[filmscanners] One more post on the PPI_DPI thread

["Michael O'Connor" <omichael@optonline.net>]

  Personally, I'm not ashamed to be concerned with linguistics, and
though I've been part of this list for a very short time I don't see how
its off topic to discuss, civilly, the terms which anyone involved with
scanning or digital imaging encounters every day.

  28 degrees means something totally different here in the states than
it would in Europe, so there has to be a frame of reference, degrees
alone (even engineering degrees) just won't cut it.

  While I do prefer PPI to DPI in referring to file resolution, since it
makes some relevant distinctions, it is actually an arbitrary
preference, since it doesn't make enough distinctions. The problem with
PPI is that its something of a mixed metaphor, the Pixels are referring
to a monitor display (I haven't really done the research here, I would
think that the term originated with displays before being applied to
files, but if I'm wrong it doesn't really change the argument, the same
term is used for both), and the Per Inch is referring (at current and
forseeable levels of technology) to print output. It would be better for
understanding if the terms for a display's picture elements, a file's
resolution, and the resolution of a print output device, all had their
own distinct terminology. Short of three distinct terms, anyone finding
relevant distinctions by referencing a files resolution as DPI can't
really be said to be wrong, and they probably have the better argument,
at least until they knock head on into output resolution - which many
people never consciously do.

  In usage, either PPI or DPI should only be applied together with
dimensions, but IMHO we all want shorthand, so its rarely the actual
usage.

  If we took available terms maybe ppi for displays, dpi for files, spi
for output devices, and lpi for halftones, would be the best we could do
(almost no one uses samples per inch, and you could use sspi). But I
think we'd be better off getting rid of the Per Inch altogether from
file references, and for easiest comprehension of why a digital image is
so variable as opposed to looking at a print or slide, it would be best
to adopt a term that would dissuade people from thinking of the file as
a captured picture. I think Bit Sets might be the best solution (BS
would certainly be appropriate to our whole discussion here), or BSR,
for Bit Set Resolution; it might fail the best shorthand test, since it
would need both horizontal and vertical references (HBS and VBS?), but
it would clear us of the most hurdles of logic.

Michael O

----------------------------------------------------------------------------------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe 
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or 
body