Basic question about contrast and gamma

On 2/21/2008 3:33 PM Monica Schulz spake thus:

> I'm just begining to dive into the photographic process, so my
> practical experience is still scarse. - I try to figure out the basics
> in a theoretical way first.

[snip theoretical worrying]

If I might be so bold, it seems to me that if you're just starting out
(at least in b&w photography, let's say), you're waaaay overthinking
this one. Unless you absolutely need to feel comfortable with the
theoretical basis of making negatives and prints, it would be far better
to just go out and start shooting, assuming you have all the stuff you
need. Shoot a bunch of rolls of film. Make contact sheets. Make some
prints. Put them up and look at them.

In other words, make mistakes. The more mistakes you make, the better
(because you'll learn something from each one in a way that no
theoretical discussion can match).


What's that old saying? "My first 10,000 prints were garbage"?
 >> Stay informed about: Basic question about contrast and gamma 

"Monica Schulz" <monica.schulz RemoveThis @gmx.org> wrote in message
news:7ec52008-4faa-4941-95b0-7e9612521257@n75g2000hsh.googlegroups.com...
Hi ng,

I'm just begining to dive into the photographic process, so
my
practical experience is still scarse. - I try to figure out
the basics
in a theoretical way first. In case of the negative/positive
process
(both b&w and color) there is the relationship of contrast
and gamma
and I simply can't get that into my head, no matter how much
I read
and think (maybe I read the wrong papers?!).
One thing is that negative gamma and print gamma should
combine to 1
if the contrast reproduction should match the scene (or
1,1/1,2 if you
take into account that most people prefer a slightly
exaggerated
contrast of the midrange values).
If that´s true and the negative is developed to a kind of
normal gamma
of 0,6, print gamma must be about 1,7. In the case of b&w
contrast
must be balanced between the negative and the print if
contrast is to
be reproduced correct. So if the neg has, say, 1,3 log units
it must
be printed on a paper that is also capable of 1,3 log units.
Not more,
not less. So if you chose the paper grade according to the
neg
contrast, doesn´t that mean that you might have to chose a
grade that
leads to a gamma which doesn´t combine to 1 with the neg?
Another thing that give me headaches in the same context are
color
prints. In the case of the color reproduction chain, the
negative is
normally developed to a gamma of about 0,6 and most papers
have a
gamma of more than 2,5, so the combined gamma is 1,5. - So
how does
that work? And why are color papers not available in
different grades
to balance with the neg contrast?
You see I´m kind of lost with those concepts and I really
hope that
someone could shed some light on them to show me what´s
going on.

