Monitor calibration

[swanseadave]

As good as your monitor calibration tutorial is I`m sure you`d agree that you can`t beat what a good calibrator would do.
Knowing how critical color is to successful photography,might it be a good idea to review some calibrators for the benefit of members.I`m thinking Xrite and Spyder.

After all how does one know that the colors you see on screen in your photo editing program are correct?Your monitor colors may be out leading to your wrongly assessing the image color during editing.It would only become apparent when your print doesn`t match the screen shot.

It`s only fairly recently that I`ve considered this and now intend to buy a calibrator.

Do you think it feasible to test some and post the results?

Thanks.

[rpcrowe]

This might seem quite elementary but, if I calibrate my monitor to match my printer's output, will it be necessarily be calibrated for correct colors when posting on the Internet - or is this two different sides of the same coin

[Polar01]

David, I have used the Spyder and now the Xrite make of Calibrator. The Spyder would only calibrate and create a profile for the monitor at the time I would only print now and then on a 4-colour shelf top printer 8.8" x 11" or view on a electronic device. I later purchased a higher end printer at too good of a deal to refuse I found with ICC profiles supplied by the printer maker that on rare times that the printer was not producing reds and oranges the same as I was seeing on the screen. The Spyder that I had could not produce profiles of what the printer could print, so I went looking for a system that could produce a profile for both my monitor and printer. What I found was that the print maker Epson and a large number of paper makers that produced ICC profiles use the Xrite line of calibrators, so I thought if it is good enough for them so be it.
That is not to say that the Spyder is not a good system to use it is, if you are posting to the net, view on an electronic device or printing the odd 8.5" x11" print. If you are printing 11 colours on sheets that are 17" x 22", @$8.00, and use $8.00 of ink a pop for someone and it does not come out correct the first time and you have to do it again on something you are now $32.00 in costs, for something that you are charging $40.00 for. I am sure that Spyder does make a system that will provide both monitor and printer profiles.
One last thing if you are only going to print the odd image or sent images out to a lab, most labs want sRGB colour space, either make that will produce a monitor profile that is all you need, do not spend the extra money to get one that will produce printer profiles. If a some point you decide to invest in a high-end printer that upgrade to one that will provide both.

Cheers: Allan

[Polar01]

Richard it will not be necessary to produce a profile for the net, the profile is created to correct any colour difference that the your monitor incorrectly displays, most monitors only display sRGB colour space which is what the colour space the net requires. Now if the image is viewed from the net on an un-profiled monitor it could very well look bad and I mean real bad. There is really nothing you can do if some one has the reds cranked up on their monitor so to make it look correct to them (think somewhat colour blind).

Cheers: Allan

[Codebreaker]

There are many reviews on the Internet about various devices but I'm not sure how you would go about comparing and posting results yourself, unless you bought each device. After all I'll never really be able to see what's on your screen.

Personally I've been a long time user of the Eye-One Display Pro. One factor that made me go for this was that at the time the likes of Epson where using Gretag devices as Allan has also stated. The devices have moved on in the years since I started and now Xrite owns Gretag and the latest Eye-one Display Pro will do more than just monitors but projectors, scanners and printers - all for a price of course.

Just a rewind on the calibration and profiling aspects. You should not be calibrating and profiling your monitor to match print output but rather doing this to accurately render the image on screen as shot. Bare in mind that cameras, printers, scanners, projectors most likely all have a different colour gamut i.e it is unlikely they will be able to display the same colour range anyway. But of course you want them to display the colours within gamut as best as possible and handle those out of gamut colours well.

Despite printer manufacturers claims of having this or that colour space, I've found this to be somewhat of an exaggeration. Yes some of the wider space colours (compared with sRGB) may be capable of being rendered on a print but not all of them. So much depends on the inks and the paper type. Having an accurate printer profile is essential.

For what its worth with Epson - my preference - I've found that the canned Epson Printer Profiles work very well with Epson Ink and Papers. Any other combination of ink types and paper types often requires a custom profile.

What colour management via calibration and profiling is doing is providing the best translation of data between different devices to achieve the best possible colour consistency. In particular how to handle out of gamut colours. The display is the starting point once the data is out of the camera - get this wrong and every thing else that follows may also be wrong. But you need a colour managed program to be able to pass colour data around the devices correctly. And not just any colour managed program but those that will use the profiles you've created or set as default.

Just one final point that's often overlooked. Assessing a print can only really be done when its viewed under good lighting conditions and for me this means using a bulb with a 65k temperature approximating bright daylight. FWIW I also set my screen colour balance to 65K.

[swanseadave]

Many thanks,that`s really helpful.I was edging toward the xrite.
I`ll let the forum know my findings in due coursr.

