Proof profile for printing

Discussions, questions, comments and suggestions regarding Capture One PRO, Capture One PRO For Sony / Fujifilm, Capture One for Phase One and Capture One Express For Sony / For Fujifilm 20.x for Mac

Re: Proof profile for printing

Postby photoGrant » Thu Feb 13, 2020 10:42 pm

I would venture to say it's less a matter of taste and more a matter of fact. How will your work be seen by the audience? Replicate that. To muddy the waters there also continues to conflate that this part of the process is not subjective, or art -- it's science.

To mix these is to add to the circles of confusion. Pun intended! :lol:
photoGrant
 
Posts: 250
Joined: Thu Sep 18, 2014 5:16 pm
Location: Midbest

Re: Proof profile for printing

Postby NN635680879799322049UL » Fri Feb 14, 2020 5:07 am

Grant,

To me, the right color temperature for the monitor is the one that reproduces what the camera sees or should have seen if it were calibrated properly. I want the printer to reproduce what the monitor displays. As you said, getting this part right is a matter of science. Done right, the reference image taken by the camera and the printed one should under the same light in which it was taken look like the subject photographed. If the light did not have the color temperature you wanted, using a reference card or neutral area of the photo, you can adjust the white balance. Perfection achieved, at least in theory.

However, replicating how your most perfect print will be seen by the audience requires adjusting the printed image to the light source the audience will be using to view your print. That means accounting for intensity and color temperature of their light source(s) and accounting for the effect of the inverse square law. This will be achievable in a museum, but not in the average home to which people take those photos they bought at the art show.

Fortunately for those of us who cannot calibrate our workflow perfectly or set up our audiences lighting, color is not in the eye of the beholder, it is in the brain. Fortunately our brains adjust, up to a point, for shifts in white balance engendered by different color temperatures of a light source. When we view an image under tungsten light that was taken in daylight and properly printed to be accurate in daylight, we tend not to notice the shift so much because everything in our environment is also shifted.

Ideally, how do you recommend making a print that is faithful to the source? What is your process for getting from the subject photographed to the print so that it appears to your audience as it did when you photographed the subject, in general terms, not in detail.

Jerry C
NN635680879799322049UL
 
Posts: 285
Joined: Sun May 24, 2015 7:12 pm

Re: Proof profile for printing

Postby photoGrant » Fri Feb 14, 2020 6:17 am

Jerry,

The display white point should equal that of your viewing condition.

If you have a small photo-booth at home with a lightbulb that equates to 6500K or D65, then your monitor should match that white point.

If you believe the viewing conditions would be closer to D55, let's say the gallery you might display your work in is busiest in the morning glow but is empty during lunch -- then you would calibrate your monitor to that white point and you would swap your viewing booth at home with a bulb that is also D55.

In general if I wanted your colours from reality to match your print, and I was aware of the general limitations surrounding gamut, metameric failure, etc -- it would look like this;

Determine image pipeline is clear of applied profiles
Characterise the printer, paper and ink combo
Characterise the monitor
Calibrate the monitor to D65 / Adobe RGB 1998
Calibrate the monitor to the printer profile & paper/viewing condition combo

Then I would operate with no colour management on the operating system, and with the D65 calibration applied to the display. When I would like to proof my image for print I would change my display calibration from 'Web/D65' to 'Print/D55'. If I profiled and calibrated against multiple papers, I would create multiple calibrations as each paper has a different white point, tonal range & response, etc. At no point with this workflow would I need to use the software to softproof or convert or interpret anything to the printer.

Shooting I would use a colour chart that at least includes the same chip values that also appeared in my monitor and printing characterisation. I would then create a custom icc profile for the baseline of the colours to match. Anything that happens to the image thereafter is moving those values away from the 'ground truth' you've tried to establish in the workflow. Your viewing conditions have to be controlled equally, but this is assumed.

All this is under the notion you want to 'match'. If it's more likely that you want to render the colours as you 'remember' them, well, all of a sudden we've popped into the world of subjectivity and none of the above would apply.

