Re: Was: [Leica] Noctilux Now: Summilux

[Jeff Moore <jbm@instinet.com>]

At 5 Jun 1998 10:55:26 EDT, Tom Shea <TEAShea@aol.com> wrote:

> The differences in quality are small enough that there is no way
> a computer monitor could adequately reveal the level of quality
> involved. ... The improvement in the new 50 1.4 would not be revealed
> by this type of comparison.

Huh?  Are you somehow assuming that one is only allowed to look at
images acanned at a low enough resolution that the entire frame is
visible on a screen?  Or (even sillier) that one is only allowed to look
at pictures at 1:1, so all images from 35mm must be 24x36mm as displayed
on the monitor?

Why couldn't someone scan at a decently high resolution, and either put
up the image of the whole frame (allowing viewers to pan around the
image), or put up cropped chunks from meaningful parts of the image?

> How many lines of resolution can a computer monitor show, even under
> the best of circumstances?  Maybe 5-10 lines per mm?

I'd guess about 4, if you're lucky.  Or would that be 2?  I get hazy
around the definitions of `lines', `line pairs' and `pixels'.  But what
does this have to do with scanning resolution?

To pick a readily-available, bog-standard scanning service: isn't the
highest resolution provided by a non-`pro' Photo CD scan from 35mm about
2k x 3k?  That's, what, 83 pixels/mm?  I wouldn't be surprised if that
resolution were sufficient to show at least some meaningful differences
between lenses.

A raw 24-bit image at that resolution would be a might hefty at 18M,
but with JPEG compression at a not-too-destructive quality setting, I'm
guessing you'd get down to about 1.2M.  Not too big at all, if you have
decent bandwidth.

> The resolution of quality lenses and the differences between the
> current 50 1.4 and the new 50 1.4 are an order of magnitude greater.

Than screens.  More, I'd hope.  So?

> Perhaps this reveals something

Yep.

> about the varying standards that different users have.

Maybe.

 -Jeff Moore <jbm@instinet.com>