Re: paleonet Digital photography again

[Stefan Bengtson <stefan.bengtson@xxxxxx>]

At 10:44 +0100 2001-10-28, Denis Bates wrote:
>I've just been scanning TEM cut film to put on file, using an Agfa
>Scanscan.  I scanned at 300dpi, which actually gave a file size of about
>8-9mb.  I then (in Photoshop) reduced the file size to between 1 and 2mb.
>This still gives excellent definition.  Even reducing the file size to
>about 500kb produces good definition at a print size of over 30x20cm.  I
>haven't converted any of these to slides.
>
>At 18:15 26/10/2001, you wrote:
>>
>>In a somewhat related vein, I'm scanning some photographs to convert into
>>slides, and I was wondering what resolution works best.  I'd rather not use
>>the maximum resolution to keep the size of my files down.
>>
>>Thanks!
>>
>  >Adam

Small file size in itself doesn't tell you much, because you can 
attain it by reducing the resolution, by reducing the pixel depth 
(the amount of information stored in each pixel), or by more-or-less 
destructive compression (such as JPEG). Also, resolution in dpi 
doesn't mean anything unless you couple it to the size of the image. 
If you scan a 24x36 mm negative at 300 dpi and then view it at 18 cm 
width the resolution will be 300x36/180 = 60 dpi, which makes each 
pixel almost half a millimeter in width.

Assuming your 500 kB file is uncompressed and has 8-bit grayscale, it 
means that you have about 500,000 pixels in the picture. With a print 
size of 30x20 cm, this gives a resolution of 
sqrt(500000x2.56x2.56/(30x20)) = 74 dpi. This is about the resolution 
of my computer screen, but it's not really acceptable for print.

For a slide to be projected onto a screen, the necessary resolution 
is surprisingly small. You could try it out on your own pictures, but 
I never use more than about 1500 pixels across the slide. This 
corresponds to scanning a 24x36 mm negative at 2000 dpi. When this is 
blown up on a 2 m wide screen the resolution will drop to a 
ridiculously low 20 dpi, but the point is that the audience is 
looking at the picture from a distance, and for them the resolution 
will be more than sufficient. The focusing of the slide in the 
projector is a much bigger problem.

Stefan
-- 
Stefan Bengtson
Senior Curator (invertebrate fossils)
Swedish Museum of Natural History
Department of Palaeozoology
Box 50007
SE-104 05 Stockholm
Sweden

tel. +46-8 5195 4220
      +46-8 732 5218 (home)
fax  +46-8 5195 4184
e-mail Stefan.Bengtson@nrm.se