Camera "Scanning" -- Use Di-Chro Head to undo Color Mask?

[Godfrey]

And yet you haven't even touched on the question how that workflow (totally) eliminates colour mask. Do you know the answer?

It's obvious.

The color mask in negative films was implemented to eliminate channel crossover problems in the chemical reversal color printing process. Without it, you couldn't get a clean RGB positive image from a color negative consistently. The color pack removed it when you made the print. It's a matter of the chemistry involved.

This is completely different from the notion of removing the crossover mask when working with CMY negative with mask -> RGB digitally. Digital negative to positive doesn't need the crossover mask, it's just a bias offset in the CMY channels in the digital space.. So I just filter it out with the customized camera calibration profile. Done.

(Likewise, you can print a negative image for chemical color printing on standard photofinishing equipment by doing a simple color inversion of a positive image and injecting an appropriate mask color overlay onto it. I used to create large format negatives like this, once upon a time, before decent inkjet printers became available in the 1990s...)

As illustration:

I grabbed a little strip of half-frame negatives I saw lying about in my drawer and snapped a photo of it using the Leica SL fitted with a Spiratone Vario-Dupliscope II. I let the sun fall on the diffuser glass to get a simple daylight color balance to the capture illumination. The DNG files were moved into Lightroom.

I exported a DNG compatible with DNG Profile Editor (it was last updated in 2012 so it can't normally open Leica SL DNGs for editing) and opened the editor, loaded a frame. First I inverted the tone curve. Then I sampled the rebate between the frames and adjusted it to come close to neutral bright white (black in the inverted image). Then I exported the edited CCP and restarted Lightroom.

After Lightroom restarted, I clicked on the Camera Calibration panel and chose my newly created "SL-colorneg-reversal" profile. Obviously, the total color matrix cannot be corrected with a single adjustment, but the resulting reversal was pretty close. A couple of minutes tweaking settings in the HSL panel netted this result:

Close enough. I'd export this as a color positive 16-bit TIFF and do my finish rendering on that (it's easier because the sense of the adjustment controls won't be inverted and they'll have more range of operation).

G

 

[stompyq]

Next, I move to Photoshop, apply the Color Neg action from Huss's reference (above).

This gives a pretty good color image.

You can correct any color cast left over at this point by converting the image in to LAB color space and doing the corrections there. I have had really good success this way

 

[brbo]

It's obvious.

... The color pack removed it when you made the print. It's a matter of the chemistry involved.

Chemistry removes the color mask?

 

[Godfrey]

Chemistry removes the color mask?

Light and the filter pack working with the chemistry. Yes. It was formulated to do that.

G

 

[brbo]

Light and the filter pack working with the chemistry. Yes. It was formulated to do that.

Is it possible to print optically and from scans (with laser light source) on the same paper?

 

[Godfrey]

Is it possible to print optically and from scans (with laser light source) on the same paper?

I not sure what you're asking here, or why a laser light source is being specified for scans...?...for both?

Do you mean, "If I scan a negative can I print it to photosensitive paper and process it in the traditional chemistry the same way I would expose photosensitive paper with a negative and process it?" The answer to that is certainly "Yes" presuming that you have a scan of a color negative that, when printed to the paper, is printed such that the same color light strikes the paper that the optical printing of the negatives do. Such a printing process will likely have its own calibration curves and filter pack requirements because it's probably a little too fussy a prospect to match the printing light from the scan perfectly with the output of an enlarger to paper.

As I said before, I at one time made large format color negatives for traditional printing machines. Given the differences in materials, although they looked pretty much the same to the naked eye, the spectral absorbtion characteristics of the ink on film or paper was a bit different from original negatives (dyes on clear polyester film base) and needed a customized filter pack to print accurately.

If you want to play with the concept, use a digital camera to capture a negative and print it to a 5x7 inch transparency. Then contact print it to color printing paper, obviously using a color pack or filtered light with a dichroic head, and process that. You can come up with some fun imagery that way...

G

 

[Huss]

Is it possible to print optically and from scans (with laser light source) on the same paper?

it is, but the paper usually is optimized for one process. An example, this Fuji paper is optimized for laser light printing:

http://www.fujifilmusa.com/products...s_printing_materials/super_type_CN/index.html

This one is for both film or digital:


http://www.fujifilmusa.com/products..._papers_printing_materials/type_II/index.html

 

[brbo]

I not sure what you're asking here

Can I optically print a negative on the same paper that minilab laser printers print on (from scans)?

