RE: VMs: RE: Map scans now posted...

["GC" <glenclaston@xxxxxxxxx>]

Very good,

I think what we've established is that maximum benefit is gained by paying
attention to the limits of the hardware in use.  In the scanning process,
the limit of the scanner is the most important element, and only fractional
integers of the scanner's upper limit should be used to avoid uncertain
results due to interpolation.  For routine display purposes, the setting of
72 or 75 dpi, depending on your scanner's preference, is the important
figure.  For "zoomable" images, such as Nick is attempting, again, the
scanner setting is the prefered setting.  If it allows 400 dpi without
interpolation, this may be a better setting.  However, most scanners have
settings that are fractional integers of their total capability - a 1200
scanner can interpolate 2400 or 4800, but still actually scans at 1200, and
does its best at 75, 150, 300, 600, and 1200.

All this relates to file size, which is the most important factor when
making a web presentation.  I note that many libraries around the world
offer detailed scans of their holdings at 300 dpi, and to date I have not
demanded more detail from these manuscripts than what has been offered.  At
this resolution, I can blow up a section on my screen more than 20x, which
is a much greater maginification than was available to the author when he
composed the manuscript, which means I have at my fingers a greater amount
of detail than the author had in writing it.  That in itself should be
sufficient to conclude that scans over 300 dpi for medieval manuscripts are
unnecessary.

As always - this is just my opinion.  (It doesn't matter that I'm right, and
I AM ALWAYS RIGHT, except when I'm not, but we just don't talk about that.)
:-)

GC



> -----Original Message-----
> From: owner-vms-list@xxxxxxxxxxx [mailto:owner-vms-list@xxxxxxxxxxx]On
> Behalf Of J M Cerqueira Esteves
> Sent: Thursday, June 19, 2003 8:55 AM
> To: vms-list@xxxxxxxxxxx
> Subject: Re: VMs: RE: Map scans now posted...
>
>
> On Thu, 2003-06-19 at 13:33, Nick Pelling wrote:
> > dpi is one-dimensional, but scan size is two-dimensional, so
> the difference
> > between 300dpi and 400 dpi is 3x3:4x4, ie 9:16 (ie, nearly 2x
> the size) -
> > so a 600dpi scan would be 3x3:6x6 = 9:36 = 4x bigger.
> > Also: if you start getting down to the grain of the image, you
> start having
> > to encode the surface's texture noise, which can worsen the compression
> > performance etc.
>
> Yes, I was familiar with that but somehow misunderstood the initial
> message and thought GC was meaning that the size difference was even
> bigger than the expected one.  Sorry for the confusion due to my hasty
> reading...
>
>                        J Esteves
>
>

______________________________________________________________________
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
unsubscribe vms-list