The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Can't get accurate colour from C1

Ben now youre this far there's a neat process to follow. Will email you a doc I wrote a few years back for Aperture and will provide a link here later for the same process in ACR. Should be easy to adapt for c1 (might even have a go)

Paul
 
The biggest colour problems I've seen are discoloured (yellowing) diffusing material and inconsistent flash colours. Shiny colour charts are also a PITA as they may pick up all sorts of things. I've recently upgraded from cheapo Asian stuff to Hensel, one of the main reasons being inconsistent colour between shots from same flash heads!
 
Ben, I've sent something through to the email on your website.

I'm still looking into various options to generate the LUT based ICC's that C1 needs. If anyone has time to look into ArgylCMS, RAWTherapee, dcp2icc (I've read both that it does and does not support LUTs), LUTbuddy, I'd be happy to pair up and pool resources.

There are some commercial apps of course, but I've no recommendation from anyone, nor have I waded through the free stuff yet to see invest the time and money in paid apps.
 

Ben Rubinstein

Active member
Paul you are a gentleman! Much appreciated. I have to look through it carefully.

Playing with the idea of something like Pictocolor InCamera software as the bosses have just said no to flying Stefan Steib out to sort out a full colour workflow and profile for us until the new studio is built as they won't pay for it to be done twice. :) They will however pay the $200 for the Picto software if I can persuade them it is worth it.
 

Mammy645

New member
Yes, very hard indeed...

Profiling for something like a scanner is "relatively" easy since there is one known light source and fewer issues with difference surfaces, difference pigment responses etc.

With a camera you are dealing with many, many variables and the results can be affected by things like metamerism, polarisation and so on...

Also some cameras "see" more than what can be represented on screen or on paper, so you have to take care of that part of the equation as well...
I'm not sure I understand. I'm talking about something like the Adobe DNG profiler, where you take a photo of a color chart and then the program evaluates it and creates a custom profile.
 

yaya

Active member
I'm not sure I understand. I'm talking about something like the Adobe DNG profiler, where you take a photo of a color chart and then the program evaluates it and creates a custom profile.
That would be an ideal solution if the results were reliably accurate and consistent. Adobe's solution is a bit different and is not suitable for reproduction work even though it does a reasonable job for many other applications.

There's a lot of math and a lot of theory involved here...the ICC has been trying to standardise a cross-platform, cross-vendor colour management system for many years.
 

Shashin

Well-known member
With a camera you are dealing with many, many variables and the results can be affected by things like metamerism, ...
Matamerism is never a problem because profiling input devices has nothing to do with metamerism. Metamerism is about two different materials looking the same under one light source and different under another. The camera should always be able to record the match/difference. Metamerism does not exist for one thing or object--it is always a comparison. Metamerism is a very misunderstood concept.
 

Shashin

Well-known member
Ben, what do you mean by accurate? You mean if you view the original and the monitor image at the same time you expect a perfect match? That would be impossible.

The hard part of this is you could post an example, but not an original.
 

Mammy645

New member
That would be an ideal solution if the results were reliably accurate and consistent. Adobe's solution is a bit different and is not suitable for reproduction work even though it does a reasonable job for many other applications.

There's a lot of math and a lot of theory involved here...the ICC has been trying to standardise a cross-platform, cross-vendor colour management system for many years.
Well, any solution is better than no solution. I've used several different techniques to make custom profiles over the years, ranging from external programs and photoshop scripts to just eyeballing it, while results varied they were all better than just using the canned profiles. I definitely think it's something Phase One should be looking into as it would be a huge asset to C1 IMHO.
 

dougpeterson

Workshop Member
Matamerism is never a problem because profiling input devices has nothing to do with metamerism. Metamerism is about two different materials looking the same under one light source and different under another. The camera should always be able to record the match/difference. Metamerism does not exist for one thing or object--it is always a comparison. Metamerism is a very misunderstood concept.
If your client is viewing the work of art in gallery lighting and you are shooting it with strobes and the pigments/dies/material of the art exhibits metamerism (when comparing those two light sources) then, by the perception of the client (the one paying/approving the job) a "correct" profile made purely based on a calibration under strobe will be wrong. You will have correctly calibrated to a widely accepted standard but you'll still be "wrong".

Moreover you may even wish to consider what light the artist created the work in, and under what kind of lighting they intended it to be viewed. Perhaps both strobe and gallery lighting are wrong.
 

Shashin

Well-known member
If your client is viewing the work of art in gallery lighting and you are shooting it with strobes and the pigments/dies/material of the art exhibits metamerism (when comparing those two light sources) then, by the perception of the client (the one paying/approving the job) a "correct" profile made purely based on a calibration under strobe will be wrong. You will have correctly calibrated to a widely accepted standard but you'll still be "wrong".

