Re: Film or Digital? The Eternal Question!

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

 



Thanks James,
You'r quite right with the DR stuff. But consider that if those subtle transitions in the shadows and highlights require 16-bit depth for preservation, then converting the image back to 8-bit will lose the them.
This has been my experience. If I had a 16-bit workflow from beginning to end, then it would make sense to continue working in 16-bit beyond the demosaicing.
But alas this is not so and I have found by experiment that it makes no difference if I convert my 12-bit Raw file to 8-bit or 16-bit, PROVIDED that in the end I convert the 16-bit to 8-bit anyway.
 
It's a scenario I found hard to accept but It's in my face repeatedly.
 
herschel
 


James Schenken <jds@xxxxxxxxxxxxxxx> wrote:
I've been watching these two threads for a while ( Dynamic Range and
Film v. Digital ) and thought it time to put in about a penny-farthings worth.

On this issue of digital v. film and dynamic range, it seems to me
that there seems to be little to no data out there on the actual
dynamic range of image sensors as installed in digital cameras. What
we do have is a difference in the encoding ability of 8-bit encoding
v. 16-bit encoding of whatever dynamic range is being recorded.

So, 8-bit encoding give about an encoded DR of 2.4 ( log(255) base 10
) while 16-bit encoding gives about an encoded DR of about 4.8 (
log(65535) base 10 ). Neither of them give much information about
the actual dynamic range of the scene being recorded in the first
place. Trying to measure recorded DR by changing f/stops and / or
shutter speeds without recording the actual dynamic range of the
scene leaves too many variables unaccounted for to make any sense to
me ( Here I'm more than eager to have my lack of appreciation
elucidated on so that it does make sense. )

So what I'm left with is that 16-bit encoding is much more useful in
recording subtle differences in a scene especially near the endpoints
of maximum and minimum brightness. 8-bit encoding cuts off a
possible set of 255 gradations at either end before you get to the
point of lost shadow or blown highlights when compared to 16-bit
encoding. So, capture it in 16-bit, do all the color corrections,
contrast manipulations, etc. while still in 16-bit mode and then and
only then convert to 8-bit for final output to a medium that will
have difficulty in recording even 255 steps of gradation when
carefully handled.

On the issue of digital v. film and which is "better", if you look at
B&W film really closely, you'll see that it is not analog at all but
is completely digital in nature. Each grain of silver compound in
the film either switches state and is developable or is not. So,
because of resolution issues, we "feel" that the image is analog when
it is in fact discrete black and white chunks. ( The issue in color
film is a little cloudy since the color image is generated into dye
clouds representing individual silver grains after development and
color clouds in different layers overlap inexactly giving a much more
nearly analog feel to the image. ) The main difference in
'digitalness' about film is that the image is stochastic in nature
while the digital sensor has a grid-like structure to it that seems
to us to be very different.

As to which is better??? Film has a long history of stability under
a variety of conditions if we neglect the post WWII period for
Kodacolor when the medium was so unstable as to almost preclude the
survival of any image at all after 30 - 50 years regardless of
storage conditions. B&W does better if properly processed and stored
but we have little data about the longevity if the current crop of
silver-based emulsions and even less about the chromogenic ones still
on the market. Almost none of the film based imagery of the last 20
years or so will survive because nobody cares to preserve them. We
know about the stuff more than 50 years old because of the tiny
proportion that is still around that has lasted and for which some
few persons actually care. In my parent's family almost all photos
were lost in a storm years ago. In other families the same is true
or they got thrown away upon the last death of that generation and
sometimes even earlier with a move to another residence. Even for
those preserved, little or no information about the images was
preserved - who are these people and why did my grandparents care to
keep this picture are common questions.

The same is going to be true of digital images captured today. Some,
most???, will be discarded immediately. Some lost due to lack of
care when the recording technology changes or format changes. ( I
still have a couple of 5.25" disk drives lying on the floor next to
me. ) I even have some files that are still intact but the software
needed to decode them is long gone. Others will be in the same
boat. Preserving digital images means not only care in recording and
duplication but making sure that the file can still be used after
some time has passed. Image corruption can creep in and careful
copying without verification only keeps dead data preserved, unreadable.

I'm reminded of this on this anniversary of hurricane Katrina. All
those families who carefully kept the albums and negatives in New
Orleans find that they are completely gone and
non-recoverable. Those who digitized and sent copies to friends and
relatives around the country can get them back and recover
completely. You can't easily make an actul duplicate of a negative
or slide to preserve it at another location ( or cheaply ) but you
can do so with digital.

Which is better? It all depends on what your goal is.

Sorry that this got to be so loang-winded.

James



At 09:42 AM 8/27/2006 -0500, you wrote:
>On 8/26/06, PhotoRoy6@xxxxxxx wrote:
>>
>>
>>
>>Film is better. You store them in a box (or in chronically ordered albums of
>>slides like me) and scan only the good ones. Why clutter up your hard drives
>>with digital stuff you might or might not want to look at in future years.
>
>I'd have to list storage on disk as one of the biggest *advantages* of
>digital over film. I can reliably find *everything* I've shot
>digitally. Of course, that may just be because I have 37 years worth
>of film around the house.
>
>And except for some early slides (which I still regret), the physical
>film files are as complete as the digital files (i.e. missing only
>total technical failures).
>--
>David Dyer-Bennet, ,
>RKBA:
>Pics:
>Dragaera/Steven Brust:
>

James Schenken




Herschel Mair
Head of the Department of Photography,
Higher College of Technology
Muscat
Sultanate of Oman
Adobe Certified instructor
 
+ (986) 99899 673
 
www.herschelmair.com


Stay in the know. Pulse on the new Yahoo.com. Check it out.

[Index of Archives] [Share Photos] [Epson Inkjet] [Scanner List] [Gimp Users] [Gimp for Windows]

  Powered by Linux