Gamma adjustment

Digital imaging software does not work like this.
Additonally you are working on your image on a computer monitor that has
entirely different characteristics than a print.You have to tailor the image
to its intended use.
You cannot think photochemically in a digital world.
But you know that and thanks for posting something other than spam
advertising. . . .

On Sat, 19 Apr 2008 02:45:37 GMT, "pickled"
wrote:

Digital imaging software does not work like this.
Additonally you are working on your image on a computer monitor that has
entirely different characteristics than a print.You have to tailor the image
to its intended use.

You cannot think photochemically in a digital world.

bull****

But you know that and thanks for posting something other than spam
advertising. . . .

On Sat, 19 Apr 2008 02:08:12 -0700 (PDT), Marc Wossner
wrote:


First of all I apologize for the mess in my starting post.
It should of course read like this:

...
Is it correct to pick two points a and b as a) upper limit of the
shadow range at 20 and b) lower limit of the highlight range at 80 and
move a from 20 to 16 (value : 1,2) and b from 80 to 96 (value x 1,2)?
Or can that task be accomplished more easily in the levels dialogs
gamma setting?
I'm asking because according to the diagrams in Hunts 'Reproduction of
color' and 'The Focal Encyclopedia of Photography' this value leads to
perceptually correct reproduction of the subjects tonal values in the
image in the presence of a normal amount of camera flare and viewing
flare.

To Lawrence: Yes I have access to Photoshop and can I reason from your
last statement that my approach is basically correct?

Best regards!
Marc Wossner

Your photo can be technically correct, but aesthetically a failure.
So the question is are you doing art, or documentary photography?

Photojournalism says "don't touch anything"

Art says "anything goes"

Science says "enhance the subject material to display best what you
are trying to illustrate"

So the real answer depends on what your intended audience for the
image is.

Generally speaking, skin tones are the _important_ element, generally
speaking, and you want those to come out correctly, as you remember
them to be. So however you have to adjust the S curve to achieve
that, this is what you want to do, being watchful not to blow out
highlight and shadow detail.

In computer software, this has become pretty easy to achieve. What
the curve ends up looking like depends on the tonal range of the
subject matter to begin with.

Lg


Lg

On Sat, 19 Apr 2008 02:28:59 -0700 (PDT), Marc Wossner
wrote:

On 19 Apr., 04:45, "pickled" wrote:
Digital imaging software does not work like this.
Additonally you are working on your image on a computer monitor that has
entirely different characteristics than a print.You have to tailor the image
to its intended use.
You cannot think photochemically in a digital world.

But don´t some basic principles still hold true no matter if the
reproduction process is analog or digital?

Yes

I mean there still is a certain amount of flare light when you take
the picture and another amount of flare light when you view the print.
Both lower the apparent contrast. And if you want to reproduce the
tones correct you have to consider those effects of flare and raise
the contrast appropriately.

If you mean by *flair* reflections? Specular reflections you ignore,
you know they are going to oversaturate and ignore them. A specular
reflection is just a washed out area, not to be considered in the
final print. Highlight detail is another matter, and that is what you
want to keep an eye on.


Best regards!
Marc Wossner

On Sat, 19 Apr 2008 02:45:37 GMT, "pickled"
wrote:

Digital imaging software does not work like this.
Additonally you are working on your image on a computer monitor that has
entirely different characteristics than a print.You have to tailor the image
to its intended use.
You cannot think photochemically in a digital world.
But you know that and thanks for posting something other than spam
advertising. . . .

What is the medium you want to use? Since my monitor is my final
medium, I have enormous dynamic range. Not so with a print. So yes,
adjustments have to be different between something intended for a
monitor and something intended for printing.

Lawrence Glickman wrote:


Photojournalism says "don't touch anything"



You are old-fashioned!

Photojournalism ... like all journalism ... says "decide what you want
to say and then fake the story and the picture to say it".

Doug MCDonald

On Sat, 19 Apr 2008 14:43:47 -0500, lid wrote:

Lawrence Glickman wrote:


Photojournalism says "don't touch anything"



You are old-fashioned!

Photojournalism ... like all journalism ... says "decide what you want
to say and then fake the story and the picture to say it".

Doug MCDonald

My world of Photography goes back to 1945 and before. If the *rules
have changed,* and I don't doubt that they have, nobody sent me a
letter of notice.

Lg

[A complimentary Cc of this posting was sent to
Marc Wossner
], who wrote in article :
What I focus on now is correct tone reproduction.
With that I mean that what is twice as bright in the subject is also
*perceived* as twice as bright in the image.
To accomplish this, contrast has to be raised to level out the
contrast reducing influence of flare light.

Wrong. Assume that flare light ADDS the same amount of intensity to
every point of the image. To remove the influence, you need to
SUBTRACT the same amount of intensity from every point of the image.

If gamma-corrected (as JPEG), the subtraction becomes (assuming 1% flare)

max(0,x^2.2-0.01)^(1/2.2)

The corresponding curve to use is (assumes fixed-width font):

1 |''''''''''''''''''''''''''''''''''''''''''''''''' '''''''''''_x"
| _x" |
| _x" |
| _x" |
| _x |
| _x" |
| _x" |
| _x" |
| _x" |
| _x" |
| x" |
| _x" |
| _x" |
| _x" |
| _x" |
| _" |
| _x" |
| x" |
| _x" |
| x |
| " |
0 ________",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, ,,,,,,,,,,,,,,
0 1

Hope this helps,
Ilya

[A complimentary Cc of this posting was sent to
Marc Wossner
], who wrote in article :
IHRoYXQgd2l0aCBkaWdpdGFsIHdlIGhhdmUgdGhlIGFkdmFudG FnZSBvZiBiZWVpbmcgdG8NCmFi
bGUgdG8gc3VidHJhY3QgdGhlIHVud2FudGVkIGludGVuc2l0eS BhZGRlZCBieSBmbGFyZSBkaXJl
Y3QgZnJvbSB0aGUNCmltYWdlIHdoZXJlYXMgd2l0aCBhbmFsb2 cgd2UgaGF2ZSB0byByYWlzZSBj

Somewhere inside this mess (DO NOT POST MIME!) is the question where
the gamma correction happens in the flow.

Answer: possibly nowhere. The JPEG from camera is gamma-corrected.
the final image is gamma-corrected. So if intermediate
transformations "know" what is the "current gamma", there is no need
to transform gammas.

Hope this helps,
Ilya

[A complimentary Cc of this posting was sent to
Don Stauffer in Minnesota
], who wrote in article :
On Apr 20, 3:01 pm, Ilya Zakharevich wrote:


Wrong. Assume that flare light ADDS the same amount of intensity to
every point of the image. To remove the influence, you need to
SUBTRACT the same amount of intensity from every point of the image.

Unfortunately, flare seldom adds the same intensity to all points in
an image. Rather, it is usually shaded.

Of course. But since the OP was obviously assuming (unrealistic)
equally-distributed flare, so I considered the same case.

Yours,
Ilya