View Single Post
  #1  
Old February 27th 14, 05:08 AM posted to sci.engr.color,sci.image.processing,rec.photo.darkroom,rec.photo.digital,comp.soft-sys.matlab
Dale[_4_]
external usenet poster
 
Posts: 131
Default dynamic range and exposure latitude?

I don't have access to current professional equipment, software, or
media, so I will rely on feedback and past experience, and the criticism
thereof

video sensor and filtration device color space assumptions face a
complexity with other output than CRT (LCD, LED, Plasma, OLED, and I
read better LCD is on the way)

ICC solution to dynamic range is based on ProPhoto RGB according to
wikipedia (there is also Adobe RGB, and Wide Gamut RGB, on wikipedia)
http://en.wikipedia.org/wiki/ProPhoto_RGB
implemented in RIMM,ERIMM, and ROMM
http://www.color.org
http://www.color.org/chardata/rgb/rimmrgb.xalter
http://www.color.org/chardata/rgb/erimmrgb.xalter
http://www.color.org/romm.xalter

ICC does not reveal the assumed film or idealistic/standard film used in
ProPhoto RGB

ProPhotoRGB chromaticity coordinates are linear with CIE chromaticity
coordinates

CIE tone is linear with light (L,Y,B,V) or linear with lightness (L*)

film due to chemical considerations of activity and exhaustion and
design for dynamic range treatments, has a toe and shoulder modeled and
designed using a quadratic as opposed to linear contrast, such as a
rational quadratic which does not bend over the toe or shoulder

since film contrast is not linear, some calculations must be performed
to have it linear

ICC does not reveal such considerations

Kodak designed ProPhoto RGB, probably in light of sRGB workflows being
accepted, as opposed to using an ideal digital camera and an ideal
digital projector as the RIMM,ERIMM, ROMM (additive color used in
displays has more dynamic range than subtractive colors used prints)

one Kodak implementation I know called Premier (a system of scanning
film, manipulating and editing film, and outputting film) used a
linearized film assumption

I believe this linearization was achieved by unbuilding the rationall
quadratic contrast, and some form of interimage of an Ektachrome film of
the time, be it spectral sensitivity interimage or chemical process
interimage, I don't really know which one or if both were used, I heard
it once but I forget

some use cases would prefer color matching as opposed to appearance
matching and would not use RIMM,ERIMM, ROMM, they would use CIE RGB

but most use cases involve a viewing and acceptance of an image, even if
it just a consumer looking at something and saying it is "good enough"

what is ProPhotoRGB? Is it a negative with exposure latitude suiting it
to ERIMM? if so this could complement many hybrid workflows where
transparency film output is used or copied in an analog fashion

in fact you could design a color negative film to be scanned,
manipulated, output, and projected with ICC color management

you unbuild the non-linearity and crosstalk of the film as long as it's
image dyes chromaticity coordinates result in something close enough to
linear with CIE chromaticity coordinates, making it easier to scan on a
scanner whose exposure, filtration and sensitivity are designed to be
linear with CIE chromaticity coordinates

manipulation algorithms relative to CIE are widely available, but you
could add analog editing, and hybrid algorithms to the mix

you can design a projector (exposure, filtration) in a digital
environment, with the right analog and hybrid manipulation
consideration, to display a color negative instead of using a
transparency intermediate

such a projector design leads to an easy film recorder design

same with monitor design ...

is film dead? should you snip most of the above? are professional
digital capture, manipulation and output already exceeding color
negatives for dynamic range and exposure latitude? even considering push
and pull chemical processing?

then why not a digital RIMM,ERIMM and ROMM?

the ICC site says color managed workflows have not taken hold in
digital, I read once, is this because of sRGB(video) and
ProPhotoRGB(film) workflow interference?

are digital RIMM,ERIMM and ROMM to far outside of CIE eye based
considerations? then why not use CIE RGB for the RIMM,ERIMM, and ROMM

probably the best bet considering printed page is not dead and
Perceptual Reference Medium (PRM) can be used in ICC
http://www.color.org/v4_prmg.xalter
is to give customers choices for their use cases of what reference suits
them

things like lighting, surround, flare, viewing angle, measurement
considerations likewise could all be incorporated

a little more education,, maybe too much for "good enough" color,
especially since "more attractive color" exists

some form of device independent "accurate" color and appearance must be
a starting point?

for consumers, white balance might not be the only scene balance that is
used, you could use analog and hybrid models to get to the "more
attractive color" achieved in consumer films like "pop" saturation

if you think I am a troll, give me a entry level imaging systems job in
my area and I will have to keep my ideas to myself, I don't quite
qualify for standards or open systems work, so here I am on usenet,
wikipedia, wiktionary and the web

I have an undergrad honors degree in ChemE with minors in polymer
science and math (forget most of this)
I know a little MATLAB and SAS
I know basic programming on punch cards and punch card readers
I know Fortran data flow programming
I know modular Pascal data flow programming
I know a little C
I know a little object oriented (OO) architecture/design
I know a little C++ OO programming
I know the basics of java systems and OO programming
I just can't get abstract IDEs like netbeans or operating system
required libraries, out of my head to write a complex program
maybe I should spend more time learning java than talking on usenet

--
Dale