 

[brbo]

Is it possible to print optically and from scans (with laser light source) on the same paper?

it is

How on earth can they do that?

 

[Godfrey]

Can I optically print a negative on the same paper that minilab laser printers print on (from scans)?

How on earth can they do that?

It's not something that is ordinarily done so the mechanism to do it would have to be contrived. But in the end, it's just paper that you expose to light and allow a negative to cast a shadow on.

G

 

[brbo]

It's not something that is ordinarily done...

Actually, it is. People optically print on the same paper that laser printers in minilabs can also use when they print from scans.

 

[Godfrey]

Actually, it is. People optically print on the same paper that laser printers in minilabs can also use when they print from scans.

So if you think you already know all about this, why are you asking these inane questions?

G

 

[ColSebastianMoran]

As illustration:

I grabbed a little strip of half-frame negatives I saw lying about in my drawer and snapped a photo of it using ... snip ...

Godfrey, thanks for the illustration and explanation. I agree, this gives color that's close enough. Nice motorcycles, nice color.

I took the jpg of your negative to look at the histogram, the red channel is well to the R of the others. Red will clip first when shooting a color negative in a camera-scan.

My aim with the cyan/magenta light is to bring all three channels up into the same range on the histogram. So that the inverted, color corrected histogram will have fewer gaps. Any thoughts on this idea?

 

[ColSebastianMoran]

I've been experimenting with different ways of doing the color correction from the inverted color negative:
- eye dropper to set WB
- curves adjustment, auto
- curves adjustment, auto, "Find Dark and Light Colors," and "Snap Neutral Midtones."

The last comes from Option while clicking the Auto button. This unknown-to-me Photoshop option I just discovered in the Action set Huss recommended.

Punch Line: This last does a good job. Damn good job. Here are the options I got from Huss's recommended article:

This auto-option is much better than the first and second, and much better than my past experience trying to get reasonable color.

Godfrey, this auto-option gives a result pretty close to your adjusted image, so it must be doing about the same thing as your profile.

This auto-option may be a silver bullet for getting reasonable color from camera-scans of color negs. Stay tuned. I'll do more with this.

 

[Ranchu]

This variable mask thing doesn't seem to be making much difference. If I'm thinking about it right, it would result in orange highlights, or blue midtones in the positive image. Does the mask vary consistently with density? It would be good to see some skin tones, though. I don't have any Adobe products installed or I'd do it, so just to mention...

 

[ColSebastianMoran]

This variable mask thing doesn't seem to be making much difference. It would be good to see some skin tones, though.

Yeah, bright colors may be easier than skin tones. I'll see what I have in negatives to try.

 

[Huss]

You boys are getting too high fallutin' for me.
I just create a profile in LR (too cheap to buy PhotoShop), export to NikFx where I use the correct color cast option. You can save that as a preset in NikFx. Works great for skin tones.

p.s. if I had photoshop I would definitely try the above suggestions.

 

[Ranchu]

So. If the channels (histograms) are matched closely by tuning the light source, the (consistent) variable mask just becomes a minor color cast removable in software?

 

[Huss]

Actually, it is. People optically print on the same paper that laser printers in minilabs can also use when they print from scans.

I think 'we' are making it out to be more complicated than it is. The colour head in the printer corrects for the orange cast on the film, whether it is a minilab printer or a home enlarger (for colour negative film).
The photo sensitive paper used is still silver halide based and works with any light source that it is exposed to. It is just that now some papers have been optimized to work better with whatever light source is being used, but they can still be used with either.
Hence the paper options that Fuji shows on its website.

There isn't a special chemical process that removes the orange etc as otherwise we would not have prints that intentionally have orange in them!

The interesting thing to me is the reasoning behind why C41 films have the orange base to them. You can also see this on B&W C41 films like BW400CN, Ilford XP2 etc

 

[Huss]

So. If the channels (histograms) are matched closely by tuning the light source, the (consistent) variable mask just becomes a minor color cast removable in software?

Yes Sir!

dot dot dot 10 characters.