Which screen?

[Ben Rubinstein]

Sorry if this is the wrong section of the forum, seemed to be the more relevant.

I need a higher end screen. 22" or bigger, doesn't necessarily need to be wide gamut (printing will never be wider than sRGB), must be designed to work with 60-80cd/m2 brightness natively, accurate colour, price up to around $1500 including calibration solution though I'd prefer nearer the $1000 mark. This is for a project using a 40 megapixel back to photograph old documents and books, mainly for digital viewing.

Eizo and NEC seem to make the most sense but I've gotten lost working out which and what, it's hard to find up to date comparisons on the web.

Thoughts?

 

[Cem_Usakligil]

Hi Ben,

Sorry if this is the wrong section of the forum, seemed to be the more relevant.

I need a higher end screen. 22" or bigger, doesn't necessarily need to be wide gamut (printing will never be wider than sRGB), must be designed to work with 60-80cd/m2 brightness natively, accurate colour, price up to around $1500 including calibration solution though I'd prefer nearer the $1000 mark. This is for a project using a 40 megapixel back to photograph old documents and books, mainly for digital viewing.

Eizo and NEC seem to make the most sense but I've gotten lost working out which and what, it's hard to find up to date comparisons on the web.

Thoughts?

I doubt whether you can find any LCD/LED monitors which work well at a brightness level as low as 60 cd/m2 . In my experience, the monitors I have worked with have exhibited accuracy/hue problems when the brightness was set to a level that low. Profiling has then resulted in a larger delta E. I think that the acceptable range for photography work would be between 80 to 130, depending on the brightness of one's work environment. I have calibrated my monitors to make the brightness of my prints match the monitors within my own work environment. I ended up using 120 cd/m2 eventually.

As you know, the best panels are the IPS ones for photo editing. You can consider many A brands which use these panels in their monitors. My personal preference would be to look at the Nec Spectraviews first. Eizo coloredge models are however equally good. Just take a look at the website of colorconfidence for a possible shortlist. There are also some very good monitors from Dell, such as the U2410 which sells for around 500 Euro. Leaves room in your budget for a ColorMunki or i1 Display Pro as a generic profiler. Just my 0.02....

 

[Ben Rubinstein]

The coloredge Eizo's are speced to be used between 60-80.

 

[Cem_Usakligil]

The coloredge Eizo's are speced to be used between 60-80.

Well, there you go then.

 

[Ben Rubinstein]

I was wondering if any others were as the cheapest coloredge is my budget and then I still need a calibration solution.

 

[Cem_Usakligil]

I was wondering if any others were as the cheapest coloredge is my budget and then I still need a calibration solution.

What about the Dell U2410 I have mentioned above?

 

[Ben Rubinstein]

I've no idea hence the questions. We're going to working in a very very dark studio so it's important. Using 120cd/m2 is out of the question, I'll use a CRT first!

 

[Andrew Rodney]

There is zero reason to calibrate a modern LCD to 60cd/m2 and I suspect few can hit that anyway. Certainly not by controlling the backlight, maybe with a huge adjustment to the LUTs. There is no such spec for this low a value.

The setting of luminance is to match a print next to the display in a viewing booth**. Even a high end now out of the box NEC SpectraView should not be calibrated new anything lower than 120cd/m2 and ideally a bit higher. If you can go higher, to say 150 or so, it means the display can compensate for when the brightness drifts up instead of down (for example when you first turn it on). Then you simply raise the print viewing conditions to match (which BTW is half power on my GTI booth so its easy to hit).

**http://www.luminous-landscape.com/tutorials/why_are_my_prints_too_dark.shtml

 

[Ben Rubinstein]

My impression was given by this article on LL: http://www.luminous-landscape.com/reviews/two-displays.shtml

Light booths aren't at all practical in an available light studio working with low powered LED lighting.

Maybe I'll just stick to CRT's. This project is working with almost monochromatic books in any case. Any more than that just isn't necessary for work where the colour will be profiled from charts in a fully static enviroment.

Not that you aren't right Andrew, you are the (world) expert on this stuff. I'm just not sure that an over bright viewing station is the answer in this situation.

 

[Andrew Rodney]

Light booths are far from real world viewing IMO. Neither are they at all practical in an available light studio working with low powered LED lighting.

