Sign in with
Sign up | Sign in
Your question

Color negative to digital pic

Last response: in Digital Camera
Share
Anonymous
January 7, 2005 4:06:00 PM

Archived from groups: rec.photo.digital (More info?)

Hi.
Has anyone figured a routine for turning scanned color negatives into
digital images??
NOT using Photoshop.
How to get rid of the red cast while maintaining the negative colors???
Thanks, G.
January 7, 2005 4:06:01 PM

Archived from groups: rec.photo.digital (More info?)

"g n p" <gpaleo@ath.forthnet.gr> wrote in message
news:1105095960.752034@athnrd02...
> Hi.
> Has anyone figured a routine for turning scanned color negatives into
> digital images??
> NOT using Photoshop.
> How to get rid of the red cast while maintaining the negative colors???

Paint Shop Pro?
Anonymous
January 7, 2005 10:47:46 PM

Archived from groups: rec.photo.digital (More info?)

If your scanner can scan transparencies it should have a choice for color
negative film.
If you scanner is supported by a 3rd party driver, such as Vuescan, that
program will automatically remove the orange mask.
Related resources
Anonymous
January 8, 2005 1:55:38 AM

Archived from groups: rec.photo.digital (More info?)

On Fri, 07 Jan 2005 11:06:00 +0000, g n p wrote:

> Hi.
> Has anyone figured a routine for turning scanned color negatives into
> digital images??
> NOT using Photoshop.
> How to get rid of the red cast while maintaining the negative colors???

Do you mean to keep the image as a negative but remove the brown
background normal for all negatives ? If that is so, all you need is a
image manipulating program, Gimp is free, with colour-balance &/or
hue-saturation manipulation facility. However, most scanners would
automatically convert the image into a positive (print-like) one. You may
have to fool the scanner by manually setting it to transparency mode to
get a negative image.


--

Gautam Majumdar

Please send e-mails to gmajumdar@freeuk.com
Anonymous
January 8, 2005 4:03:24 AM

Archived from groups: rec.photo.digital (More info?)

anyone4tennis@hotmail.com wrote:

> Hi.
> Has anyone figured a routine for turning scanned color negatives into
> digital images??
> NOT using Photoshop.
> How to get rid of the red cast while maintaining the negative colors???
> Thanks, G.
>
>
Film scanners will eliminate the red cast and produce a Photoshopable image.

You are using a flatbed scanner to scan without using the software designed
for negatives.

Use the flatbed for scanning positives only. A flatbed will produce images
inferior to a film scanner.

Moral - use the right tools for the job!
Anonymous
January 8, 2005 4:15:43 PM

Archived from groups: rec.photo.digital (More info?)

Nope, doesn't work that way for some rason, that's what I tried at first.
Thanks for all the replies, no solution yet 8-((


"Gautam Majumdar" <gmajumdar@XSPAMfreeuk.com> wrote in message
news:KrEDd.22285$C8.13995@fe3.news.blueyonder.co.uk...
>
> On Fri, 07 Jan 2005 11:06:00 +0000, g n p wrote:
>
>> Hi.
>> Has anyone figured a routine for turning scanned color negatives into
>> digital images??
>> NOT using Photoshop.
>> How to get rid of the red cast while maintaining the negative colors???
>
> Do you mean to keep the image as a negative but remove the brown
> background normal for all negatives ? If that is so, all you need is a
> image manipulating program, Gimp is free, with colour-balance &/or
> hue-saturation manipulation facility. However, most scanners would
> automatically convert the image into a positive (print-like) one. You may
> have to fool the scanner by manually setting it to transparency mode to
> get a negative image.
>
>
> --
>
> Gautam Majumdar
>
> Please send e-mails to gmajumdar@freeuk.com
Anonymous
January 9, 2005 1:25:12 PM

Archived from groups: rec.photo.digital (More info?)

On Sat, 8 Jan 2005 13:15:43 +0200, "g n p" <gpaleo@ath.forthnet.gr> wrote:

>Nope, doesn't work that way for some rason, that's what I tried at first.
>Thanks for all the replies, no solution yet 8-((
>
Give us some more info... What process and equipment are you using for doing
your negative scanning? Are you using a flatbed scanner? A film scanner? What
software are you using? Vuescan? What came with the scanner? Nothing?

I scan negatives all the time with my Nikon scanner and they come out just fine.
I've used the Nikon scan software and Vuescan and they both provide settings
specifically for negatives...
Anonymous
January 10, 2005 12:13:35 AM

Archived from groups: rec.photo.digital (More info?)

