[filmscanners] Re: Density vs Dynamic range

["Julian Robinson" <jrobinso@pcug.org.au>]

Todd wrote:

>I would presume that what is said is that "neg film can HOLD a greater
>dynamic range than slide film". In this case dynamic range is a misnomer -
>what is meant is it can hold a greater luminosity range, as the visual world
>is made up of reflected/transmitted light, not density.
>
>So, there must be a point at which either film may of itself HAVE (not hold)
>a higher dynamic range, and the parameters of that would include the scene's
>luminosity range, exposure, development, etc. IOW, how the scene is tailored
>to fit within the density range of the film.
>
>Sorry to nit-pick the issue, it's just that so much of the dynamic range
>conversation seems to hinge on exact definitions it seems relevant. Maybe
>not.

I think I agree with what you have said here and in other posts, but I
agree it is important that people not be MISLED as I think they are at the
moment. Let me have a go at putting a picture of all the scanning ranges
and dynamic ranges in one post.  I may sound overly simplistic, not meaning
to me insulting but I am trying to avoid confusion....

The term "dynamic range" is simply defined for ANY system as the ratio of
the largest possible signal to the smallest possible signal which can be
handled without changing system parameters (usually smallest is determined
by "noise" or some equivalent).

So, the world, out there, has a possible brightness dynamic range which is
very very high - looking at the sun might be the brightest we can easily
see, and inside a completely sealed box or room is pretty black.

Any particular scene in that world will have its own "range" which cannot
be truly called a dynamic range because it doesn't contain the largest or
smallest POSSIBLE. But any scene will have its own brightest part, and its
own darkest part.  Strictly speaking we are talking about the scene range,
not dynamic range.  I think you (Todd) called it luminosity range which I
think is a good term.

A typical scene range might be 5 to 15 stops (we can use D measurements
(logs) but I prefer stops; for the record this corresponds to D range of
1.5 to 4.5).  For the sake of an example, lets say the scene is a diverse
one and has a 15 stop range (or Drange of 4.5).

Now, negative film has a recordable dynamic range of around 11 stops or
D=3.3 (I am using supra 100 as my example because I have the curves in
front of me).  That is, the ratio of the brightest recordable part of scene
to the darkest recordable part of the scene is about 11 stops.

The neg film will therefore restrict the scene recorded information from
its original 15 stops to 11 stops by cutting off scene luminosity above
and/or below that range.  You will lose the brightest and/or darkest parts
of your scene.

So on the film we have recorded a scene with a range of 11 stops, as a
resulting density range which is different from the recorded scene range
because the film compresses.  The max range of *densities* the film can
hold is around 6 stops (supra again), this accurately represents the
original 11 stops range of the scene that we captured.

We can say therefore that the film itself will have its own "density"
dynamic range as seen by an observer or a scanner, which is the ratio
between the lightest possible part of a negative to the densest possible
part of the negative.  In this example for a negative it is 6 stops or D=1.8.

Now the scanner comes along.  It's job is to get that 6 stops range off the
film and into digital form. To do this it needs a dynamic range equal or
better than the range of the film "density" dynamic range.  This is not too
hard for negatives, as consumer film scanners generally have a genuine
dynamic range around 2.5 to 3 or so (8 to 10 stops).  No problem. We scan,
we have all the information for the original 11 stops that we captured out
of the 15 stops that were in the scene.  We might now have to play with the
image in Photoshop, because an 11 stop scene printed onto paper with
density dynamic range of about 6 stops will look very flat, but the
important thing is that we have 11stops of the original scene to play with.

For slides, the situation is a bit different because the image densities on
the film are not compressed but are in fact expanded - the range of
densities recorded onto slide film is in fact GREATER than the original
recorded luminosity range.  For Astia film, the recordable dynamic range
(the scene luminosity range that can be captured) is about 7 stops.  The
resulting density range on the film is about 11 stops.

To rephrase, the density dynamic range of slide film is 11 stops, or D=3.3.

This is a much trickier thing for consumer scanners to deal with, with
their realistic dynamic range of only 2.5 to 3.  In fact they cannot
capture all the information of a fully exposed slide (slide covering the
full slide dynamic range).

But, as I discuss in other posts, it MAY be that the scanner has a TOTAL
RANGE of more than its dynamic range, and you could (and people do) use
this as follows:

Adjust the exposure of the scanner so it is capturing the darkest part of
the slide, and do a scan.  The brightest parts will be burnt out - that is,
saturated, because of the limited dynamic range of the scanner.  Now
readjust the scanner exposure so that it will capture the lightest part of
the slide and scan again.  This time the shadows will be lost.  But you
have then used your limited scanner DYNAMIC range to capture in two
separate passes a larger TOTAL range that was on the film.  The scanner has
a bigger TOTAL RANGE than its DYNAMIC RANGE.

To repeat this for clarity, the slide film had a density dynamic range of
D=3.3.  We used our scanner with a DYNAMIC range of (say) 2.5 to capture
the whole of the slide dynamic range in two separate passes.  Thus we can
say the scanner has a larger TOTAL range (or INPUT range or maybe call it
NON_DYNAMIC range) than its dynamic range.

This is one reason why manufacturers specify Dmax figures instead of
dynamic range because the former  can be considerably higher.  There is no
necessary connection between Dmax (loosely describes total range) and
dynamic range.

I hope this makes sense,

Julian R




----------------------------------------------------------------------------------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe 
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or 
body