Then there is even less reason why you need a low display luminance! The brightest and darkest object to view in a digital setup is the display. Having a low ambient light in the studio doesn’t mean you need a super low display luminance. Your ambient light can’t be too low but it can be too high. Since you have no print booth to view and match to the display, you can easily run a modern LCD at its near minimum native luminance, which is nothing close to 60cd/m2 just fine.

Where on earth did you get this idea of using 60cd/m2? If my old Sony Artisan (or Barco, or PressView) could only hit that luminance, it would be retired as a door stop.

Maybe I'll just stick to CRT's.

You have a prejudice for a display technology and cd/m2 setting that is based on misunderstanding. If anything, having a lower backlight for the CRT means even more the need for a dark cave ambient condition in the studio. What rational do you have for using a dim CRT of a small size and the current studio lighting having any bearing on anything as long as the studio lighting is not brighter than the display?

With my old CRT's I get 95% colour accuracy and 100% brightness accuracy for the sRBG (like) printing that I do as a wedding photographer or the B&W fine art stuff I do.

You got those values how and compared them to the so called accuracy of a properly configured LCD in what way? Where is the science? Just explain the process to get to these conclusions please.

And note, there is no such thing as an sRGB print or printer. The sRGB color space is based on a theoretical CRT of the 1990s with a very specific set of phosphors. No print process can produce this.

 

[Ben Rubinstein]

I reread your article and deleted most of my post while you were replying to it. Sorry.

Firstly I wrote 'sRGB (like)' cause of that point about printers that you've hammered into me before on a different forum

Those figures for accuracy are my own for my needs, I don't have your trained eye and you'd probably have a fit to see what I determine is accurate but it's good enough for what I do.

If I understand your article the premise is that the ambient lighting to screen ratio should be 1:1 for accuracy and then that ratio will remain a constant if you adjust the room and screen lighting together to maintain the same ratio. Is this correct? When I darken the ambient light the more into the shadows I see, and visa versa but the amount of detail in the shadows or highlights of a print is a constant.

But how do you do that when you can't up the ambient room light any further because you need a dim cave enviroment for your shooting because a light booth would play havoc with the lighting control at the camera end? You can't lower the screen brightness as you've mentioned for technical reasons. Is it a rock/hard place situation? With a CRT I can keep the ratio far lower than with the ACD screen I had set at 120cd/m2 for working in a dark room while still maintaining what was to me more than sufficient accuracy in both colour and brightness. I agree that the studio will be far darker and perhaps there is nothing I can do other than to work with the lights turned on (i.e. not at the same time as shooting) and beg for the funding to supply sufficiently bright lighting at the computer station.

 

[Doug Kerr]

Hi, Andrew,

And note, there is no such thing as an sRGB print or printer. The sRGB color space is based on a theoretical CRT of the 1990s with a very specific set of phosphors. No print process can produce this.

There is of course no such thing as an sRGB print, just as there is no such thing as an sRGB image on a luminous screen. Every self-luminous or reflective image merely comprises pixels of various colors. sRGB refers to a convention for representing colors in digital form.

But regarding printers:

Are you saying that no available printer or print process can render the entire sRGB gamut?

Or are you saying that when we calibrate (not profile) a printer with the hope that it will honor the sRGB color space that such never works? (We assume here some norm of the illumination under which the print is to be viewed.)

Or are you speaking of the dilemma of comparing a self-luminous image output (as on a display, or a projection screen) with a reflective image output (as on a print)?

Best regards,

Doug

 

[Andrew Rodney]

Are you saying that no available printer or print process can render the entire sRGB gamut?

In the exact shape and size, probably not. Its a simple, triangular shape, its based on the description of a theoretical emissive output device. If you want to get picky, there is no such thing as such an RGB printer device anyway. The point is, sRGB is a display described with three simple primary positions, white point and gamma.

 

[Andrew Rodney]

If I understand your article the premise is that the ambient lighting to screen ratio should be 1:1 for accuracy and then that ratio will remain a constant if you adjust the room and screen lighting together to maintain the same ratio. Is this correct?

No, the correct luminance for a display to be calibrated is one that produces a visual match to the print, properly illuminated next to the display. The idea is what you see is what you get.