On Sun, 09 Jan 2005 16:25:12 +0000, dperez wrote:

> On Sat, 8 Jan 2005 13:15:43 +0200, "g n p" <gpaleo@ath.forthnet.gr>
> wrote:
>
>>Nope, doesn't work that way for some rason, that's what I tried at
>>first. Thanks for all the replies, no solution yet 8-((
>>
> Give us some more info... What process and equipment are you using for
> doing your negative scanning? Are you using a flatbed scanner? A film
> scanner? What software are you using? Vuescan? What came with the
> scanner? Nothing?
>
> I scan negatives all the time with my Nikon scanner and they come out
> just fine. I've used the Nikon scan software and Vuescan and they both
> provide settings specifically for negatives...

I am not really sure what problem OP is having. When I scan a negative
either in a flatbed Epson scanner or a PF3650Pro3 film scanner I get a
positive (print-like) image. I can set the PF as if I am scanning slides &
it produces images that look like negatives - brown background and all.
It is possible to colour correct it with image editing programs though I
have never tried it. However, I cannot fool the Epson scanner - it has a
non-overridable autodetect system and negatives are always scanned as
negatives - i.e producing positive images. OP may have this same problem.

--

Gautam Majumdar

Please send e-mails to gmajumdar@freeuk.com
Anonymous
January 10, 2005 4:16:27 AM

Archived from groups: rec.photo.digital (More info?)

It certainly sounds like you are trying to either flatbed scan, or
digitally photograph, a negative, and then hoping you can retrieve a
good color image from it.

Not likely to work. Any device that doesn't understand what a color
negative is all about, will simply do a small white balance on it, or
none at all. Which will leave you with the !real! color information
squashed up horribly into a very small range of numbers. Once you then
correct away the orange/whatever cast, there is not enough data left in
the 3 channels for a useful image. Your imaging device must be driven
in such a way that it subtracts the mask !before! giving you the data.
Doing a scan in 16-bit mode might help a bit, but you would have to
stay in 16-bit mode for all processing and it sounds like you are doing
this on a shoestring, so that's not likely to happen.


Take the negative to your nearest photo shop, and ask them to scan it
onto a CD, for heaven's sake.
Anonymous
January 10, 2005 4:23:50 AM

Archived from groups: rec.photo.digital (More info?)

Oh, and if all you really want is to get rid of the cast and see a
negative image... well, what I said above still applies, but almost
every image browsing or editing program I have ever seen has an
'invert' function. So if your scanner automatically gives you a
positive, just go into your image editor./browser and invert it.. You
are simply undoing what the scanner did, and will get the exact same
result.
Anonymous
January 10, 2005 2:58:59 PM

Archived from groups: rec.photo.digital (More info?)

Pretty close to what's actually happening...
My wife photographs using film (ugh...;-)) ) and I thought I might use my
plain vanilla (but good quality) flatbed to scan some negatives and then
have a go at them removing the color cast-negativing-white
balancing-etc...etc... (I'm ~fairly~ adept at that sort of thing).
Well, IT DOES NOT WORK THAT WAY!!! The orange cast will NOT go away and if
it does, pretty much nothing remains!! Also, the negative of the negative is
just a horrible cyan mess!!. It's as if the negative has only the orange (or
cyan) component on it. My gut feeling is that one must work in the CMYK
color space and not the RGB, seeing that the negative is a "transmission"
object and not a "reflection" one like a picture, using, of course,
different procedures.
I've Googled around for this type of software (color negative -to- positive
picture) but nope. Apparently only done at the scanner, if the scanner is so
equipped.
Thanks again, G



<chrlz@go.com> wrote in message
news:1105348586.961999.165770@c13g2000cwb.googlegroups.com...
>
> It certainly sounds like you are trying to either flatbed scan, or
> digitally photograph, a negative, and then hoping you can retrieve a
> good color image from it.
>
> Not likely to work. Any device that doesn't understand what a color
> negative is all about, will simply do a small white balance on it, or
> none at all. Which will leave you with the !real! color information
> squashed up horribly into a very small range of numbers. Once you then
> correct away the orange/whatever cast, there is not enough data left in
> the 3 channels for a useful image. Your imaging device must be driven
> in such a way that it subtracts the mask !before! giving you the data.
> Doing a scan in 16-bit mode might help a bit, but you would have to
> stay in 16-bit mode for all processing and it sounds like you are doing
> this on a shoestring, so that's not likely to happen.
>
>
> Take the negative to your nearest photo shop, and ask them to scan it
> onto a CD, for heaven's sake.
>
Anonymous
January 10, 2005 9:18:51 PM

Archived from groups: rec.photo.digital (More info?)

chrlz@go.com writes:
>Oh, and if all you really want is to get rid of the cast and see a
>negative image... well, what I said above still applies, but almost
>every image browsing or editing program I have ever seen has an
>'invert' function. So if your scanner automatically gives you a
>positive, just go into your image editor./browser and invert it.. You
>are simply undoing what the scanner did, and will get the exact same
>result.

