This column will attempt
to provide solutions to problems readers may have in getting into and
using digital cameras, scanning, and using digital photographic images
with a computer and different kinds of software. All questions sent
to me will be answered with the most appropriate information I can access
and provide. However, not all questions and answers will appear in the
column. Readers can send questions to me addressed to Shutterbug magazine,
through the Shutterbug web site, directly via e-mail to: firstname.lastname@example.org
or by US Mail to: PO Box 2830, Lompoc, CA 93438.
Q. I read with interest
the letter and answer in Shutterbug (October, 1998) regarding scanning
from Rolfe Rieckerd of Tempe, AZ. I would like to contribute that my
Minolta Dimâge Scan Multi uses a formula to establish input (for
printing). It is (dpi of the printer)(magnification factor)=input dpi
from the scanner. Thus, if one is using a 35mm or 6x6, the input scanning
dpi is quite different for the same size of print. Because Adobe gives
dimensions in pixels one can calculate the ppi and eventually the dots/pixel.
I can also use the Minolta to enter numbers in pixels per inch, I believe.
I have a great deal of difficulty figuring out the information on page
51 and 53 of the Adobe Photoshop 5.0 user guide because on page 52 they
are scanning to a screen and then on 53 to a printer. I also do not
understand the interconvertibility of the units: dpi, lpi, and frequency.
Can you enlighten me, please?
I thank you in advance. I find it very difficult to get a handle on
all this and I think part of the problem is that sufficient care with
units of measurement is not taken. I gather that the high-end printers
(Epson Photo EX) vs. my HP 890 is that they print more dots/pixel. But
is dots/pixel resolution? I thought resolution was pixels/inch and that
the resolution needed was dependent on the output device. Is that about
answer your question about convertibility of the three different resolution
measurements, please understand that they apply to different kinds of
devices. Therefore, the factors are not convertible on a purely mathematical
basis--apples and oranges.
First of all, dpi (dots per inch) is primarily a measurement that is
appropriate to printers. A dpi figure states how many dots per inch
of ink are applied by the printer determining how fine the printing
is. Second, ppi (pixels per inch) refers to the digital image resolution,
including the output of a scanner and defines the fineness of the raster
of a digital image. A higher ppi number indicates more, smaller pixels
per square inch of image. If you change the size of a digital image,
you essentially change the ppi count by increasing or reducing the size
of each pixel.
Finally, lpi (lines per inch) refers to the resolution count of offset
or lithographic printing and is based on the fineness of the screen
used to make halftones. For most photographers, unless a digital image
is intended for separation, CMYK conversion and halftoning, directly
in preparation for making offset printing plates, lpi should not be
Q. My fiancée
does stained glass and has been working on a web page for over a month
now. As a photographer, I of course volunteered to be her default photographer
to photograph her stained glass for the web site she's working
on. I just finished doing four different pieces on 4x5 Fujichrome yesterday,
but at this point I'm not sure what we should have the lab do
with the images. They offer two choices. First, have the 4x5s duped
onto 35mm copy chromes, then scanned for web site use, or second, have
them drum scanned which is way more expensive per image as you must
know. It just seems a little silly to go to the trouble of shooting
in 4x5 just to have the images duped onto 35mm film and then scanned.
Is there a third or fourth way to approach this? We tried making photographs
on color negative film and then doing scans on our own scanner, but
the results were very unsharp and unacceptable. One more question for
you if you have the time. Do you know of a good book on the techniques
of photographing stained glass? I have not been able to find anything
dealing with stained glass photography anywhere. Thanks much for your
time. I look forward to hearing from you.
First of all, check those 35mm color negatives directly with a loupe
and if they are sharp, have the "lab" scan them or have
them scanned by a Kodak Photo CD service. Then using an image editor
to open them, apply an Unsharp Mask filter, and I think they will have
more than adequate sharpness for a web site. Web images need only be
at screen resolution, which in VGA mode is just 72dpi. That's
not very high-resolution, so don't expect a lot of highly detailed
For other output your 4x5s may be useful. I would then suggest a scanned
file size for a specific print size from a service bureau. To avoid
the higher cost of drum scans of 4x5, shop around for a service bureau
that has a good flat-bed with a transparency adapter like a Linocolor.
They are a bit faster than drum scanners and the fee should be more
modest, depending on the file size of the scan. For an 8x10 print, scanning
at 400dpi should be adequate and 600 dpi is the minimum for a 11x14"
print. Do not go to duped 35mm for the very reasons you stated, that
would cost extra anyway.
As to your question about a book on how to photograph stained glass,
I cannot recall one that has that specific coverage. However, some of
the better studio photography how to books usually include glass photography.
The principles involved are similar. First you want to use a clean white
background that's evenly lit to provide backlight for the transparency
of the stained glass. Then, to show the front surface texture, an additional
soft light of a much lower level, and at an angle of about 45° is
needed. Unless you need really big prints in addition to web images,
35mm is sufficient for your purpose. It is also faster and cheaper,
so you can do some experimenting with the light balance, as well as
some bracketing to assure getting an appearance that works.
Q. My question concerns
the issue of color management. I am sure that every beginner to digital
imaging notices that the image on his/her computer screen rarely matches
what comes out of the ink jet printer in terms of color. How does a
novice approach the subject of getting monitors, scanners, and printers
to all show the same color qualities so that what you see on the screen
is what you get on the printer?
Thanks for bringing up the color management topic. It is a subject of
prime interest to many these days, as well as to me in what I do on
a daily basis. First of all let me say that color management is a work
in progress. It is reasonably well implemented and effective on the
Mac in the form of ColorSync 2.5. It is less so in Windows 98 unless
you are working with Photoshop 5.0.2, and have the good fortune to be
using devices for which there are ICM profiles.
