toe and shoulder applied to digital

[Dale]

sensitized photographic film and paper have a toe and shoulder in their
contrast, it is an artifact of chemical activity and exhaustion that
just so happens to provide a system to deal with dynamic range, such as
highlights and shadows when such chemistry is optimized, and perhaps
even before optimized which I don't know

the contrast of chemical media, is modeled with rational quadratic and
other non-linear but not cubic terms, leaving it to remain within
measurement and hybrid systems modeling/engineering ease and simplicity

pure digital has gamma, a linear contrast

gamma works just fine when the capture, manipulation and output can
handle the dynamic range of a scene or artistically created image in a
wide enough gamut working space

what I glean from the ICC website http"//www.color.org is that output
profiles are mapped to the maximal machine code for that device, this
means additive systems will have the brightest display, and subtractive
systems will have the darkest display, regardless of accuracy, perhaps
leaving a linear toe and shoulder from the accurate gamut to the machine
maximals

this is really not a capture issue, but it is an output and manipulation
issue

if you used a rational quadratic like Kodak uses for its film and paper
contrasts from the region of accurate to machine maximal color, you
might deal with highlights and shadows better, it involves a
non-linearity in calculation but still a one dimensional look-up table,
probably not as applicable to embedded systems which I don't know

in the case of ROMM, RIMM, ERIMM, I can see why you might not want to do
this to maintain linearity to CIE standards and the end points are not
mapped like output devices as far as I know, except with the
linearization to ProPhoto RGB which I will address in another post
search for ROMM, RIMM, ERIMM and ProPhoto RGB
on http://www.color.org
and www.wikipedia.org

my means are much limited compared to when I worked in Kodak R&D 17
years ago, I can study stuff, but my applications are consumer capture
and display pretty much, color manipulation is maturing in GIMP which I
no longer have since Redhat uses a version predated color management,
and my cell phone is not a smart phone

I see clipping of shadows in consumer capture when printed, how printed
I don't know
I see clipping of highlights on at least television display (I have
CRT),, haven't really viewed my LCD computer monitor enough to say there

sRGB has failed the consumer use case it wanted, at the expense of some
long term development of open systems of color, ProPhoto RGB delivers a
centralized system,, but in gamut is not big enough to accommodate
digital capture, impending digital display and probably digital
projection as I gather, maybe a filter set and associated spectral
considerations of sensitivity or density would be a better ROMM, RIMM,
ERIMM

[Floyd L. Davidson]

Post by Dale
what I glean from the ICC website http"//www.color.org is that output
profiles are mapped to the maximal machine code for that device, this
means additive systems will have the brightest display, and subtractive
systems will have the darkest display, regardless of accuracy, perhaps
leaving a linear toe and shoulder from the accurate gamut to the machine
maximals

This reminds me of a group of very talented old geezers I knew a
few decades ago. They were diesel overhaul mechanics who worked
on power plants at communications sites. Hence they were often
around young telecom technicians, most of whom thought they were
pretty significant compared to a mechanic. These guys came up
with a few statements that used some of the often heard "key
words" comm techs might use. When the mechanics met a tech that
they thought might just be a know it all phoney, they'd rattle
off a statement just like the one above. Lots of words but with
absolutely no sane meaning.

If the tech acted as if he understood them, they had one. If he
was polite and didn't laugh, that was okay. The best response
was someone who broke up laughing and asked who'd taught them to
say that.

[Dale]

Post by Floyd L. Davidson

Post by Dale
what I glean from the ICC website http"//www.color.org is that output
profiles are mapped to the maximal machine code for that device, this
means additive systems will have the brightest display, and subtractive
systems will have the darkest display, regardless of accuracy, perhaps
leaving a linear toe and shoulder from the accurate gamut to the machine
maximals

This reminds me of a group of very talented old geezers I knew a
few decades ago. They were diesel overhaul mechanics who worked
on power plants at communications sites. Hence they were often
around young telecom technicians, most of whom thought they were
pretty significant compared to a mechanic. These guys came up
with a few statements that used some of the often heard "key
words" comm techs might use. When the mechanics met a tech that
they thought might just be a know it all phoney, they'd rattle
off a statement just like the one above. Lots of words but with
absolutely no sane meaning.
If the tech acted as if he understood them, they had one. If he
was polite and didn't laugh, that was okay. The best response
was someone who broke up laughing and asked who'd taught them to
say that.