[Manfred M]

+1 to what the others have written.

Unless you have a calibrated and profiled computer screen, you don't know if what you are seeing in an accurate representation of what your camera has captured. To me a colour managed workflow goes from camera to computer to the print. I don't really care about what others see on the internet as I have no control over the quality of their computer screen or the settings they use on it.

I replaced my rather ancient x-Rite i1 for an x-Rite ColorMunki Display earlier this year. The commercial photographers I have discussed the x-Rite (i1 and ColorMunki) versus DataColor (Spyder) have all told me that in their experience, both companies make good products, but the x-Rite solutions are better.

I use the paper manufacturer supplied profiles during printing and have gotten good results with Epson and third party papers (I use an Epson 3880 printer).

[James G]

I have only ever used the Spyder system and use both the Spyder Monitor & Spyder Print calibration systems.

Like Manfred what others see on the internet is of little interest to me since I cannot control the systems that others use to display on.

For some time now, I have stopped using manufacturers' ICC profiles for papers, preferring to create the profiles myself, (Hence the Spyder Print system!). My reason for this, is that I use a 3rd party continuous ink system in my Epson printer.
Additionally, although specific printers are constructed to a standard, they all have a performance/printing profile that is unique to the machine itself. The differences from machine to machine are subtle and the manufacturer's ICC profile is of necessity generic for the printer model and a 'best fit'. By generating my own profiles I establish end-to-end control of the whole process from camera to print.

One point I would make, (specific to the Spyder system, since I have no experience of the other systems), is that the default/recommended monitor brightness (120cd/sqm) used for monitor calibration, is probably best reduced by at least 20% for assessing images for print. Alternately you will almost certainly need to experiment with reducing the brightness of the 'displayed' image prior to printing.
From my experience of printing, this is the single factor that the calibration process cannot address effectively.

(Harking back to darkroom processes, this is the equivalent of the exposure test print process.)

[Manfred M]

the default/recommended monitor brightness (120cd/sqm) used for monitor calibration, is probably best reduced by at least 20% for assessing images for print.

I have found exactly the same thing James as x-Rite defaults to the same value. I am thinking of recalibrating to 110 cd/m2 to see what that does. To print I just add an adjustment layer and knock down the brightness by ~ 20% and it comes out fine. This is consistent from image to image, so that factor stays the same.

[James G]

To print I just add an adjustment layer and knock down the brightness by ~ 20% and it comes out fine. This is consistent from image to image, so that factor stays the same.

I used to do the same, but finally tried toning down the brightness to 100-105 cd/sqm. The screen looked quite dull initially, but after a few days I accommodated and I no longer notice.

It was suggested to me once opening a white screen on the monitor and then that holding a blank white sheet against the screen is a useful check if is too bright. (Basically the paper will look black/very dark against the screen.) Toning down the brightness, until the sheet appears more white/gray will then better mimic print brightness of an image.

[Codebreaker]

I'll go back to what I said about print viewing conditions. I have my displays set to 120mcd / 65k and view the prints under a 65k Light with CRI of 98% - they are a very good match to the display and I never have a need to adjust the image specifically for print output.

If you use any other light source to view a print it may well throw off the appearance.

Holding paper to a screen is a no-no in my opinion. Firstly what is a white screen and secondly paper 'whiteness' varies considerably depending on type - gloss, matt, archival, etc and between makes.

It comes back to the fact that screens transmit light and prints are viewed by reflected light.

[davidedric]

That's fine so far as it goes, but my prints don't end up in galleries, they end up on my walls or occasionally a relative's. I may adjust the brightness of a print depending on where I think it will hang. And if I subsequently move it, well I can always print it again (the original will always be a virtual print proof in Lightroom.

What do others do?

Dave

[xpatUSA]

After all how does one know that the colors you see on screen in your photo editing program are correct?Your monitor colors may be out . . .

Because I don't print, Dave, I am far less bothered about monitor calibration than many here.

I trust NEC . . . but if I did not, surely I could just take a shot of the screen showing 'pure' R,G and B and see if the captured image data hues match the published primaries? A sort of round trip, so speak.

That would, for me, answer "how does one know?" without having to actually buy the stuff and mess with it . . .

By 'match' I mean to be within a 'just noticeable difference' in hue angle for each primary. What's good enough for paint or fabric manufacturers should surely be good enough for us photographers, eh?

[Manfred M]

By 'match' I mean to be within a 'just noticeable difference' in hue angle for each primary. What's good enough for paint or fabric manufacturers should surely be good enough for us photographers, eh?

I seem to remember colorimeters in use in both those industries. X-Rite's Gretag-MacBeth division has long supplied colour matching equipment to those industries.