Monica Schulz

First of all "gamma" is a measure of contrast but is not
often used now. Gamma is the slope of the straight line
portion of the film or paper characteristic but, since most
characteristics have only small straight line sections, or
sometimes none another measure is used. This is average
contrast or G-bar, the slope of a stright line drawn between
a shadow and a highlight point on the curve. Kodak goes one
step further and uses Contrast Index or CI which is the
average contrast between two specified points on the curve.
Since these are constant CI of different materials is easy
to compare.
You are right that the overall contrast of a print
should be on the order of 1 but reflection prints are
limited in the range of brightness they can reproduce so the
unity contrast is usually true only of a range of mid tones,
the shadows and highlights generally being compressed or
simply going out of the range the paper can reproduce. Since
transparencies don't rely on ambient light their range can
be greater but the target contrast is still 1. However, the
eye tends to judge "correct" reproduction of tones by the
gray tones and, for the most part ignors deep shadows and
bright highlights, so, if the mid-tone contrast is right the
eye will see the print as being right even though the
overall range of brightness that can be reproduced is
substantially less than in the original scene, or, for that
matter, less than what the negative will record.
For many scenes of relatively high contrast a low
contrast print that attempts to reproduce the full range of
tones recorded on the negative will simply look flat.
To some degree this can be compensated for by localized
printing of details, either holding dark areas back to
brighten them or printing in (burning in) highlights, but
this must be done with skill or the manipulation will be
obvious to the viewer.
For black and white work, where there is little
standardization, it is common to develop negatives to a CI
of anywhere from about 0.4 to 0.7 depending somewhat on the
kind of illumination used in printing. The printing light is
important because of something called Callier effect. This
is the effect on effective density caused by the relative
diffusion or specularity of the light source. Because most
B&W films have an image made up of descrete silver
particals, which can scatter light, there is a fairly large
Callier effect. Between a completely diffuse light source,
like a color head or a cold light head, and the usual partly
condensing head, the contrast difference is about one paper
grade, the condenser head having the greater contrast. Note
that Callier effect has no effect on the one reproduction
scale other than overall contrast which can be completely
compensated for by either a change in development time of
the negative or choice of paper contrast. The method of
compensation is a matter of convenience.
Because color films, and here I include chromogenic B&W
film, the image is made up of semi-transparent dye, there is
very little Callier effect so the printing light makes
little or no difference in effective contrast. While there
is some choice in contrast of color printing materials it is
much less than for B&W. This is because the eye is much more
sensitive to variations in the contrast of color. If color
is too low or too hight in contrast the effect is quite
obvious. So, the contrast of color recording and printing
materials is much more highly standardized than for B&W. In
addition to this the three recording layers (some films have
many more) for the three basic colors must have
characteristics which match very closely if the color is to
track for all brightnesses. Anothere way of putting this is
that the neutrality of a gray scale must be constant. For
this reason also there are strict limits on the contrast
range of color materials.
Color negative materials, which are intended to printed
to make postives, have relatively low contrast and can be
controlled to some extent in processing, as long at the gray
scale tracking mentioned above remains acceptable. Color
positive materials are generally meant for direct viewing
and are much higher in contrast, generally slightly more
than 1 to compensate for flare. There ARE some transparency
materials intended for duplication or other printing which
are lower in contrast bu they are rarely encountered by
amateur photographers.
Some color materials intended for direct reversal
printing from positive transparencies offer a lower contrast
grade for use when the usual grade is too contrasty. This
was sometimes the case with Cibachrome/Ilfochrome. Another
method of reducing contrast is masking, the use of a very
low contrast negative image overlying the postive image to
reduce the contrast.
I think I am getting deeper into this than you want to
go.
You can find some elementary material on sensitometry,
the name for the study of tone reproduction from film, on
the Kodak site. Search for H-740. If you can't find it
write me privately and I will send it to you.


--
---
Richard Knoppow
Los Angeles, CA, USA
dickburk RemoveThis @ix.netcom.com
 >> Stay informed about: Basic question about contrast and gamma 

On 22 Feb., 10:57, "Richard Knoppow" <dickb... RemoveThis @ix.netcom.com> wrote:

<snip>

> You are right that the overall contrast of aprint
> should be on the order of 1 but reflection prints are
> limited in the range of brightness they can reproduce so the
> unity contrastis usually true only of a range of mid tones,
> the shadows and highlights generally being compressed or
> simply going out of the range the paper can reproduce.

<snip>

> While there
> is some choice in contrast of colorprinting materials it is
> much less than for B&W. This is because the eye is much more
> sensitive to variations in the contrast of color. If color
> is too low or too hight in contrast the effect is quite
> obvious. So, the contrast of color recording and printing
> materials is much more highly standardized than for B&W.

On the risk of beeing pushing I must still insist on the relationship
between neg contrast and paper contrast.
As we agree that both must balance to about 1 and color negative film
is normally developd to about 0,6, the paper should show gamma of 1,7.
But almost all color printing papers have gamma values far in excess
of 2. So how does the steeply raised contrast of the print fit into
your above statement? Does the contrast have to be raised that much so
offset the unavoidable flare? If that´s true paper contrast cannot be
calculated by simply using the reciprocal of the neg. Maybe that´s all
what blocked me from recognizing the true connection?!