Your point about the human eye being quite forgiving (or lazy) when it comes to 'how close is close enough?' is important and shouldn't be understated. The further down this rabbit hole you go, the less fun you have. Enjoy the feel and the reality of what you're outputting and save the quest for perfection or complete understanding of the theory for those that revel in arguing about it, as even the very international consortiums that set these standards cannot agree amongst themselves and change how they define things every few years.
photoGrant
 
Posts: 250
Joined: Thu Sep 18, 2014 5:16 pm
Location: Midbest

Re: Proof profile for printing

Postby BeO » Fri Feb 14, 2020 8:16 am

photoGrant wrote:
If you believe the viewing conditions would be closer to D55, let's say the gallery you might display your work in is busiest in the morning glow but is empty during lunch -- then you would calibrate your monitor to that white point and you would swap your viewing booth at home with a bulb that is also D55.


Grant,

I think you gave me a moment of "enlightment", let's see...

Let's say you'd want to hang the print on a wall to be lit by halogen or tungsten light, without daylight in that room, would you set your monitor to a profile with white point calibration of 2800K - 3000K, i.e. the respective color temp. of the light bulb in that target room? It seems logical but I never considered going so low as the lowest "paper" profile I remember having read about or seen was 4500K ("Print" profile in Eizo Color Navigator 7), maybe 4000K.

This assumed, a very "practical" question regarding C1 white balance: If you have an image without a white balance card in it, e.g. a street shot or building during the day, not sunrise nor sunset, I would choose the WB picker and try to find some reasonable "white" in the image, clicking on it several times gives me a different setting in the WB tool every time I click (in case I have repositioned the cursor of course), so I usually adjust the tool settings then by eye/taste. So, in this scenario, wouldn't it be a practical approach to compare the white part in the image on screen with a sheet of empty target paper to print on, or printed with pure white which should be the same, lighting it with the same light source, so about 3000K halogen bulb, and do the fine adjustments on the WB tool accordingly?

Not a scientific workflow but a practical approach?

Thanks
BeO
BeO
 
Posts: 600
Joined: Thu Nov 13, 2014 4:18 pm

Re: Proof profile for printing

Postby tenmangu81 » Fri Feb 14, 2020 11:01 am

Hi BeO,

I think adjusting the white balance of an image is one thing, finding the match between your display and the print is an other. But you are right at least in one thing : the human eye is very adaptative, and can adjust the white values and the right tones much faster and better than all our sophisticated devices and softwares !!
Robert
tenmangu81
 
Posts: 945
Joined: Thu Dec 04, 2014 6:33 pm
Location: Paris, France

Re: Proof profile for printing

Postby NN635680879799322049UL » Fri Feb 14, 2020 5:51 pm

Thanks to all for the theory and practical advice.

Excellent details, Grant. It answered my question. What equipment are you using to achieve the suggested workflow? In particular: Color chart in the field, camera(s, monitor, device and software for calibrating the monitor, and printer(s).

As a practical matter, in achieving good color management, how much should one be prepared to spend?

For most of us, the forgiving nature of the human eye is essential for not breaking the budget.

Jerry C
NN635680879799322049UL
 
Posts: 285
Joined: Sun May 24, 2015 7:12 pm

Re: Proof profile for printing

Postby photoGrant » Fri Feb 14, 2020 10:10 pm

BeO wrote:I think you gave me a moment of "enlightment", let's see...

Let's say you'd want to hang the print on a wall to be lit by halogen or tungsten light, without daylight in that room, would you set your monitor to a profile with white point calibration of 2800K - 3000K, i.e. the respective color temp. of the light bulb in that target room? It seems logical but I never considered going so low as the lowest "paper" profile I remember having read about or seen was 4500K ("Print" profile in Eizo Color Navigator 7), maybe 4000K.


