At 03:26 29/01/03, Bob wrote:
>Fuji may just have come up with a solution to that problem. They are
>reported to have a chip with a second layer of detectors under the first, so
>that the second layer only detects the very bright light that saturates the
>top layer. By merging the two they claim to be able to extend the dynamic
>range of the image. Oops, I probably shouldn't have used the word 'dynamic';
>I'm bound to upset Austin!
Bob,
Very interesting, and the whole subject is worthy of discussion. The point
that people mostly miss is that there are two conflicting desires - a) to
gather wide brightness range, and b) to present an image to the viewer that
has about the same contrast as we subjectively see in real life.
These are contradictory. If you gather more than about 6 stops (maybe it
is 7 or as high as 8) and print it or display it on screen, it looks awful,
flat and low contrast. So if you do collect your 9 stops worth of image on
a digicamera, you have to re-expand it for presentation to something like 6
stops.
I wonder what Fuji are doing about this? I don't have a digital, so I
don't really know how digital cameras handle this problem. Does the raw
file contain all the image brightness range? If so it should be very flat
when viewed directly. Or is the "raw" file already expanded?
I assumed until now that they just threw away the info outside the 6 or so
stops that they were presenting, but this may not be so if they keep it all
in the raw file.
Can someone enlighten me? And what are the figures for modern cameras
for what I will call here dynamic range? And what range do they expand to
for presentation in the "final, non-raw" image? Does the user have any
control over this?
>For the time being I am putting extra range in my prints from a D100 using
>Nikon Capture. It appears that this program allows the exposure of the raw
>file to be corrected by + or - two stops. So by running the raw file through
>Nikon Capture three times (with normal exposure, overexp, and underexp) one
>can layer the three images (without any of the possible registration
>problems in doing this with a scanner) and paint in the extra detail into
>the highlights and shadows of the normally exposed image. I dread to think
>what Austin is going to say about this, but it seems to work.
That makes sense - exactly what I do with my non-digi neg scans - it's a
bit of work, but it improves the appearance of wide-dynamic-range images a
lot. Maybe you have answered my doubt here, maybe I can do what I want to
do after all. Anyone got a spare digital camera? ;)
Finally one thought - not much in the way of slides from a digital camera.
Julian
Julian
Canberra, Australia
Satellite maps of fire situation Canberra and Snowy Montains
----------------------------------------------------------------------------------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or
body
