At 10:07 PM +0000 9/21/03, olympus-digest wrote:
>Date: Sun, 21 Sep 2003 12:36:58 +0100
>From: "Julian Davies" <julian_davies@xxxxxxxxxxxxxx>
>Subject: Re: [OM] Nope, It missed it by few hundred miles
>
>Er...
>My fault, in that I was trying to draw a parallel between two entirely
>different results which have the same cause!
>Quite correct that film doesn't alias. It fogs.
>When you hit the limiting frequency, a fixed uniform sample system begins to
>form beating artefacts. Since this is a property of the sampling interface,
>the downstream electronics don't know they are artefacts, and can do nothing
>about it. Nasty. Digital samplers have to take great care not to exceed this
>frequency, however it is defined. There was a debate on the list a while
>back about whether Nyquist was right, but that exceeds my ability to follow
>or discuss by a country mile. Nonetheless there is a must - not - exceed
>frequency in the system which is always enforced by a filter of some kind.
>Analogue recording systems don't form beating patterns when a particular
>frequency is reached (for the purists this is only true when the recorder
>works without modulating a base frequency). They suffer from an increase of
>noise with frequency. In photographic film, this is base fog.
Nyquist was right, and the ultimate must-not-exceed frequency is the Nyquist
Frequency.
Base fog is *not* caused by aliasing, either in film or digital cameras. One
wat to tell is to notice that base fog affects film not in the image, like
between the frames. Base fog is caused by random thermal "exposure" of silver
grains, and is reduced by cooling. Base fog happens even in total darkness.
The classic test is to take unexposed film, develop it, and then read the
resulting film density. This value is what is reported as base density in film
spec sheets.
The haze seen in the image area is from lens flare, and 1 0s considered good in
a photographic lens. You can test this by constructing a photgraphic target
from a sheet of white foamcore board with a 25-mm diameter hole in the center.
Behind the hole put a black velvet bag (held open somehow), so light that gets
through the hole bounces around inside the bag and becomes quite lost, never to
return. Photograph this target and develop the film. By rights, the image of
the hole should have no exposure whatsoever, and yet it will have a slight
exposure. The amount of exposure in the image of the hole compared to the
image of the white board around the hole is the amount of flare, and is a
property of the lens.
Note that in this setup, there are no high frequencies to be aliased.
A few percent flare can have a large effect on image contrast, so studio
cameras very often have big lens hoods on them, to keep side light out of the
lens. Simple and very effective.
>Now I don't understand the precise physics of this, but Geoffrey Crawley
>published an article in Amateur Photographer making just this point.
>If you think of the ultimate limit of three grains of film capturing two
>separated photons (white - black - white bar signal), if the spatial
>frequency is higher than the separation, the outcome is One black, two
>clear. - by definition if one photon hits a grain, the other must miss the
>other two.
>It seems that because of the physical construction of film, light scatter in
>the emulsion makes "both grey" the result of the miss, and also makes the
>limiting frequency considerably lower than the limit case above.
>When you have a set of grey film grains, you have fog, assuming that the
>levels concerned are far below the levels of light at lower spatial
>frequency, i.e. there will be an image recorded on top of the fog.
>Since this effect is a rising with frequency one, there is no defined must -
>not - exceed point. It just becomes one of the factors in assessing lens
>performance, i.e. as we know, the MTF graph does nor reveal all.
>Since film formulation is not a static art, it is also a factor in classic
>lenses being thought of differently in the modern world compared with their
>original production time. Whether it is an enormously important aspect of
>the "Contrast Vs Sharpness" debate, I don't know, but bear in mind that it
>would be easy to produce lenses with far higher MTF results than are
>currently targeted, and would be possible within the price of certain German
>manufacturers.
I haven't read Crawley's article, but this doesn't sound right to me.
For one thing, it takes from 4 to 40 photons absorbed in a certain period of
time to render a grain developable, and the developed grain has about 10^6
atoms in it, whereas the above assumes that a single photon can expose a grain,
which isn't true and cannot be true. Why cannot it be true? Because random
thermal motion would then expose the film, leading to a hopelessly high fog
level. Requiring multiple photons arriving within a short period of time
sharply reduces the probability that thermal motion will randomly expose grains.
Nor are film grains ever "gray". A grain either develops or not.
According to "Photography - Its Materials and Processes", sixth edition,
C.B.Neblette, Van Nostrand 1962, Chapter 24 "The Microstructure of the
Developed Image": One can resolve the individual silver grains at 2500x
magnification, and using an electron microscope at 25,000x, one can see the
filamentary structure of those grains. These are far beyond the magnifications
used in photographic enlargement. The film tests in the magazines talk of
perhaps 40x, so the "grain" one sees in these test shots are in fact granules
of grains.
The bottom line is that the actual grains are far too small for the film's
overall MTF to have any effect at the level of the finest spatial frequency
supported by the film.
Joe Gwinn
< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >
|