This
article was originally printed in the winter 2000 issue of
JEA's magazine Communication: Journalism Education Today.
Why
you can't just enlarge a digital photo
by Bradley
Wilson
At the time a photograph is scanned (or otherwise
digitized), the amount of digital information in that image
is fixed, based on the settings at the time of the scan. It
will never contain any more information than at the time of
capture.
Still,
time after time, page designers try to enlarge photos beyond
the original size resulting in a loss of quality. When a photo
is enlarged in a page layout program, the pixelation (“jaggies”)
becomes evident.
 |
 SCANNING
INFORMATION
When scanned, an image has a set number of pixels
per inch. In this example, the one-inch square “photo”
has 64 pixels per inch. Half of the pixels are white,
and half are black. If this photo is simply enlarged,
all that the pagination software can do is enlarge
the pixels making the photo look “pixelated.”
If the photo is made smaller, the pixels move closer
together, and the output device throws this extra
information away. CLICK
HERE for illustration of this point.
PHOTO
CAPTION
Tom Malchow, who owns numerous swimming records, looks
up as he points to the sky, Aug. 12, following his
first place finish in the men's 200m butterfly, at
the Indiana University Natatorium. Malchow finished
in 1:56.87, and was under world record pace at one
point. Indianapolis Star Photo by Rob Mattson.
|
The
best procedure to follow is the same one page designers followed
in the “old days.” Sketch out the page and decide
the approximate size of the photograph and then digitize the
image at that size with enough information to reproduce it
clearly.
For example, if you know a photograph is going to be printed
in a newspaper that uses a line screen of 85 lines per inch
(information that you’ll have to obtain from your printer)
and you know it’s going to be about 4 x 6 inches in size,
you can scan it with that information in mind. Entering the
appropriate size and resolution (85 * 2 = 170 PPI) into the
scanner settings dialog box.
Of
course, programs such as Photoshop will allow a user to resample
an image – adding information where necessary to fill
inthe
gaps. The software will take the information that is available
and extrapolate the necessary data. However, this data will
only be a rough approximation of what was really there, and
the photograph will look blurry at best.
Given
that enlarging a photograph past 100 percent results in a
loss of quality, some users prefer to scan images at a resolution
and size higher than they could possibly use. For example,
a newspaper photographer may scan a photograph at 300 PPI
even when they know they’ll be using it at 170 PPI. Because
the photograph contains much more information than necessary
(300 pixels per inch instead of the required 170 PPI), the
output device will ignore the extra information.
Some people are taught to scan everything at 300 PPI regardless
of the output resolution or file size.
This
uniform approach presents its own set of problems, not the
least of which are slower operating speed and an increased
chance for disk errors because the file size will be much
larger than necessary. Larger files are slower and prone to
error.
For example, a 3 x 5 inch photograph at 300PPI will occupy
1.29MB at 300 PPI but only 424KB at 170 PPI (about one-third
of the disk space). The smaller file size results in faster
operations and less chance for disk errors.
To
expedite production when the final size is not known, it is
always better to have more information than not enough. For
example, if you know that a photo is going to be used on the
news page but do not know the exact size, it is better to
scan it a little larger than necesssary perhaps even at a
higher resolution than necessary. Then, when the final size
is determined, resample the image (down) in Photoshop.
For
example, scan a 4 x 6 inch photograph at actual size (100%)
for output at 133 LPI and it will occupy 1.63 MB. If you later
decide that this same photo is going to be reproduced at 2
x 3 inches, go into Photoshop and, under Image/Image Size,
make sure Constrain Proportions and Resample are turned on.
This will keep the dimensions of the photo in proportion.
The photo won’t look stretched. It will also force Photoshop
to discard the unneeded data. Then type 2 inches in for width,
and you will notice that the height changes proportionally.
And the final size (at the top) changes from 1.63MB to 415KB,
substantially smaller and more efficient.
TOP
ILLUSTRATION 1
EXERCISE
1
EXERCISE 2