If your print was in a box, it would receive no illumination. The colour of 'red' in your image is as good as 'black' as it requires a light source in order for your eye to see and interpret those colours. Those colours are pigments of ink on paper that are mixed with the temperature of light hitting it. What your eye see's is the result of a physical equation happening. If the light hitting the paper has a temperature/kelvin that is considered very 'warm', then all of the values bouncing out of that print and into your eye would shift in tandem. The white point of the paper is an additional variable. So with the print process alone you have these variables affecting your authorship: viewing light source, distance, ambient conditions, paper characteristics, ink qualities, framing & glass choice.

All of these combined variables create a 'reality' that you should profile and calibrate against. Characterising the printer with no conversion profiles applied and to the multiple conditions you expect is key. You want to remove guess work and assumptions and work with what you actually see/can record. You cannot profile the printer for one paper type and viewing condition and expect it to work for another the same way you cannot expect one display profile to work for a different display, that lives in a different location.

BeO wrote:

This assumed, a very "practical" question regarding C1 white balance: If you have an image without a white balance card in it, e.g. a street shot or building during the day, not sunrise nor sunset, I would choose the WB picker and try to find some reasonable "white" in the image, clicking on it several times gives me a different setting in the WB tool every time I click (in case I have repositioned the cursor of course), so I usually adjust the tool settings then by eye/taste. So, in this scenario, wouldn't it be a practical approach to compare the white part in the image on screen with a sheet of empty target paper to print on, or printed with pure white which should be the same, lighting it with the same light source, so about 3000K halogen bulb, and do the fine adjustments on the WB tool accordingly?

Not a scientific workflow but a practical approach?

Thanks


No, a piece of paper would not function as a way to find neutral gray. It only acts to get you close, and if that is more than good enough for you (it often is me), then you are not in the domain of repro level colour matching. Nor should you be! But paper itself as we discussed above has inherent and designed reflective properties. If the paper is altering your 'input' light and 'outputting' an additional few points of warmth then it cannot be considered an accurate way to find neutral gray. You would be better off with some PTFE tape :)

https://photographylife.com/diy-reliable-and-cheap-universal-white-balance-reference-device

If your prints are destined to live in a 3000K environment no matter what, then I would adjust your working conditions to match, yes.
photoGrant
 
Posts: 250
Joined: Thu Sep 18, 2014 5:16 pm
Location: Midbest

Re: Proof profile for printing

Postby photoGrant » Fri Feb 14, 2020 10:12 pm

NN635680879799322049UL wrote:Thanks to all for the theory and practical advice.

Excellent details, Grant. It answered my question. What equipment are you using to achieve the suggested workflow? In particular: Color chart in the field, camera(s, monitor, device and software for calibrating the monitor, and printer(s).

As a practical matter, in achieving good color management, how much should one be prepared to spend?

For most of us, the forgiving nature of the human eye is essential for not breaking the budget.

Jerry C


You could spend tens of thousands to achieve 'better' but you would see no difference between that and an X-Rite i1d3 and spectro setup.

The key in my opinion is a display capable of loading a 3D Lut directly, then you can emulate whatever you desire within the scope of that displays capabilities.
photoGrant
 
Posts: 250
Joined: Thu Sep 18, 2014 5:16 pm
Location: Midbest

Re: Proof profile for printing

Postby NN635680879799322049UL » Fri Feb 14, 2020 11:03 pm

Good practical answer. Sounds like containing cost here is a lot like it is with audiophile equipment. Going from really good to near perfect is an exponential climb. Perhaps the first place to start is in a really good spectrophotometer.

Jerry C
NN635680879799322049UL
 
Posts: 285
Joined: Sun May 24, 2015 7:12 pm

Re: Proof profile for printing

Postby BeO » Fri Feb 14, 2020 11:20 pm

Grant,

Regarding the practical approach, I probably didn't make my point.