That's not true.

The "invert" function in photo editors is a linear inversion, generally
just

out = 255 - in

for 8-bit channel data.

A proper conversion from scanned negative to positive has to take into
account the actual response of the negative to light. In its simplest
form, which assumes a "straight line" response of the negative, the
conversion looks like

out = (in / 255) ^ (-1/gamma) * K

where "gamma" is the negative gamma of about 0.55, and "K" is a value
used to set the colour balance and overall brightness of the result.

If you want, graph the two functions above. They are quite different,
and produce different "positive" images.

More sophisticated negative to positive conversion would consider the
fact that the red, green, and blue gammas are different. Or it would
use a model of the actual shape of the film response curve, rather than
assuming it's a straight line in DlogE space.

Dave
Anonymous
January 10, 2005 9:26:44 PM

Archived from groups: rec.photo.digital (More info?)

"g n p" <gpaleo@ath.forthnet.gr> writes:
>Pretty close to what's actually happening...
>My wife photographs using film (ugh...;-)) ) and I thought I might use my
>plain vanilla (but good quality) flatbed to scan some negatives and then
>have a go at them removing the color cast-negativing-white
>balancing-etc...etc... (I'm ~fairly~ adept at that sort of thing).
>Well, IT DOES NOT WORK THAT WAY!!! The orange cast will NOT go away and if
>it does, pretty much nothing remains!!

In order to get reasonable RGB values from scanning a negative, you have
to get rid of the orange cast *before* the image is digitized. On a
flatbed, you may have no choice but to use cyan filters (pretty much the
same way colour enlargers filter the light source to eliminate the
mask colour). If you were building a negative scanner from scratch,
you'd also be able to give red, green, and blue different exposure times
to balance colour. But, either way, you want the RGB histograms *as
they come from the scanner* to fill most of the available code range.
Right now, you probably have almost no blue data.

>Also, the negative of the negative is
>just a horrible cyan mess!!. It's as if the negative has only the orange (or
>cyan) component on it.

If you convert from negative to positive without getting rid of the
orange cast, of course it will be cyan. Also, negative to positive
conversion is *not* correctly done by the "invert" function in an
editor; something more complex is needed (see my other post in this
subject thread).

>My gut feeling is that one must work in the CMYK
>color space and not the RGB, seeing that the negative is a "transmission"
>object and not a "reflection" one like a picture, using, of course,
>different procedures.

No, RGB works just fine. You just need to do the right operations in
the right order.

I've written software to digitize movie negative film, and the
resulting positive images looked fine. (No, I can't send you a copy of
the software; someone else owns it. But I did just describe the basics
in two messages in these two postings).

Dave
Anonymous
January 10, 2005 11:16:51 PM

Archived from groups: rec.photo.digital (More info?)

"g n p" <gpaleo@ath.forthnet.gr> wrote in message
news:1105351137.720402@athnrd02...
> Pretty close to what's actually happening...
> My wife photographs using film (ugh...;-)) ) and I thought I might use my
> plain vanilla (but good quality) flatbed to scan some negatives and then
> have a go at them removing the color cast-negativing-white
> balancing-etc...etc... (I'm ~fairly~ adept at that sort of thing).
> Well, IT DOES NOT WORK THAT WAY!!! The orange cast will NOT go away and if
> it does, pretty much nothing remains!! Also, the negative of the negative
is
> just a horrible cyan mess!!. It's as if the negative has only the orange
(or
> cyan) component on it. My gut feeling is that one must work in the CMYK
> color space and not the RGB, seeing that the negative is a "transmission"
> object and not a "reflection" one like a picture, using, of course,
> different procedures.
> I've Googled around for this type of software (color negative -to-
positive
> picture) but nope. Apparently only done at the scanner, if the scanner is
so
> equipped.

