Web targeted color space

[Eric Myrvaagnes]

add.
To avoid confusion I always deliver an old fashioned contact sheet with the files to show them how it should look.
If their monitors show something else at least I'm not blaimed or passed for another photographer who delivers bad work that shows nicely on their monitor.

Excellent point (and less expensive than buying calibrated monitors for all clients.)

[csp]

5500 is sometimes used for certain offset presses and their paper.
We photographers/retouchers should stick to D6500.

why should we stick with 6500 ? can you explain this ?  most professional shot images still end on a offset press, so it seems smart
to  adopt prepress standards from the beginning to avoid troubles. 5500 - 6000 gives a much better match to paper used for proofing and printing
than 6500. handing out a contact sheet which is not produced with a similar standards than a proof is useless how do you  decide what is
correct if you have a mismatch between your print and a monitor. ever heard about  metamerism ?

[tho_mas]

most professional shot images still end on a offset press, so it seems smart
to adopt prepress standards from the beginning to avoid troubles. 5500 - 6000 gives a much better match to paper used for proofing and printing than 6500.

agreed. But it does not depend on the final output process that much. It depends on the ambient light you are working with. The white point of the display should be "neutral" to the viewing conditions. If it's a "digital darkroom" with a D50 viewing box and D50 ambient light the white point of the display should match the paper white in the viewing box (which is mostly somewhere between 5200K and 5800K and at a luminance level somewhere between 120cd/qm and 160cd/qm). If you work with soft natural daylight indoors the colour temperature is somewhere around 5500K by average. Only - and really only in this case - if you work with D65 ambient lights and/or a D65 viewing box 6500K will match the right white point.
The most important thing prior to all the ISO standards is that the monitor matches the targeted output visually. The numbers are not that important.

[Frank Doorhof]

Hi,
When we go back to the ISF norm than a display should be calibrated on D6500.
It's as simple as that, meaning if you want your shot to look accurate on the industriestandard D6500 is the way to calibrate, there's really not much more to it.

If you want to adjust for certain paper or printtypes it's better to use a profile to simulate that.
D6500 is simply put the industries standard for all displays displaying video/film/photography.
Especially with nowadays alot of photographers also doing video or having photos in slideshows displayed on TV sets or projectors or digital bill boards D6500 is more than ever important.

When you only deliver on print and everything is calibrated to one colorprofile/temparature in the end it would not matter, but as soon as you deliver a file to someone else to view or print it will.

And now comes the problem.
I've been doing calibrations since 2001 and have calibrated many thousands of displays and projectors but I'm almost the only one in the Netherlands doing the ISF calibrations, only the last 2 years more calibrators are doing the ISF calibrations.
In other words the customers will often watch your material on displays that are in 99% of the cases too blue.
HOWEVER, I STRONGLY believe that we should NOT change our work for that, bare with me.

When we look at film everything is mixed on D6500 monitors with a Rec709 or Rec601 colorspace and gamma between 2.3 and 2.4.
People with a BAD display are USED to their "problem" if you as a photographer steer away from that by subtracting blue for example the customers will see your work different than what they are used to and in that case not perfect, the really bad thing is that people who DO have a perfect setup will ALSO see you work in the wrong way.
And those people are often the ones that DO care and will make a note to it.

So instead of adjusting for bad displays I would say, adjust for the perfect display because the customer is used to how "shitty" their display looks (or even likes it) and will view your material just as bad as the rest.

[csp]

Hi,
When we go back to the ISF norm than a display should be calibrated on D6500.
It's as simple as that, meaning if you want your shot to look accurate on the industriestandard D6500 is the way to calibrate, there's really not much more to it.

so  you think it is better to fulfill a  a kind of video industry norm than follow what  organizations like swop, gracol, ugra and fogra recommend ?

[tho_mas]

D6500 is simply put the industries standard for all displays displaying video/film/photography.