[xpatUSA]

I seem to remember colorimeters in use in both those industries. X-Rite's Gretag-MacBeth division has long supplied colour matching equipment to those industries.

I was just explaining what I meant by "match"!

Are we saying that a 'just noticeable difference' can not be measured without the use of a colorimeter?

Are we indeed saying that one could not just take a shot of the screen showing 'pure' R,G and B and see if the captured image data hues match the published primaries?

[Manfred M]

I was just explaining what I meant by "match"!

Are we saying that a 'just noticeable difference' can not be measured without the use of a colorimeter?

Are we indeed saying that one could not just take a shot of the screen showing 'pure' R,G and B and see if the captured image data hues match the published primaries?

Ted - Having worked in the garment industry (a long time ago - I left that industry in 1988), I understand some of the manufacturing issues.

The receiving department would send samples of received goods to the Quality Control group and they would do a visual check using a standard D65 illuminant using a MacBeth colour booth using a set of GO / NO GO sample gauges. If the sample looked like it might be out of spec, the colorimeter would be used to for the final test to accept or reject the shipment.

That might have taken care of the spec, but when dealing colours and people actually buying these garments, metamerism was also an issue. Stores and people shopping don't really care how close the colours are under a standard illuminant. They make their buying decisions based on the in-store lighting, which as you can probably guess is highly variable.

[Codebreaker]

That's fine so far as it goes, but my prints don't end up in galleries, they end up on my walls or occasionally a relative's. I may adjust the brightness of a print depending on where I think it will hang. And if I subsequently move it, well I can always print it again (the original will always be a virtual print proof in Lightroom.

What do others do?

Dave

My prints end up in various places over which I have no control of the lighting, therefore I find it important to know that my prints match my calibrated and profiled screen as close as possible before I sell or pass them on.

I've seen some hideous lighting conditions in galleries, and shops and it's impossible to tailor a print for them which will then go somewhere else of which I have no knowledge.

It has to be accepted, in my opinion, that different lighting conditions may well produce different appearances. That also includes those that are on my walls, although I do control the lighting as best as possible.

[Codebreaker]

Ted - Having worked in the garment industry (a long time ago - I left that industry in 1988), I understand some of the manufacturing issues.

The receiving department would send samples of received goods to the Quality Control group and they would do a visual check using a standard D65 illuminant using a MacBeth colour booth using a set of GO / NO GO sample gauges. If the sample looked like it might be out of spec, the colorimeter would be used to for the final test to accept or reject the shipment.

That might have taken care of the spec, but when dealing colours and people actually buying these garments, metamerism was also an issue. Stores and people shopping don't really care how close the colours are under a standard illuminant. They make their buying decisions based on the in-store lighting, which as you can probably guess is highly variable.

An excellent example of metamerism and one which I use to relate to students - the reason why a lot women, in particular, will take a garment or colour sample out of the store to view in daylight. - without getting arrested for shoplifting that is :-)

[Manfred M]

An excellent example of metamerism and one which I use to relate to students - the reason why a lot women, in particular, will take a garment or colour sample out of the store to view in daylight. - without getting arrested for shoplifting that is :-)

One wedding well known wedding photographer in town told me a story of what had happened to him on a shoot.

He had shot a wedding and delivered the proofs when the mother of the bride came down hard on him for totally screwing up her wedding outfit. I vaguely remember that he said it was a brilliant fuchsia or similar colour and the top and bottom were noticeably different colours in the image. Metamerism (probably helped by dye lot differences too) at work and of course, it became a photographic problem for the photographer.

[xpatUSA]

Originally Posted by xpatUSA
I was just explaining what I meant by "match"!

Are we saying that a 'just noticeable difference' can not be measured without the use of a colorimeter?

Are we indeed saying that one could not just take a shot of the screen showing 'pure' R,G and B and see if the captured image data hues match the published primaries?

Ted - Having worked in the garment industry (a long time ago - I left that industry in 1988), I understand some of the manufacturing issues.

The receiving department would send samples of received goods to the Quality Control group and they would do a visual check using a standard D65 illuminant using a MacBeth colour booth using a set of GO / NO GO sample gauges. If the sample looked like it might be out of spec, the colorimeter would be used to for the final test to accept or reject the shipment.

That might have taken care of the spec, but when dealing colours and people actually buying these garments, metamerism was also an issue. Stores and people shopping don't really care how close the colours are under a standard illuminant. They make their buying decisions based on the in-store lighting, which as you can probably guess is highly variable.

Thank you, Manfred.

I see that illuminants and metamerism have entered our sub-discussion; while I was referring to shooting a monitor screen. While realizing nobody normally does that, I'm struggling to understand where lighting and metamerism might be involved?