hybrid system the hard way? including film modeling
both empirical (easy way) and mechanistic (hard way) are supported
mathematically by the ICC, but you could do your own system http:/www.color.org empirical characterization entails printing an equipment code value target to the "calibrated" equipment and relating it mathematically as a profile to the color of the profile connection space, usually cubic, a three dimensional profile for 3 colorant mediums, I know there are at least or there once was 4 colorant mediums from Fuji, I'll allow you to derive this from my post yourself, it is not hard if you know it even B&W colorants like silver halides have a hue that must be either maintained or translated in the ICC profile like a three colorant system, the eye is a three colorant system, I will allow you to derive B&W yourself, it is not hard if you know it with the advent of RIMM, ERIMM, and ROMM in ICC you can use digital manipulation for hybrid systems (you can search for these on the ICC site and they are from ProPhoto RGB according to wikipedia) so why would you want to do it the hard way, mechanistically? 1) want to retain analog manipulation methods 2) want to have analog manipulation algorithms within digital 3) want to an analog capture of scene colorimetry 4) multi-stage analog/hybrid systems do not calibrate (steady-state calibration is a prerequisite for profile characterization) 5) want to design new analog equipment or manipulation 6) want to design new sensitized media for a hybrid system ( a film program was around 5 million at Kodak 17 years ago) 7) want to design a better analog RIMM, ERIMM or ROMM so how to do it the hard way? first, you will need a lot of information, the preferred way of getting this method is from analog media, equipment and software companies, as opposed to the investment yourself, some analog technology really requires single layer coatings to resolve crosstalk from spectral sensitivity and chemical processing, chemical processing can be just the way it is, process variability across or inside labs, or by design with things like DIR or DIAR couplers intended to reduce or optimize chemical crosstalk http://en.wikipedia.org/wiki/Color_motion_picture_film Kodak has licensed some hybrid or analog technology to IMAX, the analog industry may be willing to deal at this point, additionally some analog and hybrid information may be patented, when I was at Kodak R&D many things were not patented due to other nations not respecting intellectual property, PhotoCD was patented as a last ditch effort to leverage capture film into digital systems so what type of information will you need? spectral sensitivity of capture mediums (for some systems digital capture sensitivity needs resolved to sensor and filtration) spectrophotometry of print (subtractive) output mediums (spectral data might have to be resolved to light source and filtration) spectroradiometry of display (additive) output mediums spectroradiometry of analog and hybrid printers (for some systems such radiometry of equipment needs resolved to light source and filtration) chemical colorant response to light of medium (DlogE) interimage, overall crosstalk of medium single layer coatings of mediums to resolve chemical versus sensitivity crosstalk what are the use cases? 1) captures (digital, hybrid or analog)(scene or like printing density) 2) manipulations (digital, hybrid or analog) 3) outputs (digital, hybrid or analog) spectral information is a one dimensional look-up table without crosstalk crosstalks are at least a linear matrix DlogE is best represented with a rational quadratic, higher math effects the central linearity, complete linearity effects toe and shoulder, highlight and shadow detail where dynamic range is low, this is still a one dimensional look-up table digital contrast is linear, gamma hybrid input/output contrast is calibrated for gamma in most cases multi-stage systems typically use some standard assumptions, mostly what equipment/software/measurement the systems engineer is working with any mathematician can take it from here to get all use cases if you want me to do a use case, just reply, I have a lot of time on my hands by the way, there is a book about "making" Kodak film, but not "designing" it, maybe the author might want to add a understandable compilation of this to his book http://www.makingkodakfilm.com/ -- Dale |
hybrid system the hard way? including film modeling
On 02/26/2014 12:07 AM, Dale wrote:
so why would you want to do it the hard way, mechanistically? 1) want to retain analog manipulation methods 2) want to have analog manipulation algorithms within digital 3) want to an analog capture of scene colorimetry 4) multi-stage analog/hybrid systems do not calibrate (steady-state calibration is a prerequisite for profile characterization) 5) want to design new analog equipment or manipulation 6) want to design new sensitized media for a hybrid system ( a film program was around 5 million at Kodak 17 years ago) 7) want to design a better analog RIMM, ERIMM or ROMM so how to do it the hard way? Kodak has a lot of things they could license in the area, or at least leverage the last days of film, probably not develop equipment, but maybe software 1) a FORTRAN library of systems modeling 2) a VAX and a portable (POSIX?) software called PLUS to implement the modeling in digital mathematic terms, and hybrid analog manipulation 3) scene quality (not accuracy) QsubA, leading to better scene balance algorithms and "more attractive" images, implemented in consumer scan/print clas35 printer 4) optimization of mathematics around ICC conversion resulting better scenes, manipulation and storage 5) colorimetry patent from Eikonix -- Dale |
hybrid system the hard way? including film modeling
On 02/26/2014 12:07 AM, Dale wrote:
so why would you want to do it the hard way, mechanistically? also might want to maintain multi-stage approvals that might span across labs and sites too -- Dale |
hybrid system the hard way? including film modeling
On 02/26/2014 12:07 AM, Dale wrote:
so what type of information will you need? spectral sensitivity of capture mediums (for some systems digital capture sensitivity needs resolved to sensor and filtration) spectrophotometry of print (subtractive) output mediums (spectral data might have to be resolved to light source and filtration) spectroradiometry of display (additive) output mediums spectroradiometry of analog and hybrid printers (for some systems such radiometry of equipment needs resolved to light source and filtration) chemical colorant response to light of medium (DlogE) interimage, overall crosstalk of medium single layer coatings of mediums to resolve chemical versus sensitivity crosstalk you can translate the spectral data above to CIE chromaticity coordinates or tristimulus values if you prefer rotations and adjustments in an independent space, compatible with ICC and digital imaging -- Dale |
hybrid system the hard way? including film modeling
On 02/26/2014 04:51 PM, Dale wrote:
photons are material chemical particles okay, photons do not have mass maybe the distinction between Physical Chemistry and Particle Physics is a consideration of mass,, maybe there does not have to be a distinction electrons have mass, etc. I got an A+ in both undergraduate Physical Chemistry 1 and 2 my best and most enjoyable classes, and in my opinion a very elegant description of nature, undergraduate though and not mature enough to say whether PChem or Physics needs a distinction without some study, wikipedia is something I really like -- Dale |
hybrid system the hard way? including film modeling
On 02/26/2014 04:51 PM, Dale wrote:
a good thing to start with might be the attempt at co-optimization of C-41 film and RA-4 paper, and the failure thereof even with the assistance of hybrid tools, and this is only a two-stage multiple stage system, don't know if a system paradigm or ideal paradigm was used if this was a consumer pursuit it would have been good enough with scene balance algorithms, perhaps as simple as white point balance no less editing work was achieved with this "as far as I know" I do know they replaced the RA4 co-optimized paper with their hybrid RA4 paper when systems testing showed enough results to simplify things with one product for analog and hybrid -- Dale |
All times are GMT +1. The time now is 12:22 PM. |
Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
PhotoBanter.com