That's all great but, again, it doesn't matter that much. On my display calibrated to around 5400K I see white as neutral white under my viewing conditions. When I send you one of my images you will see it as I see it (beside limitations of the monitors...) even if your display is set to 6500K... because you see the white on your display as a neutral white (if this is the case). That simple.

[Frank Doorhof]

That's all great but, again, it doesn't matter that much. On my display calibrated to around 5400K I see white as neutral white under my viewing conditions. When I send you one of my images you will see it as I see it (beside limitations of the monitors...) even if your display is set to 6500K... because you see the white on your display as a neutral white (if this is the case). That simple.

Not quite true.
The human eye is very easily fooled in what we see as neutral gray.

When you work with a 1000 degrees difference you will most certainly see the difference.
It could be that we both see the file as neutral but when you set both monitors next to each other there is a huge difference.

In my setup I have no light hitting my screen due to a hood and blinds on the windows were I do the editing.

There's really no disscussion needed for calibration to D6500 it's writen almost in stone that a display should be calibrated to that point on the grayscale.
for more info do a search on for example ISF or visit www.imagingscience.com.

Some photographers will claim that the settings/rules for TV or projectors will not apply for them as photographers but this is simply not true anymore.
As explained before most of our work is shown in slideshows, digital billboards, TV sets, projectors or many different monitors.
For the REFERENCE monitors/projectors the D6500 is chosen as the point to calibrate.
It's only natural that we as photographers/videographers also calibrate conform the industry standard.

REMEMBER this is totally different than a few years back when photographer shot film or only delivered prints.
With those workflows only the end result was show to the client and you must choose the colortemp fitting your other workflow.
Today however all the rules have changed and our work is hardly ever seen on print in most cases expect at the very end of the cycle.
When we all calibrate to the same standard (the industry standard) it would mean that everyone sees exactly the same thing.

[tho_mas]

It could be that we both see the file as neutral but when you set both monitors next to each other there is a huge difference.

Certainly. But which one is correct under which conditions. When we both set them up on a glacier they are both quite too dark and quite warm but yours will "match" a little bit better. In an average office in the afternoon mine will match better - I bet. And under D50 anyway.

There's really no disscussion needed for calibration to D6500 it's writen almost in stone that a display should be calibrated to that point on the grayscale.for more info do a search on for example ISF or visit www.imagingscience.com.

There is a lot to discuss (but we don't have to). I follow the latest recommendations of Fogra and Ugra and especially my eyes rather than the TV standard.

Some photographers will claim that the settings/rules for TV or projectors will not apply for them as photographers but this is simply not true anymore. As explained before most of our work is shown in slideshows, digital billboards, TV sets, projectors or many different monitors.

All my images look perfect on screens or beamers or other displays - when they are calibrated! It really depends on the viewing conditions, nothing else. The white point is not "in" the images (though there is a "white point tag" in profiles but this is meaningless here as it is just the illumination of the PCS)... it's just the colour temperature of the neutral white of a certain display under certain viewing conditions.

edit:

When we all calibrate to the same standard (the industry standard) it would mean that everyone sees exactly the same thing.

wrong! when we all calibrate to the same standard and arrange exactly the same viewing conditions... then we could see (more or less) the same.

[Frank Doorhof]

All my images look perfect on screens or beamers or other displays - when they are calibrated! It really depends on the viewing conditions, nothing else. The white point is not "in" the images (though there is a "white point tag" in profiles but this is meaningless here as it is just the illumination of the PCS)... it's just the colour temperature of the neutral white of a certain display under certain viewing conditions.[/qoute]

I think there is the thing that's giving confusion.
When calibrating you are building a profile for your monitor, this is not the profile you deliver your work in.
The D6500 is for the grayscale, after that the profile is build with the colors the monitor shows and that the monitor can show (with that whitepoint of course).
When you have the profile this is used against the LAB table to converce with the other profiles like sRGB/ARGB/Prophoto etc.

