Digital BW, The Print

on 4/5/02 1:52 AM, Martin Wesley wrote:

> Austin,
> 
> I am not sure I have much to add at this point. Still mulling things over.
> Charles Francis wrote me off list and brought up the term "internal
> contrast(s)" which I am familiar with. Sometime just thought of as "local
> contrast". This can be a problem in prints where the overall contrast
> balances out nicely in the print but there will be some area typically with
> fine detail of interest that covers just a tiny portion of the density range
> and winds up looking like mud. This seems to be a common problem with
> contracted developments where there has been a strong compression of tone
> from the scene to the negative. Is this in anyway related to what you see in
> the quality of the tonality of the print?
> 
> Martin

See if these clarify anything. I doubt they will, I think they'll point to
the the complexity of all this:

*******
Abstract

This paper is an attempt to integrate a wide variety of psychophysical
experiments into a computational model to calculate color appearance. Having
described the fundamentals of such a model, we turn to applying this model
to printing wide dynamic range, real-life scenes and finding the best
reproduction of an image with limited printer gamut.

<http://www.imaging.org/pubs/reporter/articles/14_6_mccann/index.cfm>


*******
Nonlinear Effect of Modulation on Image Quality
Petet Barten, Barten Consultency

In most image quality metrics, a linear relation is assumed between
modulation and perceived image quality. However, in practice it appears that
this relation is nonlinear. Granger and Cupery (1972) found at their
investigation of photographic pictures, that there is a linear relation
between the subjectively perceived image quality and the number of
just-noticeable differences. From our investigation of sinusoidal luminance
patterns, we found that the number of just-noticeable differences is
proportional to the square root of the modulation. The constant in this
relation becomes independent of spatial frequency if the modulation is
divided by the threshold modulation. From measurements by Cannon (1985), we
found that this rule also holds for the subjectively perceived contrast of
sinusoidal luminance patterns, in agreement with measurements by Watanabe et
al.(1968). The above given principles are used in the square-root integral
or SQRI for the description of image quality. In this method a logarithmic
integration over spatial frequency is applied to account for the
contribution of the different spatial frequency components of an image to
the total image quality. Here, various examples of measured image quality
will be given which will illustrate the given rules.
Peter Barten studied physics at the Technical Univ. of Delft and worked at
Philips in Eindhoven, where he was in charge of the development of color
CRTs. In 1987 he became an independent consultant and developed the SQRI
method for the evaluation of perceived image quality, and a model for the
spatio-temporal contrast sensitivity of the human eye. He is the author of a
recent book on the contrast sensitivity of the human eye and its effects on
image quality. In 1999 he received a Ph.D. degree from the Technical Univ.
of Eindhoven. 

<http://www.imaging.org/conferences/pics2002/keynote.html>

******

Todd