After twenty five years of selling my Fine Art
Photography on the national Art Show circuit, I've found it a natural
evolution to move from the chemical darkroom to the digital darkroom. In
purchased a computer and hired fellow artist Chris Maher to build a web site for
me. Being short on time, he agreed to do it if I would learn how to
prepare the images. This started a working relationship that grew into a
partnership. We now specialize in building "image intensive" web sites for
photographers and artists and have established a reputation for creating web
images that are true to the original artwork. Knowing that the
higher the image quality, the larger the file size, the slower the page
made it my objective to create rapid loading, good looking images in
small file sizes.
My personal system consists of a Dell Pentium III
at 550 MHz, with 384 megabytes of ram running Windows 98
SE. I use a Matrox G400 graphics card that features dual monitor
support with a 21 inch Sony monitor to work on the images full
screen and a 13 inch Sony monitor for my Photoshop tool pallets.
For scanning I use a Polaroid Sprintscan 4000 for slides and an Epson
1200S for prints. I also use a 4x5 transparency adapter for my
print scanner for the few times I've only had large format transparencies
to work from. I've found it adequate for the web, especially since
it's a $99 alternative to a $7500 investment. The programs I use
most are Photoshop 5.5 for image editing and ACDSee for image management.
While working, I leave ACDSee open in thumbnail view on my second
monitor, and drag images into Photoshop or my web design program.
The initial approach is to discuss with the artist, in
detail, their expectations and what their options are for the site. I need to consider the
page background color and image size before I begin. I always ask
about the client's monitor and screen resolution, so I can visualize how
they are going to view the web site.
In preparing the image I add a single pixel stroke
(line) around the image so it stands out on the page. For instance,
if the page color is black, I will use a white stroke so the dark
areas of the image will still stand out. If the page color is
to be white, a black stroke and maybe a drop shadow will be used,
so it looks as if the image is hanging on a wall. I usually recommend
sizing the image to 450 pixels long dimension. That size is big
enough to see clearly, but not too big that printing it out
would replace a sale.
At 72 pixels per inch, it will only print out to 6 ¼ inches (long
dimension). The last thing I do before working on the images is
to determine how the client wants the copyright to read. As a
security measure, I recommend adding a copyright, as a text layer
in Photoshop, to every picture. Ideally it should be the URL of
the web site. This way, if someone prints the picture out, one
day they might want to purchase the real picture, and will know
how to contact the artist. Another security measure I've used
is to slice the image. Each image then loads in
pieces, making it harder to capture.
This example of Image
Slicing can be seen at www.Robert
Barab.com. Note the vertical slice line through the picture
and the cell borders which include the drop shadow. Each half loads in a separate cell and is joined in the HTML. This
is how the image looks in my web editing program. As you can see,
Robert chose to use his name and phone number for the copyright
If the client gives me 35 mm slides, I scan them at 1200 dpi.
When I work from prints, I scan them at 300 dpi. In keeping all
my work files organized, I create multiple folders under each
clients name so I can save at every step of the way. I run all
my image processing on the original scanned size Photoshop (psd)
file. I never do any batch processing which affects image quality, taking
each image as a separate entity for the best results.
In this Image from the
Impressionistic.com web site, I've used curves to lighten the
darker areas without letting the highlights get too bright. In
this screen capture from Photoshop, you can also see the single
pixel black stroke surrounding the picture.
slides or prints to work with gives me a guide to match for color.
I use curves to open the shadows so the image on the computer
will look like the print (or painting) the artist is selling. Then
I resize the image by adjusting the height or width
(whichever is longest) to 450 pixels. I run the unsharp mask
filter to bring
back the sharpness lost by resizing and add a
text layer with the copyright. I save at this point by adding
the size to the file name (450-picture01.psd). Note the text is
still an active layer (not flattened), so it can be edited in
the future. Next I flatten layers and add the single
pixel stroke and save again. I am careful here to add an
"s" to the file name for stroke (450s-picture01.psd).
If the pictures require a drop shadow, this is the point where
it will get added. If not, my next step will be to convert it
to a jpeg for the web. One of the new features that made it
worth upgrading to Photoshop 5.5 is a great jpeg conversion
called "Save for the Web." It gives a side by side comparison
between the original Photoshop file and the resulting jpeg. This
way you can actually see the image quality change as the file
size is lowered and know how much you can get away with or how
small you can make the file size before the picture shows signs
of jpeg artifacts. I try to keep all my file sizes to under 40
kilobytes, and most end up between 18 and 30 kilobytes.
When constructing the web site, I use only one image per page to
maximize the chances of a sale. This way the page will load
quickly, and the viewer is not distracted. The thumbnail galleries,
which consist of multiple images of the same theme, are either 175 or 200 pixel images (again, long dimension). I
generally limit the number of images to 12 for faster loading.
All the images are then archived to CD, with both my partner and I