In practice one delivers work in for example aRGB colorspace with the correct aRGB whitepoint.
It is shown on a monitor or print with a different colorspace.
It just happens to be that for displays the D6500 point on the blackbody curve was choosen as the standard.

Again, I don't say EVERYONE should do this, but if you deliver work to other people who are viewing it on monitors or projectors you should.
Simply put, all TV programs, films etc. are mixed on D6500 displays.
Even if someone has a TV/projector that is not calibrated to that norm he/she is used to how people look on that set, when your pictures are shown it should be the same "wrong" picture that the viewer is used to.

For available light or viewing conditions you have a good point.
For me when I advice customers if they want light in their room is to use 6500 bulbs/tubes or full spectrum lights hitting the wal behind the monitor, in other words not/never hitting the screen itself. When you have available light hitting your screen you could adjust for that but remember that when you have natural light hitting your screen the quality and temperature of that light will constantly change, it's impossible to make a 100% calibration for that, that's why I always advise people to use at least a hood on their monitor and make 100% sure there is no spill light hitting the monitor.

The funny thing is that I have photographers and filmmakers (and consumers) in my client group and that the filmmakers are all 100% D6500 gamma 2.3 for Rec709 and 2.2 for Rec601 and that in the photographer poul there are still people who are in doubt between 5500 and 6500.
In most cases I make a dual setting and let them choose in the next weeks, 90% end up with D6500.

Calibrations has been my muze for some years now and I try to combine that with my photography, it's just too much fun to give up

[Frank Doorhof]

@CPs,
I checked the foga website and when you go their manual for colorcalibration/managment you can see that the ARGB/sRGB etc. colorspaces have a whitepoint of D6500.
They advise to calibrate on something else for a fixed situation (or in other words for THAT applicition it works).

I've stated in my previous posts that my story is based on the allround uses of your setup.
When I look at my own uses it's for video editing, slideshows, web and print (in house).

Only for print it would "sometimes" makes sense to calibrate on a different whitepoint.
However when you choose D6500 and make sure your whole workflow is calibrated it doesn't matter.

I print from my mac to the printer and get a correct print.
However when I edit video and play it on a calibrated projector/monitor it's also correct.

In the end it all boils down to the fact that our/your work has to look good on MANY different displays now a days, it's very logical to than choose the D6500 point.
Again when you are working in a closed down workflow with a certain type of printer/lab/papertype it does make sense to calibrate differently of course.
Again in the end it all boils down to getting a perfect result.

However I think the time of closed down workflows is mostly finished and people are delivering work that is shown on much more than one paper/display etc.

[csp]

@CPs,


Only for print it would "sometimes" makes sense to calibrate on a different whitepoint.
However when you choose D6500 and make sure your whole workflow is calibrated it doesn't matter.

as a matter of fact you can use any color temperature as long as you don't compare your screen to anything else in a dark room. but no contract proof can be used for color critical checks or adjustments under  5000k norm light  if you have set your monitor to 6500.  for me this is important .

in europe we have already a very high standard archived by eci, fogra, and ugra if you follow there advise you are  within a standard and not outside.
this is  available in german only  http://forschung.fogra.org/dokumente/uploa...uchV1_47d2c.pdf

[Frank Doorhof]

as a matter of fact you can use any color temperature as long as you don't compare your screen to anything else in a dark room. but no contract proof can be used for color critical checks or adjustments under  5000k norm light  if you have set your monitor to 6500.  for me this is important .

in europe we have already a very high standard archived by eci, fogra, and ugra if you follow there advise you are  within a standard and not outside.
this is  available in german only  http://forschung.fogra.org/dokumente/uploa...uchV1_47d2c.pdf

I know the document.
But again, remember what I said.
We use a workflow for ALL KINDS of displays, not only print/offset.

If I look at my work video is creeping in more and more, and for video it's simple D6500 is THE standard.
Also most of the work done nowadays is viewed in many different ways than print, so choosing the D6500 standard for video also for your photography is in my opinion a very good thing to do.

