RE: Converting to grayscale
RE: Converting to grayscale
- Subject: RE: Converting to grayscale
- From: "Will and Pam" <email@hidden>
- Date: Thu, 21 Jul 2005 00:25:00 -0400
- Importance: Normal
Thanks to everyone for the input. Some details about my project that might
shed some light on what the issues are:
We are scanning a large archive of bw negatives, most of which have never
been printed before. The scans will themselves become an archive that will
serve as the basis for disseminating derivative images both via the web and
via print. The intermediate goal is to get an archival master that can be
multipurposed for various outputs. The hope is to capture the densities of
the negative as accuratately as possible. We are scanning in color for two
reasons: our scanner, though very good in many other ways, does not support
16-bit grayscale scanning--only 16-bit RGB; and, some of the negatives have
condition issues such as staining and discoloration that we wish to document
before converting to grayscale.
My concern about how to convert to grayscale from RGB has to do with getting
the most accurate record of the luminosity characteristics of the negative,
since the black and white negative, when printed, is merely a means of
rendering various levels of gray. It matters then what levels of gray are
recorded in the scan of the negative. If a point in the negative scan reads
R200 G206 B198, then picking one channel by definition means arbitrarily
deciding what gray level that point should be. Furthermore, if the color
shift is not linear, each channel will also present a slightly different
interpretation of the relationship among gray levels as well--one channel's
curve is steeper or shallower than another, or one concave and another
convex and another S-shaped.
It could be argued that it really doesn't matter that much anyway since
inevitably the native scans will be manipulated in photoshop to achieve a
particular tonal appearance, which as some have already said is a matter of
the scanner/imaging technician's skill and vision and understanding of the
work of this photographer.
In the end I'd like to be able to demonstrate that we have done everything
we can do to capture the most accurate possible transcription of the
negative's densities, and that little if anything has been left up to some
arbitrary conversion routine or algorithm. At the same time I have no desire
to mix or calculate channels image by image--there are over 30,000 negatives
in this part of the project and that is just phase 1. Whatever we do for the
conversion, we have to set it up as an Action or we'll never finish the
project.
Will
> -----Original Message-----
> From: Mike Strickler [mailto:email@hidden]
> Sent: Wednesday, July 20, 2005 12:52 AM
> To: email@hidden
> Subject: Re: Converting to grayscale
>
>
> > Yes, so? As long as you have a good scanner, the information in all
> > three channels will be equally usable.
>
> Usable, yes. Identical, no. Simply pulling out the L* channel isn't
> likely
> to yield the best or most predictable result possible in very many
> cases.
>
> No, probably not. I was referring to R, G, and B channels, not L*, a,
> and b. Apparently he has some other issues with the scanner that are
> making things very tough.
>
>
>
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden