Trying to understand color management for C-prints from LF film.

[Uri A]

Dear LFers,

I am trying to wrap my head around a LF film > drum scan > C-print workflow that makes sense (to my small brain). Maybe you can explain this to me?

Note: I am committed to C-prints. Inkjet is just too expensive for my huge prints and I actually prefer the look of C-types. In any case: can you please explain to me - why was I taught to pursue A98 for the last 15+ years - what's the point of pursuing a profile that cannot be printed?

The general color space we all (every lab i've ever used) pursue is Adobe 98. A98 is much bigger than any paper ICC profile (lets not even talk about ProPhoto RGB...). So when I soft proof to the ICC profile, I see that lots of the colors that A98 can display on my monitor aren't printable in a C-print.

So when I scan, retouch, etc, I am working to get the image looking right in A98: my scans are converted to A98, my software is set to display for A98, my monitor cost $2000 to reproduce 99.3% of A98 ... for what? Just so I can dumb it down to my lab's ICC for Fuji Lustre which shows about 2/3 of the colors that A98 can display on a $2000 monitor?

I don't understand this. Should I just set my working space to the ICC printing space I am trying to print in and do all my adjustments for that space? Why is A98 this holy grail if it cannot be printed? Please explain.

Thanks in advance for your help.

[marfa boomboom tx]

what is a C-print, to you?

to me, "c" stands for chromogenic... or in sloppy lab speak, from C-nn (22, 41, ...)

the c-print is the print version of a (likely) color negative as an intermediate. this stuff goes back to 1947. in its current form, the process of print chemistry is RA (see: http://www.kodak.com/global/en/busin...als/z130.jhtml)

[Peter De Smidt]

There are device specific and non-device specific color spaces. The former include scanner, monitor and printer profiles, whereas the latter include sRGB, Adobe 98, Prophoto.....

Device specific profiles, obviously, aren't interchangeable. They characterize a specific device. In addition, they aren't necessarily uniform color spaces. There can be gaps...

Non-device specific spaces, though, have uniform steps without irregular gaps.

Monitor profiles, while devices specific, tend to be more uniform than scanner or printer profiles. That's a good thing, which is why they're often characterized as containing most of sRGB or Adobe 98.

If you edit using a scanner profile, for example, you might not get predictable results, and the results could very tremendously with different output devices.

The right way to use a color managed workflow is to use an input profile, scanner or camera, convert the file to an editing space just big enough to contain the colors, such as sRGB or Adobe 98, using a monitor that closely matches the editing space, if possible. You then use soft-proofing to make decisions regarding adjustments for a specific printer. If you make the master file look good on a specific printer, then you could be up a creek if you use another printer.

[jp]

You use a bigger color space for working/adjustment so you don't screw it up in the middle steps. It is possible you have either a poor profile for printing, or some other color management setup problem. It might be worth $100 or so to have a starving pro come look over your setup to make sure your color management isn't misconfigured or less than optimal.

[Peter York]

I think the general idea is to have an archived digital file that looks right to you. Then, you can print the file in a number of ways, making adjustments that correct for the devices' inability to render the print as you like it. In other words, the C-print is a degradation of your digital image (lower color gamut). So time and energy is spent at the printing stage of the workflow to adjust for this degradation. The workflow assumes that your digital file is the most important, because in the future you may be able to print on a device that can replicate all the colors in the file.

A98 has a set of colors (gamut) that is big enough to encompass the gamut of multiple, varied output devices (C-print, inkjet, monitor, etc.). Note that the newest Epson printers which include orange and green inks have a gamut exceeding A98 in some areas. Many suggest working in ProPhoto, which has a larger gamut that A98.

Most adjustments should be minor at the printing stage. The conversion of a larger gamut (your digital file) to a smaller one (the C-print) generally goes smoothly (the relative colorimetric and perceptual algorithms work very well), assuming the profile of the output device is good. Certain colors or images, however, are problematic and then you make adjustments in Photoshop.

IMHO because you are moving from a transmissive "print" (your image on a monitor) to a reflective one, test prints are always necessary, though if you are really good it may only take one.

[Jim Andrada]

And of course there is always LAB and which (I believe) is the actual underlying color space of Photoshop

[bob carnie]

Using LAB is amazing but can be complicated, can be described as using a sledge hammer to nail in a finishing nail.

I do all my sharpening on the L channel , and I use the curves in LAB to do finiky colour correction or enhancement.
There is nothing in RGB, CMYK, that can spread complimentary colours away from each other than LAB.

And of course there is always LAB and which (I believe) is the actual underlying color space of Photoshop

[Preston]

I do all my adjustments in ARGB. When I sharpen for a print, I create a copy of the background layer, set its blending mode to 'Luminosity', add a vector mask and paint on the mask if needed and then use USM or Smart Sharpen. This works quite well, and the results appear to be same as USM on the L-channel in LAB.

--P

[bob carnie]

Good method.

I do all my adjustments in ARGB. When I sharpen for a print, I create a copy of the background layer, set its blending mode to 'Luminosity', add a vector mask and paint on the mask if needed and then use USM or Smart Sharpen. This works quite well, and the results appear to be same as USM on the L-channel in LAB.

--P

[pherold]

People do come at this from different perspectives. The OP is writing from what might be called the "output-centric" point of view, and other posters here have given the justification for the "input-centric" way of looking at things. Steve Upton wrote about this a few years ago in a newsletter article:
http://www.colorwiki.com/wiki/Color_...u_can_print.3F