The blank sheet of paper in not part of the image but lays on my desk besides the monitor. E.g. A sheet of Fuji matte. Light in this room is halogen, same as in the room I will be putting my prints on wall. The images will be printed on Fiju matte.
My monitor is calibrated with a colorimeter to a white point of 3000K.

Now, in the street shot there is a nice woman with a white skirt walking by. I was so faszinated by her that I forgot to ask her to hold a WB card or alike, let alone her phone number. :wink:
Developing now in C1, I take the WB picker, click on her skirt in the image, look at the blank paper Fuji matte on my desk to compare the skirt with, see the skirt has a slight color shift towards green when compared with the paper, so I slide the WB slider to magenta. Now the skirt looks as white as the Fuji paper does.

I just wanted to know if the principle idea is correct, you said no. I am unsure if you misunderstood me, or if my idea is totally wrong or just not "correct enough". It certainly is not meant for repro but a practical approach for everydays photography.

Best regards
BeO
BeO
 
Posts: 600
Joined: Thu Nov 13, 2014 4:18 pm

Re: Proof profile for printing

Postby BeO » Fri Feb 14, 2020 11:31 pm

NN635680879799322049UL wrote:Good practical answer. Sounds like containing cost here is a lot like it is with audiophile equipment. Going from really good to near perfect is an exponential climb. Perhaps the first place to start is in a really good spectrophotometer.

Jerry C

If you are happy with your prints you maybe should consider trying out some other interconnects or speaker cables... :D
BeO
 
Posts: 600
Joined: Thu Nov 13, 2014 4:18 pm

Re: Proof profile for printing

Postby SFA » Sat Feb 15, 2020 12:27 am

BeO wrote:
NN635680879799322049UL wrote:Good practical answer. Sounds like containing cost here is a lot like it is with audiophile equipment. Going from really good to near perfect is an exponential climb. Perhaps the first place to start is in a really good spectrophotometer.

Jerry C

If you are happy with your prints you maybe should consider trying out some other interconnects or speaker cables... :D


Probably after an ear calibration! :evil:

For older image editor users perhaps an eye calibration might also be a useful budget spend?

For those with artistic skills maybe just take up painting - there one can play with entire genres of work where colour accuracy is entirely in one's imagination.

Grant


* Emoji above chosen because it has ears ... (or marks that are supposed to represent ears.)
SFA
 
Posts: 7720
Joined: Tue Dec 20, 2011 9:32 pm

Re: Proof profile for printing

Postby NN635680879799322049UL » Sat Feb 15, 2020 1:42 am

LOL. My brother-in-law is a sound engineer. As arcane as this discussion has been, he takes it to a new level, all the way up to eleven.

Jerry C
NN635680879799322049UL
 
Posts: 285
Joined: Sun May 24, 2015 7:12 pm

Re: Proof profile for printing

Postby SFA » Sat Feb 15, 2020 5:53 am

NN635680879799322049UL wrote:LOL. My brother-in-law is a sound engineer. As arcane as this discussion has been, he takes it to a new level, all the way up to eleven.

Jerry C


It's interesting to hear about how people go about recording music in the studio (especially, say, orchestral music with some expensive musician resources) and then contemplating what happens to the sound in post processing and then what sort of equipment is used to listen to the result in the majority of audio consumption cases. Equipment includes ears as well as playback electronics.

Much the same could be observed for images - at least for those few of the millions created that are put on permanent display or made available via the internet.

Grant
SFA
 
Posts: 7720
Joined: Tue Dec 20, 2011 9:32 pm

Re: Proof profile for printing

Postby BeO » Sat Feb 15, 2020 8:18 am

SFA wrote:.


For older image editor users perhaps an eye calibration might also be a useful budget spend?

Grant

I use a profile for my eyes, it's the IwearGlasses.icm, though it adjusts for sharpness rather than colours... :lol:
BeO
 
Posts: 600
Joined: Thu Nov 13, 2014 4:18 pm

PreviousNext

Return to Capture One 20.x Software for Mac



Who is online

Users browsing this forum: No registered users and 1 guest

cron