Color Negative Manual Conversion Advice

[Lewipix]

Hi All,

Firstly, I am aware that there are great software options for color neg conversion/ inversion eg NLP, CP, Negmaster and so on. I am also aware that manual color negative conversion is a complex topic and at times controversial eg making global adjustments for orange masks...... but wish to avoid revisiting such discussions if possible.

I am seeking some practical How-To advice based on a couple of threads here where @Adrian Bacon was a main participant.

A lot of the technical discussion is above my pay-grade but I want to give manual conversions a go.

Adrian to the best of my understanding, works with his own code and with raw samples in a non color managed environment. I work in ACR/Photoshop with 35mm film and hope to emulate some of the steps within the boundaries that the tools allow.

So, to my questions:

1) The first step, if I have this right, is to make the gamma of each RGB channel the same, aka linear gamma aka normalize to same gamma = gamma 1.0
Question - How do I do this in ACR or PS? ( In Levels the gamma is already "1.0" for the respective channels by default)

2) Apply Gamma Multipliers "apply gain/multipliers to each channel until the film base plus fog is the same exposure, which will render it as light grey to white,"
Question - How do I do this in ACR or PS?
Question - is this step different to, or related to, White Balancing?

3) Invert (no problems here unless there is a special method?)

4) Shoot a series of gray cards over a range of exposures to see if there are any color casts still apparent

5) Which Color space - perhaps this should have been step 1
I would normally use Adode RGB


As an alternative to step 2, I gather sampling the orange color onto a new layer with Divide blend mode may negate the mask but may or may not be as effective as "multipliers".
Another possibility may be to use the channel mixer tool/layer but I do not know the method.

Anyway, TIA for any tips. Hopefully this may also help others that may be curious.

Cheers
David

 

[grat]

So for difficult inversions, I typically use Darktable + the Negadoctor plugin. Generally speaking, select the film base region, select a decent shadow region, and a decent highlight region, and that gets me 95% of the way usually.

But more direct, in Affinity Photo, is to sample the scanned, orange mask to set the white balance, do the inversion, and then adjust the levels for Red, Green and Blue channels manually-- set the levels for each channel so that there is no "empty" or "mostly flat" region for that channel, or to put it another way, I move the shadow slider to where the histogram data starts, and the highlight slider to where it stops. Then I adjust each midpoint (gamma) until that color isn't noticeable. So I slide blue until the blue cast isn't obvious, green until the green cast isn't obvious, and red until that isn't obvious.

It's not perfect. But it produces a result I can work with.

 

[Lewipix]

But more direct, in Affinity Photo, is to sample the scanned, orange mask to set the white balance, do the inversion,

Hi Grat

I could be totally wrong but i think the recommendation contained in the workflow described in the Opening post is to linearize the gamma in each channel before doing any WB or inversion or otherwise color transformations? ( I don't know how to do that)

and then adjust the levels for Red, Green and Blue channels manually-- set the levels for each channel so that there is no "empty" or "mostly flat" region for that channel, or to put it another way, I move the shadow slider to where the histogram data starts, and the highlight slider to where it stops. Then I adjust each midpoint (gamma) until that color isn't noticeable. So I slide blue until the blue cast isn't obvious, green until the green cast isn't obvious, and red until that isn't obvious.

I wonder if this is equivalent to the "Multiplier" step

It's not perfect. But it produces a result I can work with.

OK and thanks

 

[250swb]

If in Photoshop go for the cheap and reliable ColorPerfect, or pricier Negmaster. I have both but use ColorPerfect, it works without any changes to default (which is good because the interface is very 'old school' and unintuitive) and that is good enough to get the image into Photoshop reasonably accurately. Always use Adobe RGB and convert to TIFF for output files even for B&W negs and don't bother with any voodoo formulas for using the 'Invert' function in Photoshop, its crap.

 

[Alan Edward Klein]

If in Photoshop go for the cheap and reliable ColorPerfect, or pricier Negmaster. I have both but use ColorPerfect, it works without any changes to default (which is good because the interface is very 'old school' and unintuitive) and that is good enough to get the image into Photoshop reasonably accurately. Always use Adobe RGB and convert to TIFF for output files even for B&W negs and don't bother with any voodoo formulas for using the 'Invert' function in Photoshop, its crap.

Steve, Where can we see samples?

 

[mohmad khatab]

Hello my dear colleague,
Allow me to express my humble opinion on this matter, and I hope that you will be generous with my opinion.
I don't know what is the ultimate goal of this project as a whole.?
In fact, my understanding is that the invention of the C41 process was with the aim of making a negative in order to print images on paper directly from the negative.
And as far as I know, the reverse process (E6) or whatever the name of the process was, it was with the aim of producing a chip that could be used on the projector..

The same thing is done in the world of black and white, there are negatives that have been set up in order to print images, and there is another type for display on a projector screen.
Here we can pause for a while, there have been some experiments that make it possible to use printing film in order to produce slides that can be displayed on the display screen.
Personally, I conducted after those experiments, and I am still doing with categories of negatives. It was impossible to produce transparency strips that could be displayed on a projector, such as (FOMA100) and the like.
And it became clear to me that there is a certain amount of success in these experiments despite their difficulty and despite the waste of money and time in them. In the end, it is actually possible to obtain transparency that can be presented on the projector. We may agree a lot and disagree a little about the quality of that slide. Transparency and The extent to which they comply with quality standards, but eventually, one way or another, we will get a transparency slide that can be displayed on the projector. From my point of view, what made this possible is that it is black and white, and everything that is not black will be white or gray.
But the matter with the color is completely different, the orange mask will give absolutely unexpected colors.

In my opinion, this is almost impossible with color negatives and in all cases it is not possible to obtain high-quality colors and you cannot get a 100% transparent film base, so it is not justified to go in this direction because the end result will be bad .

Someone will say: Let's use Photoshop to scan those slides, and here I will ask a question,,
(That man who bought a doll and announced that he had married her, did he really find the love of his life? Will he get true love? Will he really get the emotional satisfaction or is he deceiving himself?)

Using Photoshop to manipulate and edit a negative, there have been attempts to turn it into a transparency slide, is just like the man who married the doll.
The doll was produced for a specific purpose, which is to play with it, decorate the place with it, and so on, and it will not be possible to have children from it.
There is only one chance for this to work.
And is to get the film color aerial photography that does not have an orange mask.
In the past, the Agfa company produced some of it for the use of photographers, but it quickly stopped producing it, and I don't know why.

We could ask that old new company (ORWO) to consider reproducing it.

 

[grat]

I could be totally wrong but i think the recommendation contained in the workflow described in the Opening post is to linearize the gamma in each channel before doing any WB or inversion or otherwise color transformations? ( I don't know how to do that)

You're almost certainly totally right. What I posted is a very quick and effective process I've found, not for "perfect" inversion, but for a quality inversion without any excess color casts, that I can work with to produce the image I want.

Essentially, you asked for "any other tips". I'm not familiar with the process outlined in your original post, so I can't offer specifics on it. Adrian will almost certainly stop by at some point, and provide (as usual), in-depth advice and discussion. I'm looking forward to it, but thought I'd contribute my method(s) while waiting.

 

[Lewipix]

That man who bought a doll and announced that he had married her, did he really find the love of his life? Will he get true love? Will he really get the emotional satisfaction or is he deceiving himself?

The easiest person to deceive ("fool") is ourselves, to paraphrase Feynman, but I don't think he had blow-up dolls in mind when he said that

impossible with color negatives and in all cases it is not possible to obtain high-quality colors

Keeping with the Feynman theme and in relation to impossible color negative conversions, perhaps we just need a different set of Tools -Surely You're Joking, Mr. Feynman! Adventures of a curious character - A Different Box of Tools "So I got a great reputation for doing integrals, only because my box of tools was different from everybody else’s, and they had tried all their tools on it before giving the problem to me."

David

 

[Lewipix]

You're almost certainly totally right. What I posted is a very quick and effective process I've found, not for "perfect" inversion, but for a quality inversion without any excess color casts, that I can work with to produce the image I want.

Essentially, you asked for "any other tips". I'm not familiar with the process outlined in your original post, so I can't offer specifics on it. Adrian will almost certainly stop by at some point, and provide (as usual), in-depth advice and discussion. I'm looking forward to it, but thought I'd contribute my method(s) while waiting.

Thanks again and I hope so too.

 

[Lewipix]

Different Box of Tools

MatLab rgb2lin Linearize gamma-corrected RGB Hmmm

I started poking around the Image processing capabilities of Matlab. It can also be integrated into photoshop (apparently). I'm guessing its reasonably easy to learn a few specific tricks within Matlab without having to be a professional coder or learn Python or something

 

[250swb]

Steve, Where can we see samples?

Hi Alan, examples of what? I guess Negmaster has examples on it's web site, but when I use ColorPerfect or Negmaster for my own work it's for the initial inversion and pp of B&W negatives because I hardly ever use colour. The times I use it for colour negatives is when I'm occasionally working on an archiving job and the proper software saves hours of further post processing. It's not perfect every time, like when there is an overall colour cast, a case in point was some photos shot inside a steel works and the orange glow fooled ColorPerfect, but less so Negmaster. But as I neither shot the images or own them I can't really post any examples.

 

[Alan Edward Klein]

Hi Alan, examples of what? I guess Negmaster has examples on it's web site, but when I use ColorPerfect or Negmaster for my own work it's for the initial inversion and pp of B&W negatives because I hardly ever use colour. The times I use it for colour negatives is when I'm occasionally working on an archiving job and the proper software saves hours of further post processing. It's not perfect every time, like when there is an overall colour cast, a case in point was some photos shot inside a steel works and the orange glow fooled ColorPerfect, but less so Negmaster. But as I neither shot the images or own them I can't really post any examples.

I wouldn't trust the software manufacturers' samples. Obviously they're going to select the perfect ones since they're selling a product. That's why I asked for users' experience and their samples. Your experience doesn't seem to provide a strong support.

The color cast you mention is the main problem as I see it. Unless you get the picture perfect and use filters beforehand, any color cast is going to fool the programs. We often have open shade and get blue casts. The programs apparently don't deal with casts very well. It's why I shoot chromes. They're simple to scan.

 

[PhilBurton]

I hope people don't get upset with me for for asking this:

What if your digital processing workflow is centered around Lightroom, and jumping over to Photoshop for final touches for challenging images?

 

[gone]

This conversation is above my pay grade, digitally speaking, but in the past I've simply used the adjustments available w/ PS 7. There is no "correct" color, it's all in the eye of the beholder, that's why they make different color films.

Auto Color, Hue, Saturation, etc have always worked for me. I've never had to isolate certain areas or colors, do masks, any of that. When I would print digitally, it was done just like it's done in the darkroom, by inspection. Do a few trial prints, get it dialed in, and make the big prints.

Otherwise, the variables go off the charts. Differences in monitor calibrations, differences if you send a file out for printing, differences between printers/inks/papers, on and on.

 

[Lewipix]

In Regards to making gamma (approximately?) equal for each color channel in PS I have come across a few suggestions with some variations but along the lines of - in the curves clip the black and white points for each of the RGB channels. In the levels tool adjust the gamma to approximately -15 for red and +15 for blue. Do not touch the green.

 

[runswithsizzers]

I hope people don't get upset with me for for asking this:

What if your digital processing workflow is centered around Lightroom, and jumping over to Photoshop for final touches for challenging images?

For color negatives, I always start in Lightroom, but that is necessary because I use the Negative Lab Pro plugin (NLP). If there is any significant amount of dust, I do use Photoshop's vastly superior clone/heal/dust tools, but otherwise, I often use Lightroom+NLP, exclusively for my color negatives.

Before getting the NLP plugin I did try some manual color conversions in Photoshop, with variable results - some conversions I was more-or-less happy with, others not. But I found it to be a frustrating process. In the long run, I decided the cost of buying NLP - and the time I spent learning how to use NLP - has probably paid off, for me. While NLP is rarely an instant process, I now spend far less time on each image, and I am happy with the results. Examples <here>

Of course for b&w and color slide film, the process is much more straightforward and I am happy to do the whole process in Photoshop, no NLP needed.

 

[mohmad khatab]

I wouldn't trust the software manufacturers' samples. Obviously they're going to select the perfect ones since they're selling a product. That's why I asked for users' experience and their samples. Your experience doesn't seem to provide a strong support.

The color cast you mention is the main problem as I see it. Unless you get the picture perfect and use filters beforehand, any color cast is going to fool the programs. We often have open shade and get blue casts. The programs apparently don't deal with casts very well. It's why I shoot chromes. They're simple to scan.

I find your opinion and point of view worthy of respect and appreciation
Regards

 

[Lewipix]

For color negatives, I always start in Lightroom, but that is necessary because I use the Negative Lab Pro plugin (NLP). If there is any significant amount of dust, I do use Photoshop's vastly superior clone/heal/dust tools, but otherwise, I often use Lightroom+NLP, exclusively for my color negatives.

Before getting the NLP plugin I did try some manual color conversions in Photoshop, with variable results - some conversions I was more-or-less happy with, others not. But I found it to be a frustrating process. In the long run, I decided the cost of buying NLP - and the time I spent learning how to use NLP - has probably paid off, for me. While NLP is rarely an instant process, I now spend far less time on each image, and I am happy with the results. Examples <here>

Of course for b&w and color slide film, the process is much more straightforward and I am happy to do the whole process in Photoshop, no NLP needed.

Hi

do you mind detailing what your manual conversion was/is?

Like you I think for the most part I will rely on a dedicated software like Negmaster or NLP to do the 'heavy lifting'. Its really, for me, learning about the manual processes and wondering if there is a place for that approach for 'special' shots (irrespective of time spent).

 

[McDiesel]

if I have this right, is to make the gamma of each RGB channel the same, aka linear gamma aka normalize to same gamma = gamma 1.0

Mmm no. Making gamma the same across channels is not "AKA gamma 1.0". There's film gamma and digital capture gamma, you need to equalize the former (film layers have mismatched gamma, and you can't make it 1.0 because it's set chemically) and AFTER that you can adjust the digital overall gamma to your taste, and I doubt it will be 1.0

In Regards to making gamma (approximately?) equal for each color channel in PS I have come across a few suggestions with some variations but along the lines of - in the curves clip the black and white points for each of the RGB channels. In the levels tool adjust the gamma to approximately -15 for red and +15 for blue. Do not touch the green.

Correct, although +15/-15 numbers vary from emulsion to emulsion (and on the look you're after). BTW you can do both using the Levels tool, and I would suggest holding ALT when clipping - this will be more accurate than relying on the histogram alone.

Keep in mind too, that print film never meant to have a reference/scientific look. It's helpful to think of print film color it as a subject for interpretation, in other words the colors are meant to be set during printing/scanning. People get their Fuji films scanned by a local lab with a Kodak preset and then complain about "Fuji green". You have to dial in the look you want when inverting. Maybe that's why back in the day product photographers who cared about color accuracy used slide film?

Just find the gamma adjustments that make an image that's pleasing to your eye. Good luck!

 

[runswithsizzers]

Hi

do you mind detailing what your manual conversion was/is?

Like you I think for the most part I will rely on a dedicated software like Negmaster or NLP to do the 'heavy lifting'. Its really, for me, learning about the manual processes and wondering if there is a place for that approach for 'special' shots (irrespective of time spent).

Sorry I can't be more helpful, but I don't recall using any systematic process - mostly, I just muddled through. At that time, I believe I was still using a film scanner with VueScan, which did the initial conversion, but not very well. I was never able to find any combination of VueScan settings that did not require a fair amount of additional effort in Photoshop trying to get the colors to look right.

After retiring the film scanner and switiching to using a digital camera to copy film, I felt like my attempts at manual conversions were taking too much time, so I started working with Negative Lab Pro. I can't say I love NLP, but after using the plug-in for a while, I am now usually able to get acceptable results in a reasonable amount of time.

 

[MattKing]

Keep in mind too, that print film never meant to have a reference/scientific look. It's helpful to think of print film color it as a subject for interpretation, in other words the colors are meant to be set during printing/scanning. People get their Fuji films scanned by a local lab with a Kodak preset and then complain about "Fuji green". You have to dial in the look you want when inverting. Maybe that's why back in the day product photographers who cared about color accuracy used slide film?