You might try VueScan. There's a trial version.

http://www.hamrick.com/

David J. Littleboy
Tokyo, Japan
Anonymous
January 10, 2005 11:16:52 PM

Archived from groups: rec.photo.digital (More info?)

Did.
Nope 8-((



"David J. Littleboy" <davidjl@gol.com> wrote in message
news:crto7b$rgb$1@nnrp.gol.com...
>
>........................................................................
>
> You might try VueScan. There's a trial version.
>
> http://www.hamrick.com/
>
> David J. Littleboy
> Tokyo, Japan
>
>
>
Anonymous
January 10, 2005 11:53:46 PM

Archived from groups: rec.photo.digital (More info?)

Excellent and informative posts (both).
Thank you very much.
Really a (surprisingly) fairly complex process.
Thanks again, I'll go buy a scanner with color negative capability 8-((
________________
G. Paleologopoulos


"Dave Martindale" <davem@cs.ubc.ca> wrote in message
news:cruhd4$5uc$2@mughi.cs.ubc.ca...
>
> "g n p" <gpaleo@ath.forthnet.gr> writes:
>>Pretty close to what's actually happening...
>>My wife photographs using film (ugh...;-)) ) and I thought I might use my
>>plain vanilla (but good quality) flatbed to scan some negatives and then
>>have a go at them removing the color cast-negativing-white
>>balancing-etc...etc... (I'm ~fairly~ adept at that sort of thing).
>>Well, IT DOES NOT WORK THAT WAY!!! The orange cast will NOT go away and if
>>it does, pretty much nothing remains!!
>
> In order to get reasonable RGB values from scanning a negative, you have
> to get rid of the orange cast *before* the image is digitized. On a
> flatbed, you may have no choice but to use cyan filters (pretty much the
> same way colour enlargers filter the light source to eliminate the
> mask colour). If you were building a negative scanner from scratch,
> you'd also be able to give red, green, and blue different exposure times
> to balance colour. But, either way, you want the RGB histograms *as
> they come from the scanner* to fill most of the available code range.
> Right now, you probably have almost no blue data.
>
>>Also, the negative of the negative is
>>just a horrible cyan mess!!. It's as if the negative has only the orange
>>(or
>>cyan) component on it.
>
> If you convert from negative to positive without getting rid of the
> orange cast, of course it will be cyan. Also, negative to positive
> conversion is *not* correctly done by the "invert" function in an
> editor; something more complex is needed (see my other post in this
> subject thread).
>
>>My gut feeling is that one must work in the CMYK
>>color space and not the RGB, seeing that the negative is a "transmission"
>>object and not a "reflection" one like a picture, using, of course,
>>different procedures.
>
> No, RGB works just fine. You just need to do the right operations in
> the right order.
>
> I've written software to digitize movie negative film, and the
> resulting positive images looked fine. (No, I can't send you a copy of
> the software; someone else owns it. But I did just describe the basics
> in two messages in these two postings).
>
> Dave
Anonymous
January 10, 2005 11:53:47 PM

Archived from groups: rec.photo.digital (More info?)

g n p wrote:
> Excellent and informative posts (both).
> Thank you very much.
> Really a (surprisingly) fairly complex process.
> Thanks again, I'll go buy a scanner with color negative capability
> 8-(( ________________
> G. Paleologopoulos
>

The EpsonScan software that came with my 4870 offers a choice between
Reflective and Film subjects. Film offers Transparency, Color Negative,
and BW Negative. Similar choices in the HP Scanner software with the
PhotoSmart S20. If you don't need more than 2500 dpi scans, the HP item
is available for not-much-money, and does a good job within its
limitations.

Good info and review/comparisin on this site:
http://www.sphoto.com/


--
Frank ess
Anonymous
January 11, 2005 3:15:39 AM

Archived from groups: rec.photo.digital (More info?)

Fantastic site!!



"Frank ess" <frank@fshe2fs.com> wrote in message
news:c_GdnQwGKKrweX_cRVn-gA@giganews.com...
>
>.........................................................
> Good info and review/comparisin on this site:
> http://www.sphoto.com/
>
>
> --
> Frank ess
>
Anonymous
January 13, 2005 7:05:12 PM

Archived from groups: rec.photo.digital (More info?)

OK, but surely it just depends on when the actual inversion is done?
If the software does all the non-linear stuff while the image is still
reversed (and I would have thought that was the case) and then as a
last step, does a linear inversion, then my comments apply.

I'm not sure why anyone would want the 'unlinear' negative image. But
point taken, if that's how they do it.
!