Filmscanners mailing list archive (email@example.com)
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[filmscanners] Re: film and scanning vs digital photography
I'll say again something I have stated many times in the past. Humans
are analogue, not digital. We work on a cellular level and most of our
cells aren't lined up in perfect grids, far from it. We, both
evolutionarily and through learning, ignore random patterns in our
vision (and other senses), as random noise, because we are ourselves
have a very noisy perceptual system. We see, hear, taste and "feel" all
sorts of "stuff" which is just random nerve triggering, not to mention
all our noisy functional systems (circulation, our
neuro-chemical/electrical system, etc.). We learn to filter this out.
Film grain itself is not actual information. it is the random structure
used to create the image on it's smallest level. Grain occurs in three
random manners. Firstly, each color layer is laid down with the silver
halide grains in a completely chaotic manner. Secondly, the grain size
is randomized, and thirdly, the relationship between those factors
between the layers is randomized, as well. This creates a "forgiving"
structure. There will be a certain luck of the draw that the image
reproduction will sometimes follow certain grain placement closely,
while other areas may be very non-optimized. However, due to the
jumbled manner of grain, its size relative to our perception, and our
perceptual filters, we excuse much of this.
Again, because they are so foreign to most of nature, grids and
non-random placements and patterns stand out to our vision. The grid
used in a digital camera sensor is one example. Each sensor point is of
the exact same size, and perfectly located in a matrix. All three
colors use the same exact grid format.
Due to the use of the Bayer matrix, the color interpolation required,
and a number of other factors, digital images are intentionally blurred
via electronic filtering. This is why judicious use of unsharp masking
can bring so much detail back to an image.
Actual physical size of the sensor is not relevant to resolution,
however, how many discrete points of information is. So, it is the
number of pixels within the frame that determines resolution, not size.
(Of course, the larger the sensor the larger each sensor point and
therefore the lower the noise, but that's another issue).
The reason the debate regarding image resolution - film versus digital -
continues, is because instrumentation can't really answer it. Yes,
numbers of line pairs can be read, etc. but that isn't how we perceive.
Our eyes prefer random analogue and in spite of the defects in this
method, we have built in filters to deal with that because nature is
designed around random noise.
Digital sound, images, or whatever are... digital. The "edges" are
distinct, and placed without random elements. We notice these things.
No matter how "pure" digital might end up (CD music versus vinyl) we
expect and desire the random elements of noise and variability, and we
perceive these as extra "resolution", "warmth", "more natural" etc.
It's really just random noise, but "we like it". Those are the types of
errors we can live with.
So, this debate cannot be answered by machines. It can only be answered
by human consensus. At some point, the digital image components will be
beyond any human's ability to perceive as discrete components, (other
than with massive enlargement) and then the issue will be moot, and for
some it is so close to that now, that is already is moot, considering
the many other features digital images and music supply.
I will once again also mention that the environmental damage done by
silver halide photography is great enough that even if digital is not
perfectly equivalent, the options are close enough that each of us
should seriously consider our ecological footprint when making a
decision about digital versus silver based. It is not that digital is
without a footprint, but in the "big picture" it is likely much smaller,
and studies to date seem to suggest that.
R. Jackson wrote:
>On Jul 1, 2007, at 6:00 PM, Laurie@advancenet.net wrote:
>>Yes because you are mixing apples and oranges in your comparison.
>>and D2X produce a 35mm equivalent first generation capture; it does
>>to be converted into a digital file after the capture by a second
>>process. A 35mm film capture's quality after scanning will depend
>>film uses, and how it was processed, for starters, and the scanning
>>film will comprise the equivalent of a second generation capture
>>possible introduction of noise, artifacts, and other degrading
>>during the scan.
>I'm sure I'm not the only one who's going to find this a little
>suspect. I own a D200 and I like it quite a bit, but at the end of
>the day I like a scanned Kodachrome/Velvia slide more most of the
>time. Now, it's true that the slide may well end up being a
>troublesome scan and may have dust or other artifacts that I'll have
>to clean up and won't ever completely transfer to digital. There's
>always going to be a percentage of what's on that transparency that
>doesn't make it into the computer for whatever reason, but it's still
>a pretty good source, IMO. At 4800 dpi a 35mm scan is 6255x4079.
>That's over 25 megapixels. I can't really tell the difference between
>a 4800 dpi scan and a 6400 dpi scan, so I never go higher than 4800
>dpi, but it's still a pretty decent capture medium, IMO.
>Not knocking digital. It's cool. Very convenient. Very high quality.
>And I'd agree that the D200 is probably resolving as much detail as
>film, more or less. It's just that film's detail extends down to its
>grain structure and things that the lens didn't even necessarily
>resolve, as well as having a different appearance in general than
>electronic capture. A certain vibrance in things like afternoon
>sunlight seems to be there on film that I, at least, have real
>trouble duplicating with digital properly. The instant feedback is
>very conducive to a sharp learning curve, though.
>Santa Rosa, CA
Unsubscribe by mail to firstname.lastname@example.org, with 'unsubscribe
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or