Apache-Talk @lexa.ru 

Inet-Admins @info.east.ru 

Filmscanners @halftone.co.uk 

Security-alerts @yandex-team.ru 

nginx-ru @sysoev.ru 

   


   


   















      :: Filmscanners
Filmscanners mailing list archive (filmscanners@halftone.co.uk)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: filmscanners: Profiling negative films



Ed writes ...

> michael@shaffer.net writes:
>
> > > ...
> >
> >   Just to clarify ... does a simple "3x3 matrix"
> > profile allow for the capture of each R,G & B
> > channel having a different "linearity"?
>
> I don't know what "different linearity" means.  The 3x3 matrix
> transform assumes that each CCD channel's value is linearly
> related to the number of photons that hits it.

        So, there are no other issues which would affect the how well a
"linear" matrix fits the characteristics of a scanner.  I'd wonder if
besides CCD linearity, there'd be other issues like how efficient a
particular dye responds to illumination (linear?), or absorption
effects within the film (linear?).  Granted, either of these
characteristics would most appropriately be associated with the film,
and shouldn't be associated with the scanner's profile ... but I'd
still wonder if there wasn't something else associated with the
scanner which might affect the linearity of each channel(?)

        My curiosity stems from 2 profiles I have ... (1) a 3-D LUT
supposedly associated with the scanner, and (2) a matrix
simplification of this same profile.  My experiments with #2 would
imply it is very close to what Vuescan believes the scanner is capable
of.  Yet, profile #1 implies a much larger gamut(?)  If I am to
believe profile #1 is accurate (I am not confident of this), it would
imply a simplification of it to a matrix-type results in a profile
which is not accurate ... which implies something very different
between a scanner's LUT-type profile and its simplified matrix-type
conjugate.

        I can provide you with both of these profiles if you like.

shAf  :o)




 




Copyright © Lexa Software, 1996-2009.