Apache-Talk @lexa.ru 

Inet-Admins @info.east.ru 

Filmscanners @halftone.co.uk 

Security-alerts @yandex-team.ru 

nginx-ru @sysoev.ru 




      :: Filmscanners
Filmscanners mailing list archive (filmscanners@halftone.co.uk)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: filmscanners: Film Scanner Question Again

Hi Rick,

Actually, the manual is correct.  The error you are making is in the
size of the file you expect you will be creating.  If you are making a
scan of a 35mm film frame, you don't need to scan the whole flatbed
size, only 1" x 1.5", as you states.  This doesn't make a 700+ meg
file.  The size of the file depends upon a number of things.  1) The
size of then source image and the resolution used, two the number of
color layers involved (black and white, grayscale, RGB, CMYK) and
thirdly the bit depth of each color channel.

If you use 8 bits per channel (color) on a RGB scan of one frame it will
be approximately 18 megs at 2400 dpi, It will double if you scan at 16
bits per channel.  The equation you used is correct, using the 300 dpi
input to printer resolution, which is about correct for an inkjet

So, other than your assumption about the size of the file, the rest is

However, keep in mind that your flatbed has a maximum optical scanning
ability (probably 1600 ppi on your scanner).  Anything above this is
simply interpolated, and provides no additional real information, so
scanning beyond the maximum optical value has no advantage.  You might
as well allow the printer driver print at a somewhat lower dpi to
the limitation of the scanner.  Or you might compare what happens by
increasing resolution in Photoshop or another software package, via 
upsampling, and see which gives a more pleasing result.

As to if there is a point where these are no diminishing returns,  it is
higher on some printers than I originally believed, which was based upon
the older Epson printers.

About a year ago, someone sent me scanned samples from an Epson 1160?
printer and it did show minor additional detail when going from 240 to
300 to 360 to 372.7 (which was the size he ended up with without doing
any downsampling) to about 400 dpi, at which point the results pretty
much plateaued.  The differences beyond 300 dpi took some careful
scrutinizing to see.

The problems with large files are that image manipulation is slowed
down, storage space gets used up and the printer spools the file more


Rick Decker wrote:

> Now let's say that I want to take a 35mm slide (1x1.5in) and enlarge it
> to 8x12...my resolution would be 2400 according to the manua
> (12/1.5=8...8/1=8...300x8=2400)l.  At 1600, I would have a file size of
> 705 megabytes!! I am sure that this is way beyond the point at which the
> increase in file size does not result in any more increase in data.
> It sounds to me like I should leave my resolution at 300 dpi/ppi...even
> that will give me a file size of 24megabytes which I suspect is larger
> than I need.  For an 11x16 it would be 45 megabytes.
> Any advice is much appreciated.


Copyright © Lexa Software, 1996-2009.