What target settings for LCD monitor calibration?

[_Murat_Cuhadar_]

I have recently installed Spectraview II to calibrate my NEC LCD 2180WG monitor but have some doubts on which target settings to use as there are different opinions on color temp, gamma and intensity choices.

I use my system purely for photo post-processing and printing on color calibrated printers. I use AdobeRGB color space and have no interest for processing images for web.

- What color temp do you use/recommend? D50, D65 or something in between? I see that some folks use D65 and others swear by D50.

- What about Gamma choice 1.8, 2.2 or L* ? Same here, no common choice here too.

- Lastly what about intensity (brightness in terms of cd/mm2)?

NEC recommends Target Settings for Printing, which are D50, 1.8 gamma and Max. Intensity, but I'm not sure.

I'd appreciate if you could recommend correct settings for my type of work.

Thank You

[_Digital_Jim_]

I've been profiling about 10 LCD monitors at our photo lab for a couple years now, using Profilemaker 5 and an Eye-One spectrophotometer. I've settled on D50, 2.1 gamma, 95 Cd/m².

If this results in whites on screen being too dark compared to printed whites, use a higher target Cd/m².

If whites on screen are too blue, try a lower Kelvin temperature. If they're too yellow, try a higher Kelvin temperature.

If midtones appear too light on screen, use a lower gamma (raise the gamma if they're too dark).

Of course, this is assuming your printer is properly calibrated and you're using an accurate printer profile to soft-proof.

Hope this helps!

Jim

[Guest]

2.1 gamma? Is that a typo?

[_Digital_Jim_]

Marco-

No, it's not a typo. I set the gamma to 2.1 when calibrating our LCDs. Why is that so crazy? I've seen some post that they use 2.2, and others even use 2.0.

The bottom line is: set it to what gives the best screen-to-print match, right?

[Guest]

I didn't say it was crazy, did I?

But unusual it is indeed. I understand: you're splitting the difference.

If it works for you, fine. That's what matters.

[_Peter_Figen_]

In a color managed application like Ps, there will be virtually no visual difference overall using gamma 1.8, 2.2 or anything in between. There is, however, a real advantage using native gamma of the display, which usually IS 2.2. The advantage is that less correction is done in the video lookup tables, resulting in a monitor that has visibly smoother gradients. This is a particular problem on most LCD displays that use an 8 bit data path.

[_Digital_Jim_]

Thanks for the clarification, Peter. I'm going to recalibrate and re-profile my monitor at gamma 2.2 right now. Although it may only make a small difference, every little bit helps.

[_Andrew_Rodney_]

If the software allows Native Gamma, pick that. It might be 2.2. 2.1. 2.3, don't know (don't care). Why set an arbitrary value? Any of the above may be close to native but using a Native setting simply leaves that alone and records the value for the profile.

[Lou Dina]

I find the gamma setting for calibration definitely DOES makes a difference in Photoshop. Native white point and gamma give me some of the worst monitor profiles I have ever made, with either EyeOne Match or ProfileMaker 5, both with an Eye One spectrophotometer. I always confirm my white and black luminance and the gamma setting with a specially designed Lab target that I display in Photoshop after calibration and profiling. There is a noticeable difference in the midtones between 1.8, 2.0, 2.2 and native gamma.

I'd have to agree with Digital Jim's original post. My settings are very close to his. I set my LCD to 5100K (measured paper white), 2.0 gamma, and a luminance setting of 85 to 90 cd/m2. I have tried dozens of different settings time and again, but these settings always work the best.

I found that my CRT (now dead) worked with the same white temp and luminance, but I had to set it's gamma to 1.8 so that the same Lab test target would display the same as my LCD when set to 2.0 gamma. This is using the same spectro and the same software. And a consistent, accurate monitor output is what I am after. Besides, the monitor to print match is uncannily accurate. I used Solux and Philips viewing lights for all my work.

I agree that people should experiment and do what works for them. Try both approaches and verify it for yourself.

Lou

[Guest]

Bad post - I didn't notice the date(s)

cvt

[_Charles_Tintera_]

How did you arrive at "set my LCD to 5100K (measured paper white), 2.0 gamma, and a luminance setting of 85 to 90 cd/m2."?

Specifically how did you get 5100K as a paper white? Did you measure the paper?

Thanks
cvt

[Lou Dina]

Charles,

I arrived at 5100K by measuring a few of my favorite printing papers with my spectrophotometer. This was a good compromise (most measured between about 4800K and 5300K). I confirmed the results by trying various calibrations them comparing the monitor to prints (using custom printer profiles). Anywhere near 5000K looked great, but 5100K was the best match of the bunch. I used 5000K lighting for viewing (Solux and Philips light sources).

