Re: how to convert metric to pixels
"Eric Sosman" <email@example.com> wrote in message
Joshua Cranmer wrote:
Patricia Shanahan wrote:
I'm typing this on a laptop with an external display connected, in
extended desktop mode, and the composition window placed so that it lies
across the boundary between the displays. The two displays have
different characteristics. What is the pixel size for this window?
On most Windows computers, the answer will be 96 dpi: the value is
controlled by a setting in Display Properties->Settings->Advanced, and
very few people actually change it.
The setting has little to do with physical reality. I've
just measured, and on the plain-vanilla monitor I'm using right
this minute, the resolutions are 76.2+ and 71.4+ dpi, horizontally
and vertically (30 dots/cm and 28.1+ dots/cm). Windows calls it
"96 dpi" -- high by one-third -- and ignores orientation.
How did you measure that, Mr Sosman? I used the pixelRuler from Roedy
Green's site, and compared, on screen, to the inch ruler in openoffice.org
writer. My result was 96. Then I marked a paper against the OOO inch ruler,
and then measured that with a physical ruler. All are in agreement.
My measuring instrument was a flexible tape of the kind used
by tailors and seamstresses: A sturdy tape of about 3/4 inch (2 cm)
wide and -- I'm guessing -- 1/32 inch (0.8 mm) thick, approximately
59 inches (150 cm) long, the "approximately" due to the difficulty
of reading the markings at the the metal reinforcing tabs on the
two ends. The tape is yellow and slightly shiny, with black marks:
Inches in eighths on one side and centimeters in halves on the other.
When rolled up, the tape forms a coil with a diameter of about 1 inch
(2.5 cm), but the coil is somewhat more elliptical than circular and
this dimension is only approximate.
I measured the horizontal and vertical dimensions of my screen
using the tape's inch scale. To avoid the imprecision induced by the
metal tab, I set the 1-inch mark at one edge of the visible image and
measured the position of the opposite edge, then subtracted 1 from the
result. Then I consulted Display->Properties->Settings to learn the
screen dimensions in pixels, and dragged out my trusty calculator to
perform the necessary divisions.
My measuring device's accuracy is suspect. The tape is subject
to repeated coilings and uncoilings, and probably undergoes physical
deformations as a result. Then, too, it is made of a material whose
dimensions may be affected by humidity and temperature. I would not
recommend it for measuring the diameter of a human hair, the distance
between the neutrons in a tritium nucleus, nor any distance at all in
circumstances subject to appreciable relativistic effects. Nor would
I myself use it for measuring hips, girths, busts, inseams, and the
other dimensions for which it was intended -- not because I doubt the
tape, but because I doubt my own skill in such measurements (the bald
guy with the little piece of chalk, though: him I'd trust). Even so,
I am confident that my small skill with this simple instrument is keen
enough that off-by-one-third is so many sigmas away from the mean as
to be incalculably unlikely.