Just find the gamma adjustments that make an image that's pleasing to your eye. Good luck!

Somehow I think @Mr Bill might have a different outlook on some of this.
I certainly know that I did when doing colour printing back in the day for professional photographers.

 

[McDiesel]

He certainly would agree with me on the fact that the final color = emulsion color * paper color * CMY filters. This certainly can be compressed down to colors are set during printing. I am now going to keep my lips tight about scanning, as I remember getting in trouble for that in the past

 

[Lewipix]

Mmm no. Making gamma the same across channels is not "AKA gamma 1.0". There's film gamma and digital capture gamma, you need to equalize the former (film layers have mismatched gamma, and you can't make it 1.0 because it's set chemically) and AFTER that you can adjust the digital overall gamma to your taste, and I doubt it will be 1.0

Yeh thanks. I mentioned in my OP that gamma is already 1 for all three channels in PS. I'm guessing now that the gamma displayed in PS is relative gamma and the absolute gamma could be anything as multipled by X 1.0 eg gamma 2.2 x 1 = 2.2. I am wondering whether the "multipliers" referred to by Adrian are applied to this relative gamma of 1.0 in varying degrees to equalize the different RGB gammas to the same value of say 2,2 or whatever. But then both step 1 and 2 would both be addressing the same thing, correcting film gamma?

Correct, although +15/-15 numbers vary from emulsion to emulsion (and on the look you're after). BTW you can do both using the Levels tool, and I would suggest holding ALT when clipping - this will be more accurate than relying on the histogram alone.

Okay and thanks

I have also come across the method whereby you sample the rebate color onto a new layer then blend using in mode "Division". I'm not sure if this addresses the gamma inequality or something else related to the mask

Just find the gamma adjustments that make an image that's pleasing to your eye. Good luck!

Thanks and yes, It seems that PS does not have the tools to click something to mathematically "equalize gamma" and/or add gain/multipliers (for whatever purpose).

Maybe (?) a Matlab script does the job to some extent, but unless it has already been written it would be beyond me to figure out.

If correct, one of the the methods above will get the job done and involve 'eye-balling' it to judge results. That is fine in the sense that irrespective of whatever method used there will always be the possibility for tweaking to suit whatver taste or artistic intent that pleases the eye.

 

[Lewipix]

Sorry I can't be more helpful, but I don't recall using any systematic process - mostly, I just muddled through. At that time, I believe I was still using a film scanner with VueScan, which did the initial conversion, but not very well. I was never able to find any combination of VueScan settings that did not require a fair amount of additional effort in Photoshop trying to get the colors to look right.

After retiring the film scanner and switiching to using a digital camera to copy film, I felt like my attempts at manual conversions were taking too much time, so I started working with Negative Lab Pro. I can't say I love NLP, but after using the plug-in for a while, I am now usually able to get acceptable results in a reasonable amount of time.

No problems, Thank you!

 

[McDiesel]

Yeh thanks. I mentioned in my OP that gamma is already 1 for all three channels in PS. I'm guessing now that the gamma displayed in PS is relative gamma and the absolute gamma could be anything as multipled by X 1.0 eg gamma 2.2 x 1 = 2.2. I am wondering whether the "multipliers" referred to by Adrian are applied to this relative gamma of 1.0 in varying degrees to equalize the different RGB gammas to the same value of say 2,2 or whatever. But then both step 1 and 2 would both be addressing the same thing, correcting film gamma?

I am not 100% sure I am following, but generally this sounds about right. You may be missing one more thing though: the gamma set by a DCP profile, if using a camera to digitize. Some RAW converters allow you to switch to Linear Mode - that's real 1.0 capture gamma across all channels. This allows you to see the film image exactly the same way sensor saw it without camera curves applied to it.

My understanding of Adrian's algorithm is that it's a fitting algorithm, i.e. he's effectively applying custom LUT to get true grey across multiple exposures, and it's easier to do on a working image which is linear and uses a huge color space, in other words it's an implementation detail that may not be relevant to you.