You say you are not viewing prints, fine. In such a case, there is zero reason to calibrate to something that’s unbelievably dim and nearly impossible to hit: 60cd/m2.

When I darken the ambient light the more into the shadows I see, and visa versa but the amount of detail in the shadows or highlights of a print is a constant.

What ambient light? None should strike the display. You could work in a pitch black room, that be fine and useful assuming you don’t bump into anything on the way in. The only ambient light that is a factor in terms of display calibration targets is the viewing booth. You say you have none. Then any other ambient light just needs to be lower than the display and of course, not hit it.

With a CRT I can keep the ratio far lower than with the ACD screen I had set at 120cd/m2 for working in a dark room while still maintaining what was to me more than sufficient accuracy in both colour and brightness.

You are bringing in another term, ratio (I assume contrast ratio). That’s a totally different factor, discussed in the article and again something one sets based on the print. Otherwise it has nothing to do with the ambient light nor the luminance of the display (other than the ratio is that between black and max white).

The bottom line is, you can get yourself a modern LCD display, you can calibrate it to a reasonable cd/m2 (150) and unless you are trying to match a print, there will be no issues and there is no reason 60cd/m2 is useful, optimal or necessary.

 

[Doug Kerr]

Hi, Andrew,

The point is, sRGB is a display described with three simple primary positions, white point and gamma.

Well, sRGB is not a display, it is a convention for representing colors. Its structure of course comes from a display mechanism. A display chain that accepts sRGB data and produces the expected colors can of course be reasonably said to be "an sRGB display"; its profile would then be the sRGB profile.

And not exactly gamma, of course (but I know what you mean).

Best regards,

Doug

 

[Andrew Rodney]

Well, sRGB is not a display, it is a convention for representing colors.

Actually it is a display. Its based on a theoretical display whereby the primaries are based on a long ago used set of phosphors, the spec even defines the theoretical ambient conditions this theoretical display lives. Its based on a emissive device. So its not anything a colorant on paper device could produce.

 

[Andrew Rodney]

In terms of the history of sRGB, this very old post from the ColorSync list from the father of sRGB is interesting:

On 11/30/98 8:56 AM, "Stokes, Michael" wrote:

Jeff and Andrew,
I just ran across a couple of threads you wrote in response to a query about
sRGB. Being the color scientist behind this effort at HP, I would like to
clear up some apparent misconceptions.

