matt@inuxf.UUCP (Matt Verner) (02/03/88)
I am looking for references that indicate the proper way to convert
'negative' RGB information to 'positive' RGB. In other words if
I take a film negative and scan it into an image using a color
scanner what formula do I use to convert the image into a positive,
properly color balanced, image? The key here is proper color balance.
Do I just subtract from the highest R, G or B value (255 in my case).
or is it more correct to do a gamma correction on each value?
Thanks for any help...
Matt Verner UUCP: ...ihnp4!inuxc!matt
AT&T Graphics Software Labs AT&T: (317) 844-4364
Indianapolis, IN
"The whole point of this sentence is to clearly
explain the point this sentence is making."eao@anumb.UUCP (e.a.olson) (02/10/88)
In article <333@inuxf.UUCP> matt@inuxf.UUCP (Matt Verner) writes: >I am looking for references that indicate the proper way to convert >'negative' RGB information to 'positive' RGB. In other words if >I take a film negative and scan it into an image using a color >scanner what formula do I use to convert the image into a positive, >properly color balanced, image? The key here is proper color balance. > >Do I just subtract from the highest R, G or B value (255 in my case). >or is it more correct to do a gamma correction on each value? > >Thanks for any help... the negative isn't RGB, it's YCM (yellow cyan magenta) - film is a subtractive color process; video is an additive color process. so you'll need some transformation like R = (MAX_COLOR_VAL * 2 - G - B) / 2 # cyan -> red G = (MAX_COLOR_VAL * 2 - R - B) / 2 # magenta -> green B = (MAX_COLOR_VAL * 2 - R - G) / 2 # yellow -> blue Andy Hay +-----------------------------------------------+ ATT-BL Ward Hill MA | Don't try to out-wierd ME, three-eyes! | ihnp4!mvuxq!adh +-----------------------------------------------+
kurtk@tekcae.TEK.COM (Kurt Krueger) (02/16/88)
additional problem that a color negative also contains a mask - that orange color. This is compensated in the printing process by a sequence of filtration- quite heavy filtration to be exact. Filtration is done both to the light source that passes through the film and then through a filter that is in the film itself. These filters (supposedly) are used to correct non-linearities in the negative/film printing process. In order to properly construct a positive image, you will have to model this filtration/printing process, along with these non-linearities. It may be possible to scan the negatives through the appropriate filters and construct a reasonable image. It should be noted that the filtration needs to be precise, and each negative requires a slightly different filtration in order to generate a good positive. Different types of negative film require quite different filtration.
hughes@hrc63.co.uk (Andrew C. Hughes) (02/17/88)
In article <333@inuxf.UUCP>, matt@inuxf.UUCP (Matt Verner) writes: > I am looking for references that indicate the proper way to convert > 'negative' RGB information to 'positive' RGB. In other words if > I take a film negative and scan it into an image using a color > scanner what formula do I use to convert the image into a positive, > properly color balanced, image? The key here is proper color balance. > All photographic negative material has a strong colour cast (usually orange) which would have to be filtered out in your conversion. It is also based on CMY (Cyan, Magenta, Yellow) filtration and so you would need another conversion factor to get RGB values. I believe that an algorithm to do this conversion was recently posted to this group so you might be able to find it. Andrew Hughes GEC-Marconi Research Centre, Chelmsford, UK. UUCP: ..!mcvax!ukc!a.gec-mrc.co.uk!hughes ARPA: hughes%uk.co.gec-mrc.a@nss.cs.ucl.ac.uk
dave@onfcanim.UUCP (Dave Martindale) (03/03/88)
In article <169@anumb.UUCP> adh@anumb.UUCP!mvuxq (a.d.hay) writes: > >the negative isn't RGB, it's YCM (yellow cyan magenta) - >film is a subtractive color process; video is an additive color process. >so you'll need some transformation like >R = (MAX_COLOR_VAL * 2 - G - B) / 2 # cyan -> red >G = (MAX_COLOR_VAL * 2 - R - B) / 2 # magenta -> green >B = (MAX_COLOR_VAL * 2 - R - G) / 2 # yellow -> blue Film is indeed a subtractive process, using cyan, magenta, and yellow dyes, but unless you work in a processing lab you can forget this, and do everything in RGB. The film's three layers are sensitive to RGB. If you measure the density of the three image-forming layers, you do it with RGB filters in the densitometer, and the densitometer filters are labelled RGB, not CMY. If you convert film to electronic form, either with a colour video camera or a B&W sensor plus colour filters, you will have RGB information. Nowhere will you need to convert to a CMY colour space. Even if you are working with printing inks and need real CMY information, you don't use formulae like the ones above - the CMY densities are, more or less, just 1 minus the RGB densities. Where *did* you get the transformation above from? Have you actually used it, or did you just make it up because it seemed like a good idea at the time?