Best regards!
Monica Schulz
 >> Stay informed about: Basic question about contrast and gamma 

In article
<55c237f9-dc18-4ff2-8f02-5a00f838164a.DeleteThis@b29g2000hsa.googlegroups.com>,
Monica Schulz <monica.schulz.DeleteThis@gmx.org> wrote:

> On the risk of beeing pushing I must still insist on the relationship
> between neg contrast and paper contrast.

Neg contrast= 1.45 max
Paper contrast .85 <

Gamma is not a good indicator of true paper capability try .85 or if you
must use film gamma subtract .20 from the bottom and top of the
Characteristic curve your using, use a shoulder around 1.45 or less more
like 1.25


> As we agree that both must balance to about 1 and color negative film
> is normally developd to about 0,6, the paper should show gamma of 1,7.
> But almost all color printing papers have gamma values far in excess
> of 2. So how does the steeply raised contrast of the print fit into
> your above statement? Does the contrast have to be raised that much so
> offset the unavoidable flare? If that´s true paper contrast cannot be
> calculated by simply using the reciprocal of the neg. Maybe that´s all
> what blocked me from recognizing the true connection?!
>
> Best regards!
> Monica Schulz

Flare should not be an issue when using a modern lens if using something
older producing flare you should replace it with a modern Apo lens.

--
Reality is a picture perfected and never looking back.
 >> Stay informed about: Basic question about contrast and gamma 

"Monica Schulz" <monica.schulz.TakeThisOut@gmx.org> wrote in message
news:55c237f9-dc18-4ff2-8f02-5a00f838164a@b29g2000hsa.googlegroups.com...
On 22 Feb., 10:57, "Richard Knoppow"
<dickb....TakeThisOut@ix.netcom.com> wrote:

<snip>

> You are right that the overall contrast of aprint
> should be on the order of 1 but reflection prints are
> limited in the range of brightness they can reproduce so
> the
> unity contrastis usually true only of a range of mid
> tones,
> the shadows and highlights generally being compressed or
> simply going out of the range the paper can reproduce.

<snip>

> While there
> is some choice in contrast of colorprinting materials it
> is
> much less than for B&W. This is because the eye is much
> more
> sensitive to variations in the contrast of color. If color
> is too low or too hight in contrast the effect is quite
> obvious. So, the contrast of color recording and printing
> materials is much more highly standardized than for B&W.

On the risk of beeing pushing I must still insist on the
relationship
between neg contrast and paper contrast.
As we agree that both must balance to about 1 and color
negative film
is normally developd to about 0,6, the paper should show
gamma of 1,7.
But almost all color printing papers have gamma values far
in excess
of 2. So how does the steeply raised contrast of the print
fit into
your above statement? Does the contrast have to be raised
that much so
offset the unavoidable flare? If that´s true paper contrast
cannot be
calculated by simply using the reciprocal of the neg. Maybe
that´s all
what blocked me from recognizing the true connection?!