I was a systems engineer who was allowed to dabble in science every once
in awhile, so yes, I know the buzz words, and had the practical
application of systems experience to know a little about applying them,
since I only worked 9 years I didn't really become an expert systems
engineer, but I think an expert knows what I am getting at

my experience at Kodak systems R&D was

Imagining Science Training Program
1) first rotation, consumer paper test engineer, 1 year
2) second rotation, consumer C-41 process engineer, I did the beginnings
of Flexicolor Stabilizer LF, 1 year
25% of time in classes and a nice little library of books

3) first permanent job, Pro C-41 Process, had some interesting stuff
like process variability of new professional films, and a little work on
C-41 Q-Lab idea, 2 years

4) second permanent job, Professional Hybrid Systems Integration
(I'd like to think I was a little integral in color management), 5 years

would have been an interesting job but I was young and caught up in the
rat race, in retrospect I could have enjoyed life a lot more then, or at
least been financially better off

I was a part time Real Estate agent for awhile, I had a two family home
and an apartment I rented

I almost took a job as a systems engineer on Kodak's Premier PC, a
competition to Photoshop, but the funding was not there, and there was
overall workflow solution additionally needed

found my teenage loves on the streets and crashed and burned, maybe if
Rochester wasn't "smug town" thing could have been different

pecking orders other than performance, if you want to make a new Kodak
stop this

[Kevin McMurtrie]

High-end digital cameras do not use a simple formula for mapping the CCD
dump into RGB. Each pixel produces around 2^14 levels at ISO 100. When
using a very clear lens, a large amount of data beyond the normal black
point and white point is captured. The tonal curves can be adjusted and
viewed in real-time when post-processing RAW files.

Post by Dale
sensitized photographic film and paper have a toe and shoulder in their
contrast, it is an artifact of chemical activity and exhaustion that
just so happens to provide a system to deal with dynamic range, such as
highlights and shadows when such chemistry is optimized, and perhaps
even before optimized which I don't know
the contrast of chemical media, is modeled with rational quadratic and
other non-linear but not cubic terms, leaving it to remain within
measurement and hybrid systems modeling/engineering ease and simplicity
pure digital has gamma, a linear contrast
gamma works just fine when the capture, manipulation and output can
handle the dynamic range of a scene or artistically created image in a
wide enough gamut working space
what I glean from the ICC website http"//www.color.org is that output
profiles are mapped to the maximal machine code for that device, this
means additive systems will have the brightest display, and subtractive
systems will have the darkest display, regardless of accuracy, perhaps
leaving a linear toe and shoulder from the accurate gamut to the machine
maximals
this is really not a capture issue, but it is an output and manipulation
issue
if you used a rational quadratic like Kodak uses for its film and paper
contrasts from the region of accurate to machine maximal color, you
might deal with highlights and shadows better, it involves a
non-linearity in calculation but still a one dimensional look-up table,
probably not as applicable to embedded systems which I don't know
in the case of ROMM, RIMM, ERIMM, I can see why you might not want to do
this to maintain linearity to CIE standards and the end points are not
mapped like output devices as far as I know, except with the
linearization to ProPhoto RGB which I will address in another post
search for ROMM, RIMM, ERIMM and ProPhoto RGB
on http://www.color.org
and www.wikipedia.org
my means are much limited compared to when I worked in Kodak R&D 17
years ago, I can study stuff, but my applications are consumer capture
and display pretty much, color manipulation is maturing in GIMP which I
no longer have since Redhat uses a version predated color management,
and my cell phone is not a smart phone
I see clipping of shadows in consumer capture when printed, how printed
I don't know
I see clipping of highlights on at least television display (I have
CRT),, haven't really viewed my LCD computer monitor enough to say there
sRGB has failed the consumer use case it wanted, at the expense of some
long term development of open systems of color, ProPhoto RGB delivers a
centralized system,, but in gamut is not big enough to accommodate
digital capture, impending digital display and probably digital
projection as I gather, maybe a filter set and associated spectral
considerations of sensitivity or density would be a better ROMM, RIMM,
ERIMM

[Dale]

Post by Kevin McMurtrie
High-end digital cameras do not use a simple formula for mapping the CCD
dump into RGB. Each pixel produces around 2^14 levels at ISO 100. When
using a very clear lens, a large amount of data beyond the normal black
point and white point is captured. The tonal curves can be adjusted and
viewed in real-time when post-processing RAW files.

never really got the chance to have an expert eye to do such things
well, I like a curve I can import

[Savageduck]

Post by Kevin McMurtrie
High-end digital cameras do not use a simple formula for mapping the CCD
dump into RGB. Each pixel produces around 2^14 levels at ISO 100. When
using a very clear lens, a large amount of data beyond the normal black
point and white point is captured. The tonal curves can be adjusted and
viewed in real-time when post-processing RAW files.

You understand that Dale is very much a troll, with very little to
offer the World, don't you?
<<< Bait Snipped >>>

[PeterN]

Post by Kevin McMurtrie
High-end digital cameras do not use a simple formula for mapping the CCD
dump into RGB. Each pixel produces around 2^14 levels at ISO 100. When
using a very clear lens, a large amount of data beyond the normal black
point and white point is captured. The tonal curves can be adjusted and
viewed in real-time when post-processing RAW files.

You understand that Dale is very much a troll, with very little to offer
the World, don't you?
<<< Bait Snipped >>>

He's one of the few in my killfile.