The entire purpose of color management is first to provide WYSIWYG (What
You See Is What You Get) control of color. Color management is governed
by a standard set by the ICC, the International Color Consortium, which
provides the reference specifications by which digital imaging devices
can be characterized and profiled. Profiles (.ICM files in Windows,
.ICC files on a Mac) are reference information files which describe
the particular characteristic performance of a device like a scanner,
monitor, and printer. The color management system, like ColorSync or
ICM in Windows 98 is the operating system engine (CMM) which relates
the devices used on a system and informs an application what profile
to use, as well as providing the means to process the interpretation
of image file data, translating it from one device to another.
Each device like a scanner, monitor, or printer has its own unique way
of handling color. You could say the problem and function are analogous
to determining by testing to select color filter packs to match output
for a particular color film, printing paper, for use with several different
enlargers and processors in a photo lab. The uniqueness of a particular
device is described by the process of "reading" or reproducing
a standard (ICC) color reference called an IT-8. How the device scans,
displays, or prints the IT-8 reference image is the "characterization,"
which when processed by software results in a profile file that tells
the color management system how the device handles color. The color
management system can then take information from a profiled device (scanner)
and translate it to a profiled monitor. The image then displayed is
independent of either device by being consistent with the ICC standard.
This allows a user to adjust the image according to its appearance on-screen
with the expectation that the color correction applied will result in
predictable output. That is, if the output device, like a printer, is
The most recent version of Photoshop (5.0.2), when installed and run
for the first time, opens a Wizard which assists the user in making
the necessary choices to setup Photoshop to use color management to
obtain predictable WYSIWYG output for whatever kind of input the user
requires, and is connected to the system.
My experience so far with both a Mac G3 and ColorSync 2.5, as well as
with Windows 98, both running Photoshop 5, has been that when the entire
system is set up correctly and all of the device profiles are in place
and identified, I will obtain very close to what I expect from the scanner's
driv-er dialog, through editing in Photoshop to making a print that
matches what I see on-screen.
There is more to it and there can be problems, as well as the possibility
of having a device for which there is not a profile. There is also the
more advanced level of color management which involves using measurement
devices and software to create custom profiles for scanners, monitors,
and printers, which take into account any variables there may be in
particular devices caused by age, use, and other system factors. To
make any remarks beyond this general description or provide any more
specific answers, I'll need specific information about the devices
involved, the software used, the operating system involved, and what
the problem is.
Q. I wonder how I
might discover if there's SilverFast software for my Umax Astra
the Umax Astra 1220s is one of many makes and models supported by Lasersoft
SilverFast software. More information can be obtained including prices
from their web site at: www.lasersoftint.com You can also send direct
inquiries by e-mail to: email@example.com or laser firstname.lastname@example.org
The US headquarters of the company is located at LaserSoft International,
Inc., 6529 Gulfside Rd., Longboat Key, FL 34228; (941) 383-7496.
Q. When I purchased
my new monitor, a Mitsubishi Diamond Pro 1010e, I received a free copy
of "Diamond-Match" a color calibrator kit that includes
"Colorific." My question is, should I use this instead of
Adobe Gamma? I'm confused because the first paragraph of the DiamondMatch
manual said, "If your software application already has a color
management system built-in, use that." My second question concerns
color temperature. In all the information I've read on color management
none have discussed what a preferred color temp setting for a monitor
should be. My monitor has adjustable color temp. It came from the factory
set at 9300. I know a color temp of approximately 5000-6000 is considered
"daylight," but should I set my monitor to this and then
run Adobe Gamma?
follow that advice. Use Adobe's color calibration Adobe Gamma,
particularly if your primary work is being done in Photoshop 5.
You will obtain more effective color correction if your monitor matches
your transparency viewer in color temperature. Although my viewer is
ostensibly "daylight" and is a very high-end brand, I find
that my monitor most closely matches it when the monitor is set at 6500°K.
Then once the color temperature is set using the controls on the monitor
(which affects white point), rerun Adobe Gamma, and be sure to check
the box in the Adobe dialog which allows you to specify that "color
temperature is set in hardware." Then if you have a "generic"
profile for your monitor provided by the manufacturer which includes
a selection of color temperature variants, choose the color temperature
you have set the monitor to and are using, and have Adobe Gamma rewrite
a new .ICM/.ICC based on that profile which will become the new monitor
profile you'll use, making a "copy" version of the
Q. I anticipate purchasing
an Epson 1520 printer within the next 3-6 months. I recently purchased
Photo-shop 5.0 and it comes with a PostScript level driver. If I want
to use the PostScript option on this printer, do I still need to purchase
the PostScript option?
I enjoy many of your articles. I really enjoyed the recent review you
did of the Epson 836XL scanner. I have the book you did on photographic
lighting. When are you going to do one on digital photography? I thought
the book on lighting was straightforward and easy to understand.
be able to use the Photoshop PostScript driver, you will also need the
PostScript upgrade option for the printer. PostScript is a unique page
description language used in desktop publishing primarily, and a printer
requires both the PostScript software
driver and the PostScript hardware to rasterize this kind of image information
to send it to the print mechanism to put ink to paper.
Unless you are doing publishing work and require soft proofing of documents
which are designed for offset printing, or want to print .EPS files
directly, PostScript is not needed for Photoshop output.
As for doing a book on digital photography and computer image processing,
I agree there are some potential readers among photographers probably.
However, this technology is evolving too fast and the products are changing
too rapidly to make paper publishing in book form viable--the book would
be obsolete before it got on the bookshelves. But thanks for the suggestion,
the proposal is flattering.