RE: filmscanners: Vuescan: "device RGB"

["shAf" <michael@shaffer.net>]

Tony writes ...

> On Thu, 29 Mar 2001 21:01:09 -0800  shAf
> (michael@shaffer.net) wrote:
>
> >... so Ektaspace is an easily accepted compromise.
> > Ektaspace is also respected for its editabilty,
> > and its wide gamut is very suitable for
> highbit editing.
> > Its gamut is also sufficiently wide for archiving.
>
> Actually I think you're spot on here, and the space used
> *is* Ektaspace - at least according to my
> overcrowded memory of previous discussions.
> ...

        I'm not inclined to believe it *is* Ektaspace ... leastwise, I
claimed it was "most like" ektaspace, but I did see some differences
... essentially spot on, but reds were perceivably slightly different.
        I am rather inclined to believe it *is* PCD RGB, that is, Bruce
Fraser has implied the two color spaces are very much alike.

        Something else has just occurred to me.  My "test" was based on a
fresh installation of VS7 while my LS-2000 is now put away.  That is,
my test was based on a previously acquired "raw" scan 64bit TIFF, and
VS7 had no way of knowing which scanner scanned it(???!!!)

        This shouldn't change my conclusions regarding the color capacity of
VS RGB, but it does raise the question as to WHEN the scanner
characterization's transform is applied and when VS RGB enters the
picture (so to speak).  Is the "scanned" RGB data truely "raw"? Does
the transform take place only if "device=scanner"? ... and no
transform takes place if "device=disk"??  There being no difference
for selecting "scanner" versus "disk" would only be true if the
scanner transform (scanner_RGB=>VS_RGB) were applied to the "raw"
data.  Maybe it is, but I was under a different impression.

        This is where I miss Ed on this forum  :-(

shAf  :o)