1. sRGB is not base on "standard" "typical" or any other type of PC monitor,
but is directly derived from the HDTV standard ITU-R BT.709/2
2. sRGB does represent not only average PC monitors, but is within the
factory tolerances of almost all CRTs on the market today, including Barco
professional CRTs. This is due to the shared family set of P22 phosphors
which almost all CRTs use today. While this "family" of P22 phosphors has
some differences between manufacturers, these differences fall within each
manufacturer's factory tolerances. Saying that sRGB chromaticities are
"quite small" is simply saying that CRT phosphors in general are quite
small.
3. While the 2.2 gamma was directly derived from HTDV, it has been
independently verified by Sony, Barco and others to represent the native
physical state of CRTs today. It is also very close to the native human
perceptual lightness scale when viewing CRTs. This combination makes this
gamma the optimal for CRTs to physically operate at. This also goes a long
way in explaining the compatibility with Windows and PCs in general since
these systems have not imposed any arbitary or proprietary system
adjustments.
4. The white point again is derived directly from the television industry
and is the standard is televisions and also in many aspects of photography.
Achieving a bright enough D50 white point to comfortably adapt to continues
to be a technical challenge for CRT vendors.
5. I agree that there is a different in gamut shapes between sRGB and press
CMYKs. This is due the the difference in gamut shapes of CRTs in general and
press CMYKs in general. Since sRGB represents the native physical condition
of CRTs, this is an obvious outcome. I also note that the sRGB gamut in
general is significantly larger than press CMYK gamuts an most areas other
than cyan.
6. I agree that if I am in a high-end graphic arts D50 only workflow, that
sRGB is not the optimal solution and neither HP nor Microsoft claims it to
be so. On the other hand, it is the optimal solution for any display-centric
workflow such as desktop publishing in the office or home, the world wide
web or any assortment of workflows where a display plays an integral part.
7. Claiming that pure cyan in sRGB converts to 78% cyan in press CMYK is
completely dependent upon which gamut mapping technique you are using. I am
assuming you are using whatever is in Photoshop. I can tell you that this is
not the case for the gamut mapping in our own printers.
8. We've worked very hard with Pantone to provide a solid physical and
scientific foundation for their RGB representations. I am at a loss to
explain your criticims on this front and your implicit request that Pantone
base their CRT RGB palettes on something other than established standards,
physics and science. I would appreciate some input on this one.
9. Your statements that "However, I've been told that the original color
scientist from HP that
proposed this colorspace has stated that it has gone too far, that this was
a proposal ONLY for the web. . .not for printing or cameras or scanners.
I've also heard the even Microsoft is kinda backing away from sRGB for
ANYTHING other than the web." are simply untrue and I would appreciate
knowing where to go to straighten this out. I have never said that sRGB is
not for printing or cameras or scanners. I believe sRGB provides an
excellent, robust and fundamentally sound solutions for these mass markets.
HP has a lot of evidence both internally and with real customers to support
this. We also have many partners in the camera, scanner and desktop printer
businesses that have independenty confirmed this. I also believe that
Microsoft has not backed off from sRGB in any way.
10. A better web site for sRGB information is at www.srgb.com
11. I would very much like to have a discussion on the difference between
display spaces and editing spaces. I agree that a larger editing space would
be helpful, but am skeptical from my own scientific research if this can be
done in a 24bit encoding by simply changing the chromaticities without
resulting in other problems. Would you and Rodney be interested in such a
discussion?
12. Characterizing our efforts as hoodwinking seems a bit stretched since
we've gone out of our way to conform to existing international standards,
sound physics, and state of art research results.
I just wanted to clarify a few things and hope this helps,
Michael Stokes
HP
[/QUOTE]

 

[Ben Rubinstein]

Andrew, when my screen is too bright it is no longer accurate, it shows detail in shadows that are just black in print. If the ambient light is too low then the screen again looks too bright and it is no longer accurate. I think my point stands. The balance of the brightness of the screen and the brightness of the room make a difference to whether you can have WYSIWYG editing and that is my goal.

 

[Andrew Rodney]

Andrew, when my screen is too bright it is no longer accurate, it shows detail in shadows that are just black in print.

The print you are not viewing next to the display and whereby the display isn’t calibrated to view?

The display shows data that is there. You are saying that’s wrong. Seems like a false assumption! What do the RGB values read? Do you suppose those values should print? If not, then why do you see them when you soft proof (assuming you do)?

You are saying you are seeing detail on the display you don’t see in the print, therefore, its got to be the display. That fails to ask about how the numbers were printed because its quite possible the output profile is blocking up this detail. Or the device (whatever it is) isn’t behaving as the profile is describing. Or you didn’t soft proof with the profile correctly (simulate ink and paper). Or the table in the profile that describes the soft proof is out of sync with the table that makes the output values that go to the printer. You get the point. You are making a lot of assumptions about a mismatch issue. Worse, until you know where the issue comes from, you are mucking with the display in a very odd way in an attempt to presumably match this print (which isn’t next to the display). And even worse, we have no idea what conditions the display are in or what software you use to calibrate it or what instrument. The contrast ratio could be way off. In fact if we knew for a fact that the output process was sound, that be one direction I’d look at because having to lower a display to such a god awful low setting to make the display block up data that really is in the document is suspect (setting luminance to this extreme should not be necessary).

You’ve got a bigger mess to figure out and fix that can’t be resolved by calibrating a display to a condition that is basically nonsensical.

 

[Ben Rubinstein]

Lets take a step chart wedge thingy. If you pump up the brightness of the screen you're going to see more of the black steps. If you kill the lights in the room you're going to see more of the black steps. When you print however the amount of black steps you can see is a given based on your processing. It then follows that adjusting either the screen brightness or the room ambient lighting can cause innaccuracy in how you view the screen relative to the results.

Or is this something else?

 

[Andrew Rodney]

Lets take a step chart wedge thingy. If you pump up the brightness of the screen you're going to see more of the black steps.

