink is absorbed by the paper fibers and, as a result of the capillary action of the fibers, the ink spreads out slightly. This is dot gain. Essentially the dot on the paper is bigger than the dot on the printing plate. The more fibrous and rough the surface of the paper, the more dot gain you will get. With a smooth coated paper dot gain is minimized, but is still a factor.
Dot gain makes your printed image darker — simply because the ink spreads and there is more black ink used in the image.
In the past, when everyone sent their original glossy photographs to the printer for processing (making films and velox or blueline proofs) the printer took care of adjusting the exposure of the images to compensate for dot gain. With some consideration of the type of paper stock and the type of press the image would be printed on, he adjusted the exposure. Similarly, if the printer scans your images and you will be replacing low resolution images with high resolution images held by the printer (through OPI), then your printer will adjust the scans to compensate for dot gain (usually)!
Also note that many programs, Photoshop and PageMaker included, have a general setting to compensate for dot gain. I never use these as I believe that for the best results each image should be adjusted individually.
Since many of us now scan and process our own images, we need to compensate for dot gain ourselves, prior to sending the completed PageMaker or Quark files with the accompanying Tiffs or EPS files.
Okay, your monitor gamma has been adjusted and you tweaked the heck out of your image with the controls in your twain module. Now you have what you think is a perfect image displayed in Photoshop. Right? Wrong! Dot gain is going to get you. To compensate you will need to make your image look worse, a little ligher, a little washed out on screen. Remember, dot gain makes the image darker. Therefore you have to make it a little lighter, but not necessarily uniformly so. Simply adjusting the "brightness" although helping the situation somewhat, will not necessarily result in the best image with the shadow and highlight detail you want reproduced. So here’s what you do.
In Photoshop first make sure that the Info palette is displayed (go to Window\Palettes\Show Info). This palettes is going to be used to provide you with information about the levels of grey at in your image. For example, pick the eyedropper tool and run it over parts of your image. Notice how the "K" value (K = black) varies from the light to dark parts of your image.
As a general rule your goal is going to be to get the whitest value to read 3% and the darkest value to read about 90% to 95%. (Note that this is a general rule only. This can be used faithfully for printing 150 line screen images on glossy coated paper):
"Image" on the menu bar
Click "Adjust"
Click "Curves"
A graph appears with a line drawn diagonally through it. Below the graph is a bar that shows gradations of grey tones from 100% white (or 0 value) to 100% black (or 255 value). Below the bar, on the right side, are two values "Input" and "Output." As an example of how input and output work, take your mouse and position it over the exact center of the grid. Input will read 50% and output will read 50%, now click and drag the diagonal line downward. Input will stay at 50% (if your hand is steady enough to stay on the center line) and the output value will change. As you drag notice that your image changes.
Place the mouse in the center of the graph and drag until the Input of 50% produces an Output of 40%. Then release the mouse.
Now position your mouse inside the image you are working with and try to find the darkest part of the image. Watch the "K" value at the top of the Info palette (when in an image the Info palette is where the density values will be show). If the darkest part is 100%, then go to the graph and drag the top of the line downward until an Input of 100% yields an Output of 90%. However, if the darkest part of your image reads 95% then pull the Input of 100% to an Output of 95%. Check the image and see whether the darkest part of your image reads 90% (it should).
Now position your mouse inside the image and find the lightest point. If the Input reads 0% in any part of your image then this portion of the image will not contain any halftone dots when it is output on the imagesetter and therefore it will be the color of the paper. If this occurs in a "specular" highlight (like the sun glare off a metal or glass surface) then this is the result you may be looking for. However, if this 0% white area is at the edge of your image, then the absence of halftone dots will make the image loose its straight edge and the image will appear to be poorly printed (again, unless this is the effect you were after, like when you are producing a gradation that ends at 0% density).
To prevent the absence of dots, the lightest area of your image will need to read 3%. So if the lightest area is 0% drag the Input of 0% until it shows an Output of 3%. (But if your lightest area is 1% you will only need to drag the 0% Output to 2% etc.)
Your image is now close to being optimized for dot gain.
Input 0% Output 3%
Input 50% Output 40%
Input 100% Output 90%
Now, before you accept the changes in this Curve review your image and check that the detail you want to be brought out is evident. Remembering the caveats above (especially regarding the lightest point) fine tune the midpoint and dark point, or even add adjustment points in between.
Incidentally, generally speaking, the output value of the midpoint can fall almost anywhere from 3% to 90%. Its value really just depends on the effect you are trying to achieve and the amount of detail you want in your image. Adjusting the Input of 50% to Output of 40% is a general guide for compensating for dot gain while retaining as much tonal variation in your photograph as possible. As I mentioned at the start of this discussion each image requires its own custom adjustment. You will find that in some images the midpoint should not be moved at all.
Similarly, there are images where only minute portions (a few pixels here an there) read as an Input of 100%. Therefore, caution should be used when following these general rules. If large segments of the image give Output values of 95% to 100% then these areas of the image will totally block up with ink when printed and generally produce a poor image. However, if only minute areas of the image demonstrate tones in this range and the vast majority of the image’s tones are below 95%, then the best image may be yielded by leaving the Input value of 100% alone so that it still yields an Output value of 100%.
Now send your image to the printer and have them output film and make a contact print. Check the film with a loupe to see if all the dots are reproduced and that there are now totally opaque areas or solid black areas. Check the contact print for these same items and review the overall tonal variation and detail of the image.
Print the image on an offset press. When you receive the printed piece compare it to the version displayed on your computer screen in Photoshop. Get a grey scale color chart and with the image open in Photoshop and a printed copy of the image in front of you see if the 90% as measured in Photoshop is printed at 90%.
It is important to realize that the above guidelines are for images that contain the full tonal range from total white to total black. Not all black and white images do. For example, a black and white photograph of the clouds in the sky on a sunny day will more than likely not have any grey in the 70% to 100% range. Therefore, to a very large extent you need to use your artistic judgment when adjusting the curves of an image.
Additionally, results will vary. No matter how consistent you are you will find that in a one print job your 90% black was right on the money. But on the next project (produced on the same equipment, with the same paper, at the same printer) your 90% printed lighter or darker. Why? While your images are being processed on the same imagesetter and the films are the same quality (let’s assume that), the step from film to plates to ink may not be. Possible variance is introduced:
So what is the bottom line to producing images in halftone images in offset lithography? It’s that the tonal variation in the image will be compressed. There is no way around this, it is the nature of printing a halftone that some of the image information is lost and the range of tones is crunched to a shorter scale through the elimination of the lightest highlights and darkest shadows. A black and white photograph reproduced on photographic paper has the advantage of using the white of the paper as the lightest color and the deepest, darkest black that the silver crystals can produce as the darkest shade of black. Not so in offset lithography.