But again, there will be people that work mostly for print and than the story could be different for them.
In the end it boils down to getting to your client what works for both.

For me however calibrating to D6500 is the most logical thing to do due to the continues growing mix of video/photography and photography shown on normal video displays.

[tho_mas]

All my images look perfect on screens or beamers or other displays - when they are calibrated! It really depends on the viewing conditions, nothing else. The white point is not "in" the images (though there is a "white point tag" in profiles but this is meaningless here as it is just the illumination of the PCS)... it's just the colour temperature of the neutral white of a certain display under certain viewing conditions.

I think there is the thing that's giving confusion.
When calibrating you are building a profile for your monitor, this is not the profile you deliver your work in.
The D6500 is for the grayscale, after that the profile is build with the colors the monitor shows and that the monitor can show (with that whitepoint of course).
When you have the profile this is used against the LAB table to converce with the other profiles like sRGB/ARGB/Prophoto etc. In practice one delivers work in for example aRGB colorspace with the correct aRGB whitepoint.

nobody is confused here. Just saying that the white point tag in the profiles have no effect on the way there are displayed on the monitor. You can create an AdobeRGB with D50 white point and convert an image to that colour space and compare it with the original version of AdobeRGB - no difference (here). Of course I do not convert my files to the monitor profile  
Again, the white point tag in the profile is just the illumination of the PCS and does not affect the image itself.

With the new ICC V4 specifications this is history anyhow as ICC V4 dictates D50 as white point for monitor profiles. So monitor profiles created in ICC V4 all have the white point tag D50... even if they are calibrated to any other white point. My monitor is calibrated to ~5400K but the white point in the profile is D50 due to the ICC V4 specifications ( http://www.color.org/v4spec.xalter ). Eizos Color Navigator for example automatically creates profiles in ICC V4. For sure you can calibrate to D65 or whatever but the white point tag in the profile shows D50.

For me however calibrating to D6500 is the most logical thing

To me the most logical thing is to calibrate to a white point that looks neutral in a certain ambient light. And in the most common environments D65 is too cold (if you work in D65 ambient light then D65 is fine). If you work with a neutral white point and send the images to any other device in any other environment that produces a neutral white in its own specific ambient light there... then everything will look fine (as the white point tag in the profiles has absolutely no relevance here).
For beamer/cinema this is most evident as there is no reference light... because projections take place in dark (black) environments and the eyes are adapting to the white point of the projector.

[Frank Doorhof]

Hi,
Every color space has indeed it's own whitepoint, I never say they don't :-)

But when calibrating a monitor a certain whitepoint is asked, native/5000/6500 etc.
For me the video standard is the way to calibrate.

But whatever works for you is ok of course.

[bryanyc]

This discussion has been quite illuminating

I think what we can take away is that there is no agreed upon standard.  This is not surprising.  Color profiles and reproduction in print and other devices is not straight forward (even forgetting the mass of uncalibrated monitors work is seen on by consumers and AD's)    and the only thing to be wary of is anyone who gives absolute answers about what is correct.  

It is analogous to the issue of color space: Profoto or Adobe 1998 or sRGB or (remember Joseph Holmes custom color space) etc:  each of these has their advantages and big disadvantages.

Sometimes there is no standard and that is OK too.

[tho_mas]

But when calibrating a monitor a certain whitepoint is asked, native/5000/6500 etc.

no. what for? Yes, there are rudimentary calibration softwares that have just these presets. But more mature softwares allow to set x/y coordinates or different user settings for Klevin values. I edit the white point manually (but with the calibration software) to match paper white & ambient light (all D50 here) and the resultiong white point isn't even a Kelvin value... it's near 5400K but slightly shifted to green (the Kelvin values are just on a small bandwidth in the sprectrum). So I set certain x/y coordinates (and anyhow the profile white point is D50 though the real colour point of white is stored in the profile as well of course).