Best regards!
Monica Schulz

I may be going over territory I covered in private
e-mail but want to answer it here if possible.
First of all I am not sure about your numbers for
negative and paper contrast. 0.6 is probably about right for
B&W negatives but I think color negs are somewhat lower,
perhaps around 0.4. In any case gamma can be misleading
because it applies only to the straight line portion of the
film or paper characteristic. More modern methods of stating
"effective" contrast are to use an average. This is done by
drawing a straight line between the shadow density and
highlight density and stating its slope. Average contrast,
or G bar, does not have a fixed interval between highlights
and shadows, the Kodak method called Conrtast Index does.
G-bar and CI can be the same. In general gamma will not
always agree with either.
The idea is that multiplying negative contrast and
positive contrast will give the contrast of the final print.
Target contrast is usually somewhere around 1.0 but can be
varied for a special effect or to compensate for something
such as flare.
Because the range of brightness that can be reproduced
by a reflection print under normal illumination is, in
general, less than the original scene some compression must
take place. Compression means some part of the image is
reproduced with lower contrast than the original. Since the
eye tends to judge the "correctness" of an image based on
its mid gray values or equivalent in color, the compression
is typically in the shadows and highlights. Since the eye
tends to compress these values in direct viewing the results
are acceptable. If the mid tone contrast is too little, in
an attempt to compress the entire tone scale, the results
will look "flat" even though details are visible in shadows
and highlights that are not in a more "normal" looking
print. If the print is made to have higher contrast than the
original scene it will still not look natural but may be
acceptable if the original scene was very low in contrast.
Since they eye is constantly compensating for brightness
variations our impressions of original scene contast may not
be too accurate. For instance, on a brightly lighted day
with deep shadows the eye will adjust somewhat when looking
into the shadows and see detail there that is not visible
when looking at bright objects, but, because one's attention
is not concentrated in the shadows then we are not generally
aware of it. Film, OTOH, is fixed in its sensitivity. It can
record a wide range of brightness but that range is centered
according to the exposure its given. Then, when a reflection
print is made, there is a further reduction of range since
the reproducing medium is not capable of reproducing the
full range of original scene brighnesses. Note that I am
talking about the presentation to the eye. If one prints a
long scale negative will onto a low contrast paper its
possible to reproduce the full range of tones on the
negative, but, the print will look very flat to the eye
because the original brightness scale will be compressed.
There are times when such prints are desirable, for instance
for scientific or evidentiary purposes but, in general, they
are not acceptable for pictorial purposes.
Where there is not a clear reference for the eye
distortion of contrast is acceptable. This is very true of
B&W but somewhat less for color. If there is no
psychological expectation the eye simply accepts what is
presented.
There is a great deal more to this and I have
oversimplified to some extent but I think the main answer is
that I think the gamma numbers you have are somewhat off.


--
---
Richard Knoppow
Los Angeles, CA, USA
dickburk.TakeThisOut@ix.netcom.com
 >> Stay informed about: Basic question about contrast and gamma 

"____" <internetphobic DeleteThis @deletedmail.com> wrote in message
news:internetphobic-9BE424.14082724022008@newsgroups.comcast.net...
> In article
> <55c237f9-dc18-4ff2-8f02-5a00f838164a DeleteThis @b29g2000hsa.googlegroups.com>,
> Monica Schulz <monica.schulz DeleteThis @gmx.org> wrote:
>
>> On the risk of beeing pushing I must still insist on the
>> relationship
>> between neg contrast and paper contrast.
>
> Neg contrast= 1.45 max
> Paper contrast .85 <
>
> Gamma is not a good indicator of true paper capability try
> .85 or if you
> must use film gamma subtract .20 from the bottom and top
> of the
> Characteristic curve your using, use a shoulder around
> 1.45 or less more
> like 1.25
>

Gamma is a measure of the slope of exposure vs: density
of the straight line portion of a film or paper curve. Since
nearly all film or paper characteristics are curved and have
either small or no straight line portions gamma can be
misleading about overall contrast. More modern measures are
average contrast or G-bar and Contrast Index or CI. These
two are very similar but CI has a specified interval of log
exposure.
In any case, whether one is using gamma or G-bar or CI
the numbers combine by multiplying, not addition or
subtraction. That is, a gamma of 0.5 will combine with a
gamma of 2.0 to result in a gamma of 1.0. A gamma of 1.0
means that the contrast is the same as the original subject.

