Color Neg to Pos Conversion - Blue Cast

[Alan Klein]

Hi Alan, thanks for the nice words! Really, it took a bit of time the first time I attempted, but we would expect that wouldn't we? The first time, we don't know what to expect; and we move slower to study what happens at each step. Once I got past that learning curve, I did the other photo. That took less than 1 minute; and that included last final adjustments to taste.
Apologies for the late rely By The Way! Had a busy day. Oh! and no, I did not deviate from the instructions save the clipping bit. Please see my response to Pat regarding the ALT key below.

To sum it up, this is the best method I have used; and I intend to stick with it. Eventually, I will automate most of it so that one click will get me to the final adjustment stage. Just seconds per photo. I must say, I love shooting film for many reasons, and the hybrid use of film to digital is just fantastic!

I have to say thanks again to Pat here for recommending the article.

How would you automate it to one click?

[rawitz]

Furthermore, the orange mask is not just an added color, like sandwiching the negative with an orange gel; the intensity of the mask varies with the actual colors recorder on film. Which is why a correction that appears to work on the rebate leaves an orange (negative) or bluish (positive) cast on values above unexposed film.

I could not find detailed infos about the mask-technic. But surely the mask is not flat - like a yellow-orange lensfilter - but is it linear to density values or curved (gamma values)? And is there a separate mask layer in the film, or layers for eyery color, or within (added to) each color-layer?

And from the technical to the practical part. I do Color-Negfilm for many years now and do conversion within scanning.
There are different methods of digitally converting CNs. First is to directly convert within the scanner software and get a positive pic. Or to scan as rawscan and do the conversion and colorcorrection in the post with Photoshop, Lightroom or else. Or scan the negative as slide and remove the mask digitaly like described here.
But are these methods correct? Of course with some experience you will get "nice colors" with every method, but do you get the "true colors"?
Without answering my technical question about the mask I cannot answer this practical question. If the mask is "flat" every method is correct.
Presuming it is not, we have to take a Xrite-Color-Checker and make a correction-LUT for every color and density of every CN-Film. The better scanning software like Silverfast have done this with Presets for the most common CN-Films, or you can do it for your personal film within the software.
My own way is much easier. I do conversion within the scanning, but exactly manually defining the Black and White points of the picture. This will correct the mask linear for each color. Additionall defining a middletone grey, I can approximatly correct the (curved) color gammas.
I also tried the rawscan method, but you have to be very experienced in software to bring out the real picture from this extremly false raw picture and you have to work with 48bit data for all this color pushing and stretching.
For the mask subtracting method I´m in doubt. It cannot give true colors if the mask is not flat.

[Alan Klein]

I could not find detailed infos about the mask-technic. But surely the mask is not flat - like a yellow-orange lensfilter - but is it linear to density values or curved (gamma values)? And is there a separate mask layer in the film, or layers for eyery color, or within (added to) each color-layer?

And from the technical to the practical part. I do Color-Negfilm for many years now and do conversion within scanning.
There are different methods of digitally converting CNs. First is to directly convert within the scanner software and get a positive pic. Or to scan as rawscan and do the conversion and colorcorrection in the post with Photoshop, Lightroom or else. Or scan the negative as slide and remove the mask digitaly like described here.
But are these methods correct? Of course with some experience you will get "nice colors" with every method, but do you get the "true colors"?
Without answering my technical question about the mask I cannot answer this practical question. If the mask is "flat" every method is correct.
Presuming it is not, we have to take a Xrite-Color-Checker and make a correction-LUT for every color and density of every CN-Film. The better scanning software like Silverfast have done this with Presets for the most common CN-Films, or you can do it for your personal film within the software.
My own way is much easier. I do conversion within the scanning, but exactly manually defining the Black and White points of the picture. This will correct the mask linear for each color. Additionall defining a middletone grey, I can approximatly correct the (curved) color gammas.
I also tried the rawscan method, but you have to be very experienced in software to bring out the real picture from this extremly false raw picture and you have to work with 48bit data for all this color pushing and stretching.
For the mask subtracting method I´m in doubt. It cannot give true colors if the mask is not flat.

How do you know the true colors anyway with negative film?

Does it matter as long as it looks OK to the eye?

[rawitz]

You can know, if you compare the film or the scan with your Colorchart in origin.
If it matters is of course your decision, and even if not, I want to know why I like my personal result better.

[JLNims]

How would you automate it to one click?

Hi Alan. Just by writing a macro and/or an apple script for the first basic steps. The final adjustments still must be carried out by hand because of personal preference.

[Alan Klein]

Hi Alan. Just by writing a macro and/or an apple script for the first basic steps. The final adjustments still must be carried out by hand because of personal preference.

I was afraid you were going to say that.

[PatrickMarq]

Thanks Pat!
As far as the ALT key is concerned, what I did instead was watch my color scope set for RBG Waveform. I increased the channels until just before the tops of the channels hit the top of the scope. Bang! Done. Hope that helps.
I'm an Affinity Photo worshipper, so if it can be done with AP, I'll find a way (By god!).

In Affinity there is a ‘bug’ for the ALT key, but it only works for all the channels.
Getting there slowly

[PatrickMarq]

But, at the end I have chosen to use Silverfast as scanning software. There Negafix conversion for color is exactly what I like. Not cheap !
This way from 35mm to 4x5 and the future 6x17 all my Kodak Porta looks exactly the same.
For me life is to short to play around with software and my time out in the field is more important. To frustrated at work and going out to photograph keeps me healthy.
And I’m a IT person for more than 40 years.

[JLNims]

Gotcha Pat. Me, on the other hand am a retired electrical engineer with plenty of time and pain killers to play away with the editing!

[SergeyT]

I will through some ideas here. Maybe someone will bring them to fruition one day...

If by converting digitized color negatives into positive images one is looking to reproduce the manufacturer-designed properties of particular film type then a proper conversion formula(or algorithm ) must be content-unaware, meaning it must not analyze the content of the frame to perform the conversion of that frame. It must not sample\analyze the unexposed part of the frame neither (during the conversion of a particular frame), unless used for the purpose of auto-detecting film type.
Conversion "formula" must be film emulsion type (Ektar 100, Portra 160, etc), film exposure(under\normal\over) and film development time (push/pull) specific. Only these 3 parameters should be "fed" into the proper conversion formula (algorithm) to control the resulting output (colors and tones).
The output produced by application of such the formula to a digitized color negative data would be a more or less accurate reproduction of "positive image qualities" intended by the manufacturer that are based on film emulsion type , exposure and development times. It should not be considered as a final image and therefore can, and in most cases should, be adjusted further to suit the intended by photographer or image editor purpose.

A few clarifying examples :
* presence and size of dust in digitized color negative frames must not impact the output of the formula (algorithm)
* presence of sprocket holes on 135 format film frames must not impact the output of the formula (algorithm)
* manufacturing and development flaws on film frames must not impact the output of the formula (algorithm)