Image Acquisition can involve scanning, digital photography, screen
capturing, etc. Despite following our Images
and Scanning guidelines, there are still many variables and problems
that can come up. Moiré patterns, inaccurate color, 'hot' spots,
banding, jagged edges, and many more. We'll describe some of these briefly,
with a short answer. For better technical support and specific solutions
for your job, please give us a call.
Moiré (Moire): This is a term used to describe wavelike patterns
in fabric, but applies to printing in some ways. Technically a "moiré"
is an interference pattern. It generally occurs when an image that you
are attempting to scan has already been printed. Printed images, as
opposed to photographic prints or original art, have been converted
to halftones. This means that they have been broken up into fine rows
of dots with overlapping colors to reproduce the image. If you examine
a printed piece closely, you will see small rosettes or rows of dots
that make up the image.
A scanner takes samples of an image at
regular intervals, those intervals are smaller and/or more frequent as
the scanner resolution is increased. When scanning a previously printed
image, the scanner 'hits' and 'misses' these previously printed dots,
and the result is an interference pattern of dark and light spots, or
wavy lines, or honeycomb patterns.
There is no practical way to scan a previously
printed image such that the scanner will 'hit' all the rows of dots
in the image, but there are a couple of techniques to help make the
scan better.
First, descreening is sometimes
helpful in the scanner software. If you know the lpi of the previously
printed piece, you can input this into the scanner software and it will
make automatic adjustments to produce the best scan. Sometimes this
does a very good job, other times it is very poor. One trick is to scan
at resolutions that are purposely not divisible by your scanner's
optical resolution. This forces the scanner to approximate each sample,
and can help eliminate unwanted interference patterns by slightly blurring
each sample.
Next, filtering in a raster image
program such as Adobe® Photoshop® can reduce or eliminate unwanted
patterns. Filters such as Median and Dust/Scratches are sometimes helpful.
Inaccurate Color: Scanning colors accurately is extremely important
if the intent is to reproduce those colors as closely as possible. Scanner
optics are improving continually, and color reproduction is becoming
easier. Many factors can affect your scans, and there are simple techniques
to calibrate your scanner. There are also expensive color management
solutions if your workflow warrants it.
The two biggest helps in scanning color
are linearization and color cast removal.
To linearize your scanner means
to calibrate it to some extent such that gray levels are being recognized
correctly. In other words, 100% black is seen as black, 50% yields 50%,
and white yields white. This may sound elementary, but many scanners
are factory programmed with incorrect gamma which makes good-quality
scans, but does not generally produce desired results for commercial
printing. A grayscale strip (target) with known values can be obtained
from Kodak, Agfa, or perhaps your scanner manufacturer. Scanning this
target allows you to determine the values that your scanner is registering
for various shades, and (depending on your scanner software) make corrections
for it.
Removing color cast means removing
any tendency of the scanner to reproduce images with heavy red, green,
or blue values. For example, a neutral color in an image, such as a
white dress or a gray area, should have equal values for red, green,
and blue once it is scanned in RGB. If the area has high green values,
for example, the scanner is said to have a green color cast. Removing
color cast can be done by scanning the grayscale target and changing
the curves for red, green, and blue channels. The idea is that you want
your scanner to register neutral gray values as having equal parts red,
green, and blue.
Hot Spots are found in images having high-key (white) areas
of sharp contrast. Glass-glare, chrome, or even people's foreheads can
have hot spots. In photography, such areas should be reproduced with
pure white. In commercial printing, once these images are converted
to halftones, they will look much better if these light areas are about
5% black (95% white). This means they will retain a small halftone dot
in high-key areas. If they do not, they will gain 'hot spots,' or an
absence of halftone dot. Since the image is printed with halftone dots,
high-key areas look unusually conspicuous if they are not produced the
same way.