>> As we agree that both must balance to about 1 and color
>> negative film
>> is normally developd to about 0,6, the paper should show
>> gamma of 1,7.
>> But almost all color printing papers have gamma values
>> far in excess
>> of 2. So how does the steeply raised contrast of the
>> print fit into
>> your above statement? Does the contrast have to be raised
>> that much so
>> offset the unavoidable flare? If that´s true paper
>> contrast cannot be
>> calculated by simply using the reciprocal of the neg.
>> Maybe that´s all
>> what blocked me from recognizing the true connection?!
>>
>> Best regards!
>> Monica Schulz
>
> Flare should not be an issue when using a modern lens if
> using something
> older producing flare you should replace it with a modern
> Apo lens.
>
Flare comes from other sources than just lenses, for
instance light scattered in the optical path of a projector.
Also, ambient light in a room where slides are being shown
has an effect similar to flare. Projection transparencies
and motion picture prints are often slighly higher than a
gamma of 1.0 to compensate a bit.

Another note: The prefix APO on most modern lenses in
meaningless. It is permitted in German made lenses because
of an error in the German DIN standard which mis-represents
the term "apochromatic" and allows its use for lenses which
are not so corrected. I think this has been changed but the
terminology continues.
A true apochromat is corrected to bring three colors to
a common focus. The curve of longitudinal chrmomatic
aberration will be S shaped crossing the zero line three
times. Most lenses are achromats, that is they are corrected
for two colors. The curve for an achromat will cross the
zero line twice. More important than the number of zero
crossings is the deviation from focus between the corrected
points: the less the deviation the better the lens is
corrected. Its possible to have an achromat that has better
overall correction than an apochromat. In any case, nearly
all of the current crop of lenses with APO as a prefix are
NOT apochromats but are simply well corrected achromats. I
think the term APO used for such lenses is misleading and
should be eliminated.
In any case, an APO lens, regardless of its true color
correction, has no less flare than any other. Flare is
mainly a function of the number of glass-air surfaces but
other things can contribute such as reflections from the
inside of the lens mount. Surface reflections from the glass
can be controlled by anti-reflection coating. Such coatings
have been known for over a century but were made
economically practical only after WW-2. Modern coatings are
complex and are effective over the entire range of colors of
interest in photography. Modern complex lenses, especially
zoom lenses, would be impossible without advanced coating
techniques.
Even single coated lenses have substantial reduction in
flare and modern multiple coated lenses are nearly flare
free. Probably the use of a modern lens would pretty much
eliminate having to increase print gamma to compensate for
flare but one might still want to compensate for other
factors such as the viewing conditions.



--
---
Richard Knoppow
Los Angeles, CA, USA
dickburk DeleteThis @ix.netcom.com
 >> Stay informed about: Basic question about contrast and gamma 

On 24 Feb., 23:29, "Richard Knoppow" <dickb....DeleteThis@ix.netcom.com> wrote:

> Gamma is a measure of the slope of exposure vs: density
> of the straight line portion of a film or paper curve. Since
> nearly all film or paper characteristics are curved and have
> either small or no straight line portions gamma can be
> misleading about overall contrast. More modern measures are
> average contrast or G-bar and Contrast Index or CI. These
> two are very similar but CI has a specified interval of log
> exposure.
>      In any case, whether one is using gamma or G-bar or CI
> the numbers combine by multiplying, not addition or
> subtraction. That is, a gamma of 0.5 will combine with a
> gamma of 2.0 to result in a gamma of 1.0. A gamma of 1.0
> means that the contrast is the same as the original subject.

Uuh, now I see what you mean. You talked about the difference between
gamma and contrast index before but I didn´t realize it´s importance -
sorry for the delay! So using the contrast index method leads to other
figures for the negative and the paper and they combine to about 1. So
I finally got rid of this problem I had with the curves of these
materials. - Thank you!
Now there is only one point left from my starting question: Why is
there not such a wide variaty of paper grades in color reproduction as
there is in b&w? As far as I understand, in b&w we need this wide
variaty to make the dark tones of a low contrast negative as black as
the dark ones of a high contrast neg. What´s thze difference in color?
Is it because a high density there really means a more saturated color
and not so much a darker color?

