Re: Best CI for process

From: Richard Knoppow ^lt;[email protected]>
Date: 10/02/05-09:17:46 PM Z
Message-id: <004301c5c7c9$09a947d0$49f75142@VALUED20606295>

----- Original Message -----
From: "Katharine Thayer" <>
To: <>
Sent: Sunday, October 02, 2005 12:54 PM
Subject: Re: Best CI for process

> The more you guys talk about this, the less I think I
> understand.
> If CI is arrived at by choosing a starting point density,
> counting off
> 2.0 log exposure units from there to find the end point
> density, and
> calculating the average slope of the line between those
> two points, and
> if log density and log exposure are related in a lawful
> way, which I'm
> told is so, then negatives with exposures that fall
> beyond the end of
> the portion of the curve for which the slope is
> calculated, like the
> three negatives Sandy listed earlier, will have to have
> the same CI
> almost by definition, simply as an artifact of the
> arbitrary truncation
> of the curve. It looks to me the CI is hitting an
> artificial ceiling,
> and that in order to get a meaningful CI for these
> negatives, you'd need
> to calculate the slope of a curve section that extends
> into a longer
> exposure range.

   This is a situation where average density makes more
sense than CI. You would be right if the characteristic
curve continued to rise at a constant slope. But, take a
film like Tri-X sheet film (ISO-320). This has a
characteristic curve which is upward deflected practically
its entire length. That means that the contrast is not a
constant value but increases with density all along the
curve, at least until it reaches the shoulder, which is at a
rather high density. Average Contrast is simply drawing a
line between some low and some high density point on the
characteristic curve and measuring its slope. The difference
between Average Contrast (also called G-bar) and Contrast
Index is that Contrast index is measured by extending arcs
from the exposure baseline to the density curve at two
specified log exposure values. The lower value is specified
to be at a minimum contrast value on the toe but that is
hard to determine simply from the curve. This is the reason
that CI is measured with an overlay. However, it can be
approximately measured by finding the point on the density
curve which has a density of 0.1 above base density and fog.
Use this as an end point and draw an arc which intersects
the exposure scale at a distance of log 2.0 from this point.
Draw a stright line between the two points and determine its
slope. This will be very close to CI as measured using the
   There is a good explanation of the various methods of
measuring contrast in:
_Photographic Materials and Processes_ First Edition,
Stroebel, Compton, Current, Zakia 1986, Focal Press
ISBN0-240-51752-0 Call No. TR145.P47 1985 There is a
second edition which may have the same material but the
first is more comprehensive about most subjects.
 See p.51 for CI and Chapter 11 for a good exposition of the
theory of tone reproduction.

Richard Knoppow
Los Angeles, CA, USA 
Received on Sun Oct 2 21:17:59 2005

This archive was generated by hypermail 2.1.8 : 11/07/05-09:46:18 AM Z CST