Moreover you may even wish to consider what light the artist created the work in, and under what kind of lighting they intended it to be viewed. Perhaps both strobe and gallery lighting are wrong.
So Doug, are you simply trying to prove that metamerism is a misunderstood concept. LOL.

Everything looks different under different lighting--and that is not metamerism. If you made two prints from a Epson and HP printer and they looked identical under tungsten light and then you took the same two prints and put them under daylight and the prints looked different, then the pigments are not metamers. However, neither print is "wrong," just different--unless you have really bad pigments.

Metamerism requires a comparison of TWO or more thing under different light sources where the spectral response causes two materials that appear to match under one source will not match under another--unless they are me tamers and then they will match. A sample of one has nothing to do with metamerism.
 

dougpeterson

Workshop Member
So Doug, are you simply trying to prove that metamerism is a misunderstood concept. LOL.

Everything looks different under different lighting--and that is not metamerism. If you made two prints from a Epson and HP printer and they looked identical under tungsten light and then you took the same two prints and put them under daylight and the prints looked different, then the pigments are not metamers. However, neither print is "wrong," just different--unless you have really bad pigments.

Metamerism requires a comparison of TWO or more thing under different light sources where the spectral response causes two materials that appear to match under one source will not match under another--unless they are me tamers and then they will match. A sample of one has nothing to do with metamerism.
Unless you're shooting paint samples then a single piece of artwork can contain more than a sample of one. More than one kind of paint/substrate/material.

Imagine a painting of a red kite next to a red house. The kite and house could appear to be an identical color of red under one type of lighting, but different when viewed under another light source.

In output (e.g. your Epson vs. HP example) you're usually using one set of inks. In input, especially in the Cultural Heritage market things are not always as straight forward.

Mixed media, various modern art, as well as pieces which were worked on over a long period of time (wherein the artist may have changed their tools during the creation of the artwork), or started by one artist and finished by another can all exhibit such issues.
 
Last edited:

Shashin

Well-known member
But metamerism comes basically down to one thing--one color, two materials, and different light sources. And my point was, this is not a color profile issue. The camera profile cannot and should not correct for this.

But Ben is not having this problem.
 

Shashin

Well-known member
But I am now wondering if Ben is viewing the originals under the same light as he is shooting with. If he is shooting with strobes and viewing with daylight tungsten, he will not get a match between the original and image.

Or maybe the problem is with expectations. There can never be a perfect match. The image should look like the original, but it is not a copy of the original. Glowing monitors cannot look like ink on paper.
 

Ben Rubinstein

Active member
Not quite that stupid, as mentioned in the original post, using daylight balanced bulbs for viewing the originals. We don't expect 100% but better than 70% would be nice. At present blacks are grey, pages are muddy dull colours, coloured ink is nothing like the originals in colour.
 

MGrayson

Subscriber and Workshop Member
FWIW, I have been unable to PRINT from C1 and get anything remotely color matched. If I export a TIFF and print from Lightroom, Photoshop, or Aperture, then all is well. If I print a test jpeg from C1, it looks good as well.. What has not worked at all is printing directly from a RAW file within C1. (This is on a Mac Pro with OS 10.7.4, and C1 6.x)

I'm not the only one, as I've seen this mentioned on the C1 fora, but haven't checked recently to see if it has been fixed. Ill try again when C1 7 comes out.

--Matt
 

Shashin

Well-known member
Sorry, I was not implying you are stupid. I am thinking about a solution and rather handicapped by not seeing anything. Sometimes we can overlook an obvious factor.

I have shot books for an exhibition catalog. Books don't have a black and the pages are dull--they are low-contrst targets. When you pump the contrast, the colors are hard to control and you see unwanted color in the paper. The problem does not come down to simple color management and profiling.

After I had finished this, I came across the LAB color space. I am wondering if that would be a better way to control the reproduction as the color and contrast are separate. You can increase the contrast while desaturating the color. LAB seems a difficult space to work in, but once you figure out a few basics, it is really not that hard, like if the A or B color curve intersects the center, there is no color shifts. The slope of the curve changes saturation and the movement of the curve impacts hue. The lightness curve is all about contrast.
 

EH21

Member
Hi Ben,
I'm using an AFi-ii 12 (aptus 12) for art repro work. Typically I light the work with profoto strobes using a cross polarizing technique. I've been scratching my head wondering how you could be so far off on color because with my experience the files have been really very close using the LF3 Product 5 or Profoto RGB profiles. I may make minor color adjustments or I may not depending on the job but only minor adjustments seem to be needed. I may use the linear curve setting and make my own curve however. On a few occasions I have also brought along my hasselblad CF 528 to shoot some work side by side. There are some very minor color differences between the respective backs using their stock color profiles but nothing most people would notice. I'm wondering if there isn't something else going on in your case that could be messing with color such as the OS applying a 2nd profile or something?
Eric
 
Top