Lets. And try this: http://www.ppmag.com/reviews/200412_rodneycm.pdf

You should see the difference between 0/0/0 and 1/1/1 in a prefect world. On the display and on the print. How close you can get is based on many factors.

If you kill the lights in the room you're going to see more of the black steps.

Only if the ambient light is not controlled properly, then you should not. And as I have already said, the brightest and darkest object in view should be the display. So now it sounds like you have another area to possibly fix, your ambient light conditions.

 

[Ben Rubinstein]

So what should they be then Andrew? I've had a system for the past 8 years where I've matched my ambient light (low, two 100 watt tungsten bulbs) to my screen (low) so that my desktop image is rendered exactly for brightness compared to the comparative print. It's been perfect for printing whether inkjet or at the lab and in pretty much every lighting that the prints are viewed in. It just works and has done for years. It doesn't work with a flat screen though, I just can't get the contrast and brightness to match print for normal viewing conditions (using a spyder 3).

Can I ask you to look at my test image? http://www.studio-beni.net/B&W3_test.jpg I've lost the original so I only have a low res version which I've used on my desktop. Picture of my wife. My system works like this. Calibrate as instructed. Adjust brightness to max. Lower until the marked triangle of light becomes black I'm at accurate brightness and that's exactly how it appears in print. However bright I shine a light on the print that triangle never stops being black. If I have the room lights on too low or the screen turned up too high then it does.

Crude of course though easy as it's my desktop picture anyway. A step wedge would do the same thing.

Am I making any sense?

 

[Andrew Rodney]

The image you provided looks totally blocked up in blacks and whites and the histogram and numbers back this up.

 

[Ben Rubinstein]

Did your test (first part), got to 10 before I saw changes but the changes from there were all uniform. Again it still matches what I'm getting in print but I'm no master printer, I print at a pro lab where consistency is usually more valued (to me) than absolute accuracy or showing a shade of red which can't be seen on screen, etc, etc.

That image is a horrifically resized jpg. That isn't the point, the point is there is shadow detail I see when the screen is too bright or the room to dark that is not apparent in print however much light I shine on it. Period.

I think therefore I'm misunderstanding you. I understand that you are saying that it is irellevant how bright your screen is if you can match the brightness of the viewing station to make the print look the same side by side. The reason that is confusing me is that if I turn the brightness of my screen all the way up and calibrate there then no matter how much light I shine on my print I will still never get it to look the same as on screen. That triangle will be there on screen and not in print (and of course the difference in the rest of the tones).

Brightening the screen for me is like applying a levels adjustment compared to print. It was the same with the ACD that I had ditto my laptop. Not just a brighter rendition of the same picture but a brighter picture period.

 

[Jerome Marot]

A quick web search shows that the target room light for photography editing is usually around 80 cd/m2, so it does not seem that 60 cd/m2 is an impossible target to meet, quite on the contrary. Maybe Andrew is confusing the numbers with the maximum brightness of LDC monitors? That figure is closer to 200-300 cd/m2.

 

[Andrew Rodney]

A quick web search shows that the target room light for photography editing is usually around 80 cd/m2, so it does not seem that 60 cd/m2 is an impossible target to meet, quite on the contrary. Maybe Andrew is confusing the numbers with the maximum brightness of LDC monitors? That figure is closer to 200-300 cd/m2.

Target says who? Again, the brightest and darkest object should be the display. So 80cd/m2 is fine IF the display luminance is higher and the lights don’t strike the display. I don’t care what luminance value, in cd/m2 or NITs you provide, it should not strike the display! In fact, the higher the ambient the light, the more spill, NOT direct light will eventually strike display and affect perception of black. None of this means you should use a ridiculous low backlight intensity for your display.

 

[Andrew Rodney]

Did your test (first part), got to 10 before I saw changes but the changes from there were all uniform.

Which sucks eggs! On a good display like the old Artisan or a newer NEC I can see the difference between zero and 1.

 

[Ben Rubinstein]

Which sucks eggs! On a good display like the old Artisan or a newer NEC I can see the difference between zero and 1.

Do your prints?

 

[Jerome Marot]

Target says who?

The screen manufacturers.

 

[Andrew Rodney]

Do your prints?

Nope. They closely match my display.