Question about neutral pipette (white balance)

General topics about imaging
User avatar
Jossie
SilverFast User
SilverFast User
Posts: 82
Joined: Sun Nov 11, 2012 4:39 pm
Scanner: Reflecta DigitDia 6000
SilverFast Product: Archive Suite
Location: im Kraichgau

Question about neutral pipette (white balance)

Postby Jossie » Wed Feb 01, 2017 2:54 pm

Hello,

in the job manager only the positions of the neutral pipette measurements are copied from the source to the destination image. Quite often one has the problem, that a given image lags neutral areas so there is no way to use the neutral pipette. Thus one would like to transfer the results of the neutral pipette operation to the destination image.

In the dialog for the neutral pipette there are values given for R, G and B for the source and the destination point. These values can be edited. But as long as one does not know what these numbers mean, it is not possible to do anything with these. I have tried the following: I marked 2 positions with the neutral pipette and got four sets of RGB values for these (2 source and 2 destination). I have calculated the differences between source and destination values for both and averaged them. Then I deleted the second entry and applied the average differences to the first data point. This would be an operation I could image is performed behind the scene. But the result is not the same as for the original 2 positions marked with the pipette. So my assumption obviously is not true.

So my question is twofold:
1) What is the meaning of the RGB values (source and destination) in the panel of the neutral pipette results?
2) How is the averaging done and how can I transform the white balance result from one image to another manually?

Of course, the transformation of white balance cannot always be done with valid results, e.g. if the illumination changed between the scenes. But if one has slides from the same film reel and with identical illumination conditions but different motives, then the transformation should give good results.

Hermann-Josef Röser

User avatar
LSI_Ketelhohn
LSI Staff
LSI Staff
Posts: 4265
Joined: Wed May 28, 2008 11:19 am
Scanner: all
Location: Kiel, Germany
Contact:

Re: Question about neutral pipette (white balance)

Postby LSI_Ketelhohn » Mon Feb 06, 2017 4:55 pm

Dear Mr Röser,

copying the pipette values unfortunately does not help because the histogram is individual for each image. Neutral grey values from one image can not be used to correctly remove casts from another image.

The values in the Neutral Pipette window are RGB values. The values are not simply averaged. Each point works based on the previous one.

Kind regards,
Arne

User avatar
Jossie
SilverFast User
SilverFast User
Posts: 82
Joined: Sun Nov 11, 2012 4:39 pm
Scanner: Reflecta DigitDia 6000
SilverFast Product: Archive Suite
Location: im Kraichgau

Re: Question about neutral pipette (white balance)

Postby Jossie » Mon Feb 06, 2017 7:12 pm

Dear Mr. Ketelhohn,

thank you for your answer. As I had mentioned in my last paragraph, copying will not work in general. But same film reel and same illumination should not pose a problem.

I think it should possible to copy the results from one image to the other due to the following thought experiment (please tell me, if I am wrong): Imagine three images taken from the same film reel: two portrait images #1 and #2 taken next to each other (like in a panorama shot) and the third one #3 in landscape mode, covering a good fraction of the two other images. Let only #1 have neutral grey areas, which are also covered by image #3. You will agree that using the neutral pipette on images #1 and #3 should give the same result. And then automatically also image #2 (with no grey areas) must be corrected with the same settings.

If I see it correctly, the neutral pipette produces shifts (and perhaps stretches / compression) of the R,G, B histograms. So why should it not work using the same operation in an image with no grey areas? I did the experiment on two images of the same scene with slightly different settings. I used the neutral pipette on the first image and on the second image, where I placed the measuring point deliberately in a bright red area. Of course, the colours in image #2 got ruined. But when I copied the values for the neutral pipette from image #1 to image #2, the colours in image #2 became as good as those in image #1! Q.e.d.

Here is the result with 4 positions marked with the neutral pipette in the left image and copied manually to the right image (after marking 4 dummy positions):
ohne_NP.jpg
Both images without neutral pipette
ohne_NP.jpg (73.79 KiB) Viewed 783 times

mit_NP.jpg
Neutral pipette used on left image and copied manually to the right image
mit_NP.jpg (69.63 KiB) Viewed 783 times

As is evident, the colours in the second image come out perfectly after manually copying the pipette data.

So I do not understand your argument, that it is not possible to transfer the results from one image to another. Doing it manually works, also with several settings. It would only be convenient if the copy process could be done in the job manager.

With kind regards

Hermann-Josef Röser

User avatar
Jossie
SilverFast User
SilverFast User
Posts: 82
Joined: Sun Nov 11, 2012 4:39 pm
Scanner: Reflecta DigitDia 6000
SilverFast Product: Archive Suite
Location: im Kraichgau

Re: Question about neutral pipette (white balance)

Postby Jossie » Mon Feb 06, 2017 7:24 pm

Just one addition to my previous post.

Here are the histograms of the two images. They are quite different. And nevertheless, copying worked perfectly!

Hermann-Josef Röser

Histo_4423.jpg
Histo_4423.jpg (20.65 KiB) Viewed 783 times
Histo_4424.jpg
Histo_4424.jpg (21.17 KiB) Viewed 783 times


Return to “Imaging in general”

Who is online

Users browsing this forum: No registered users and 2 guests