"Bill Hilton" wrote in message
oups.com...
snip

For example, in deciding if I want to go digital for Canon, I might be
happy
with the 16.7 MPix 1Ds MII for scenics, but it only shoots 4 fps, so
would
not work well for wildlife

Right, so get the 1Ds M II for scenics if you need to print that large
and get the 1D Mark II for wildlife, which is what many of us like
myself and Roger are using for birds and bears etc. 8 Mpixels is
enough for wildlife, I feel.


This was pretty much my conclusion. But I like my large wildlife shots to
be as sharp as the landscapes, if not sharper. For certain human portraits,
soft focus is desirable, but I like to see the snot in the bison's nose, or
the feather detail on an eagle. Surely these would be better at 16 than 8
MPix. In your experience, how large would one have to print in order to see
this difference? I'm currently only going to 13x19, but would like to plan
for when I need larger.

My 1V is 10 fps, and even with just a desk scanner (4k dpi) I
get 24 MPix (with film)

This was explained to you previously by Roger, about why you can't just
say one is better than another because it has a higher pixel count. To
use your 24 Mpixel value (it's actually more like 21 Mpix if you scan
to the edge of unmounted film, less if scanning mounted slides), if you
shot a crappy high speed film, say 400 iso pushed one stop to 800, it
would be grainy and not as saturated as slower films. You scan it and
you have 24 Mpix, you scan Velvia or Provia 100F and you also have 24
Mpix ... are they the same even though they have the same pixel count?
No. You could make this more absurd by scanning the grainy, low
saturated film with a drum scanner at up to 12,000 dpi and have around
182 Mpixels ... is this 7 or 8 times better than the Velvia scanned at
24 Mpixels or is it inferior for practical purposes? All you've done
is scan grain and you still have poor colors.

For the same reasons, pixels from the better digital cameras are better
than scanned film pixels.

I'm still trying to wrap my cerebral cortex around that one. I guess I just
need to experience it.

When scanning grainy film, the film is the weakest link (garbage in, garbage
out).
But with Velvia under ideal conditions, I assume a 4000 dpi scan is the
weakest link. An R print (or other direct print or drum scan) would lose
less in the translation, so it should be best, right? (At least for
resolution.)

Shooting the same scene under the same conditions (e.g., ISO 50 film and
digital set to 50, same lens etc.), you're saying that an 8 MPix (or 16?)
digital is better than film (resolution AND color?), using the best
availible method of printing each?

Bill


- Alan Justice