Sometimes there is no standard and that is OK too.

yes. There are still standards... but due to the advancements of the last years they are less strict as different workflows require different needs and especially because in front of every monitor there is a human being with its individual perception. But all the standards are defined not only for a certain device but for the entrie worflow and all devices involved. None would recommend to calibrate the monitor to D50 in a D65 environment... or the other way around.

[Frank Doorhof]

no. what for? Yes, there are rudimentary calibration softwares that have just these presets. But more mature softwares allow to set x/y coordinates or different user settings for Klevin values. I edit the white point manually (but with the calibration software) to match paper white & ambient light (all D50 here) and the resultiong white point isn't even a Kelvin value... it's near 5400K but slightly shifted to green (the Kelvin values are just on a small bandwidth in the sprectrum). So I set certain x/y coordinates (and anyhow the profile white point is D50 though the real colour point of white is stored in the profile as well of course).

It depends on the software indeed, I use 5 different analyzers two are professional ones including a spectrum radio meter.
Those are way more sofisticated than the spyder software for example.

But as mentioned I'm an ISF tech normally we also dial in the settings we want like HDTV(Rec709) CIE NTSC/PAL etc. but for grayscaling for "us" it's all D6500.
With video that is fixed luckely

With the software for PC/MAC calibration it's often very rude by only calibrating the IRE100 field manually but than again for that kind of work it works great.

[csp]

the problem in this discussion is that mr. doorhof thinks that his background in calibrating tv sets and home entertainment systems is the holy grail and video and web  as important as print.
maybe this is true for him but the majority of professional advertising and fashion photography is still printed on paper and i can't see this will change dramatically in the next years.   recommending a video standard for professional image editing  is kind of odd.

working for print accurate color is much more critical than publishing on the web where images can be exchanged in seconds or video (what is a different story anyway)  because of high production costs.  f**king it up can get really expensive and as a matter of fact when something went wrong everybody from the ad to the prepress  will first point at you.  following  the recommendations of fogra or similar organizations will keep us on the save side  and none of them recommend a 6500 white-point for good reasons.  in europe fogra and others  also have  archived a wide adopted iso standard within the printing industry which helps everybody including photographers to avoid problems.

i see it as amateurish to stick with a certain white-point anyway if it does not fit the workflow.  switching calibration settings on eizos for example s done very easy with CN and i'm sure there are other solution out.

[tho_mas]

maybe this is true for him but the majority of professional advertising and fashion photography is still printed on paper

but the question in this thread was: how to match the (uncalibrated) consumer displays! in this sense Frank's idea to calibrate to a standard that quite a wide range of displays "match" (roughly) ex factory is not wrong. Except that the viewing conditions are far different in most environments. There was a study from Kodak (as far as I remember correctly) a few years ago that the average colour temperture in typical indoor environments (offices and others) all over the day is around 5500K. In such an evironment 6500K is always much too cold. Without a kind of ISO standard ambient light I find 5800K to be quite a good compromise (even in D50 ambient light 5800K is still quite okay).
But I don't think that the white point is that essential finally as we are all adapting to the white point of the monitor within 10 minutes or so (as long as the monitor is the brightest light source). Only in direct comparisions to certain outputs/devices the white point is important (and in this case not only important but essential).

[Jonathan Wienke]

but the question in this thread was: how to match the (uncalibrated) consumer displays!

The bottom line is you can't, and it is pointless and stupid to even bother to try. sRGB is the closest thing out there to a rough average of the colors a given uncalibrated/unprofiled monitor might display; that is what it was designed to do. Educating the client(s) about the futility of trying to critically judge colors on uncalibrated/unprofiled monitors is a far better investment in one's time than trying to come up with some BS kludge that is only going to work properly on one particular monitor and make things worse everywhere else.

Export your web images to sRGB, and when people complain about the color, educate them about the importance of properly calibrating and profiling their monitors and controlling ambient lighting around the monitor. It's all you can do.