Digital BW, The Print

on 4/6/02 12:25 AM, John Brownlow wrote:

>>> So just when are you going to define "noise" WRT to a silver print and an
>>> inkjet print so that we have a basis to judge whether it's existence is
>>> RELEVANT?
>> 
>> We have shown how to do it with inkjet prints with pretty reasonable
>> certainty, and shown some proposed methodologies for doing so with chemical
>> prints.
> 
> I would have to think about chemical prints but yes I think that has been
> addressed convincingly. The noise floor is the variation in the paper
> base and the minimum discernible signal has also been defined.

John, Austin, 

Forgive me but what I saw was a theoretical basis for taking some
measurements, but without hypothetical realistic values there is no basis to
know if the result will have relevancy. John, correct me if I'm wrong, but
you even ended your description wondering if it would yield anything of
relevance. 

(No offense, but It also doesn't help that neither you or Austin restate
what definitions you are referring to. ;-)

Please humor me and make a demonstration for me, as I think this more than
anything else will clarify (for me) what this thread has been about.
Honestly, right now, I'm not sure.

Earlier today I spoke of a three print scenario, which I will restate:

> Lets consider a common situation we as photographers
> experience. Say through careful exposure and development we have a negative
> with a beautiful delicate distribution of tones throughout it's useful
> range. It would perfectly fit all it's tones on a given grade 2 paper. Let's
> say that print's density range would be 2 units. If we could measure the
> print on a pixel by pixel basis it would yield a perfect distribution of
> tones in a histogram, right to the ends with no clipping. We could then
> print that negative on a grade 3 paper, which may also yield a density range
> of 2 units, but we "clip" tones from the image on both ends due to increased
> contrast. It's histogram shows the clipping accordingly. We could even
> extend the scenario to where we keep the print on grade 3 paper, but dodge
> the shadows and burn the highlights, thus maintaining a higher internal
> contrast through the midtones, without clipping the ends. It's density range
> is still 2 units, and it's histogram looks more like the grade 2 print, at
> least at the ends.


Lets assume all of these prints were on a smooth glossy immaculate RC print,
which as expected, has minimal surface texture.

Please restate the formula each of you would apply to analyze the difference
in DyR between these three prints. Please plug into the formula as realistic
numbers as you can guesstimate for each print, and then show how you would
compare the results to make a meaningful conclusion about how and where the
prints differ. Finally, please be sure to fully address the difference in
"noise" between them, and explain how it has played a role in the different
DyR values recorded between them. Extra points will be given to a clear
verbal explanation, for the math impaired, throughout. ;-)

Please participate, I anxiously await your replies. I love the smell of
rubber hitting the road in the morning. ;-)

Todd