Best!
Monica Schulz
 >> Stay informed about: Basic question about contrast and gamma 

Maybe I did not make myself clear enough in my last post. What I mean
is that there is no need for such a wide variaty of paper grades in
color printing as there is in b&w because contrast can´t be
manipulated that much in color as in b&w. I came thinking about that
after reading another old thread so I will use the important words of
the relevant post.
To adjust contrast means to make the image brighter or darker. The
only way to do that is to make the dyes thinner. This lets more light
reflect off the paper. But as a dye gets, say, thinner it removes less
of its anti-color until it finally disappears and no anti-color is
removed. And the other way around if the dye gets thicker. So the
scale doesn´t go from, say, a dark magenta to a light magenta but from
dark magenta to white. Or from a saturated magenta to an unsaturated
magenta.
The digital process can compensate for this unwanted increase or
decrease in saturation if contrast is adjusted in the luminosity
channel of lab-mode. In this case neither hue nor saturation is beeing
changed. There is a nice little example of that on
http://www.cambridgeincolour.com/tutorials/photoshop-curves.htm. But
we can´t do the same in analog color printing.
It could well be that this is complete garbage as far as the
subtractive mixing of the analog process is concerned and if so I´d be
happy if someone tells me so that I can get this idea out of my head
and can think in another direction.

Best regards!
Monica Schulz
 >> Stay informed about: Basic question about contrast and gamma 

"Monica Schulz" <monica.schulz RemoveThis @gmx.org> wrote in message
news:1c545b43-672d-4790-9803-c00e025cb69a@e10g2000prf.googlegroups.com...
Maybe I did not make myself clear enough in my last post.
What I mean
is that there is no need for such a wide variaty of paper
grades in
color printing as there is in b&w because contrast can´t be
manipulated that much in color as in b&w. I came thinking
about that
after reading another old thread so I will use the important
words of
the relevant post.
To adjust contrast means to make the image brighter or
darker. The
only way to do that is to make the dyes thinner. This lets
more light
reflect off the paper. But as a dye gets, say, thinner it
removes less
of its anti-color until it finally disappears and no
anti-color is
removed. And the other way around if the dye gets thicker.
So the
scale doesn´t go from, say, a dark magenta to a light
magenta but from
dark magenta to white. Or from a saturated magenta to an
unsaturated
magenta.
The digital process can compensate for this unwanted
increase or
decrease in saturation if contrast is adjusted in the
luminosity
channel of lab-mode. In this case neither hue nor saturation
is beeing
changed. There is a nice little example of that on
http://www.cambridgeincolour.com/tutorials/photoshop-curves.htm.
But
we can´t do the same in analog color printing.
It could well be that this is complete garbage as far as the
subtractive mixing of the analog process is concerned and if
so I´d be
happy if someone tells me so that I can get this idea out of
my head
and can think in another direction.

Best regards!
Monica Schulz


Something very like the "curves" effect can be done in
chemical photography be means of elaborate masking and by
using such techniques as coltrolled flashing. Both
techniques were widely used in making the plates for
four-color printing. However, its very much easier to do it
in Photoshop and one can see immediately the results. Most
of the tailoring done in Photoshop can be done in analogue
photography but can be very complex to do there.
Contrast and brightness are different. This is
especially true of most reproduction media where the maximum
brightness is determined by illumination. What adjusting
contrast does is to fix the points that are the minimum and
maximum brightness of the image. As contrast is increased
something must be lost. That is where the "brightness"
adjustment comes in. That controlls what is to be the
midpoint of the gray scale. Its possible to extend the range
of tones that will reproduce without falling off one end or
the other by shaping the transfer curve and that is what
Photoshop is doing. Many films and papers have curves which
affect the reproduction of gray tones because they are not
straight lines. Similar curves are available in color films
and papers. This one of the things that determines the
difference between "portrait" material and commercial
material. The overall conrast may be the same but the
placement of reproduced tones in relation to the originals
is different.


--
---
Richard Knoppow
Los Angeles, CA, USA
dickburk RemoveThis @ix.netcom.com
 >> Stay informed about: Basic question about contrast and gamma