Luminance settings were done in a similar fashion. First I compared a ppure white Photoshop document (on the monitor) to the intensity of a sheet of white printing paper displayed under my viewing light. With the monitor luminance set to the 85 cd/m2 range, the two documents look about the same brightness of white. I try to view my prints under "moderate" light levels rahter than ultra bright levels, since I know they won't be lit by spot or flood lights when placed on a wall. If you do have spots on your hung prints, there is a case to use a higher luminance level on your monitor. Anyway, on many LCDs, if you set the luminance too high, your blacks can start to look a little washed out. When set to about 85 cd, my tonal range in the print is a great match to my monitor. If I have the monitor luminance set to 100 or higher, I find that my prints look dark and muddy in comparison (unless viewed under unrealistically bright lights). So, a lot of trial and error, plus some initial brightness comparisons.

Gamma is a tougher one. I have a special viewing target that I bring up in Photoshop (Lab based to prevent any possible conversions). When viewed in Photoshop at full magnification, it helps me to set the gamma, which affects mostly the midtone density. On my LCD, I found )again by trial and error) that 2.0 gave me the best tonal distribution. My old CRT worked best at 1.8 gamma, and other LCDs sometimes work best at other gamma settings, such as 2.2.

The above settings work beautifully given my equipment, my viewing levels, ambient working conditions, etc. My monitor to print match is excellent from both a color and tonal range standpoint. Hope that helps clarify my thinking and approach.

Lou

[_Charles_Tintera_]

I am trying to decide just how much $$.$$ to put into this. We have the Eye1 Photo bundle with the Rev B Pro sensor already. But I have not found anyway to get it to work as a spectrophotometer.

What did you use to measure your paper(s)?

cvt

[_Andrew_Rodney_]

When you say you have the EyeOne Photo bundle, that's the EyeOne Pro Spectrophotometer or the EyeOne Display-2 Colorimeter?

You can't get the Colorimeter to work this way unfortunately. Otherwise, the EyeOne Display is fine for calibration and profiling of your display.

[_Charles_Tintera_]

The EyeOne Pro Spectrophotometer, Rev B.

cvt

[_Andrew_Rodney_]

That will measure the paper white, the ambient light etc. You need either i1 Share or MeasureTool which will run without a dongle using that instrument with limited functionality (but enough to measure paper white).

[Lou Dina]

Charles,

I use an Eye One Pro spectrophotometer. If I am not mistaken, your Eye One Photo bundle includes a full spectrophotometer, right? If you can build custom printer profiles then you have a spectrophotometer (not a colorimeter) and your device is capable or reading reflected paper white.

Now you need software that can give you Lab, XYZ or direct color temperature readouts. Eye One Share (free download) can give you Lab readings, and MeasureTool can also give you Lab or XYZ readings. Once you have those, you can plug them into the CIE color Calculator at http://www.brucelindbloom.com/ and get a color temperature readout. Most "white" papers will read somewhere between 4700K (pretty yellowish stuff) to 5300K (bright white, usually with optical brighteners). Both of the above software products are made by Gretag (now XRite). There are lots of other software modules that can read your paper white, but one of the above will do the trick.

I use Gretag ProfileMaker which allows me to enter XYZ coordinates directly into the monitor profiling software as a target white setting. Eye One Match software allows you to enter a custom white temperature (in degrees K), but I don't think it allows you to enter Lab or XYZ coordinates. Nonethless, you can set the white point for monitor calibration once you know the desired target temp. From extensive experimentation, I'd say anywhere in the 5000K to 5250K area typically works fine to get the blue/yellow balance correct. I find 6500K WAY too blue for good monitor to print matching.

Lou

[_Charles_Tintera_]

Rodney & Lou - thanks!

It, the Eye-One Pro, is a spectrophotometer though it's presentation is more like the older Xrite instruments & suffers for that.

I will get Eye One Share & try that.

cvt

[_JayM_]

"I always confirm my white and black luminance"

What is your black luminance value? I'm calibrating with a new Spyder3 Elite and the menu option allows manual input of both white and black luminance. I've found lots of information in this thread and others re: approaches to setting white along with a range of values that seems to run from 85 to 120 depending on ambient light conditions, paper white, and anticipated viewing conditions for the final print. But I have seen nothing about ranges for the black level.

Any input would be much appreciated. Thanks.

[Lou Dina]

Jay,

My black luminance usually comes in at 0.2 and I set the white luminance to 85. I am using both EyeOne Match3 and ProfileMaker 5 with my LCD, and neither program lets you set the black luminance, but EyeOne Match3 tells you what it is when your profile is done.

Lou

[_JayM_]

Thanks Lou. Much appreciated. Is the black level influenced by a higher white target? I run at 110-120 for white luminance.

[Lou Dina]

Jay,

I'm not sure about other software, but with EyeOne Match and ProfileMaker, if you raise the white luminance, the black tends to follow.

For monitor to print matching, I have always found 120 cd/m2 to be way too high (unless you are using extremely bright viewing lights). I recommend trying a white luminance of 100 or lower, and for my own personal work and setup, I find 85 to 90 to be just about perfect. You want a pure white document displayed in Photoshop to have the same approximate brightness as a sheet of blank paper displayed under your viewing light.

If white luminance is set too high, you will edit more darkness into your files (since the high monitor luminance makes your files bright and washed out looking). These numbers, when sent to the printer will tend to look very dark and muddy, unless you are using a very bright light for evaluation. And if you will display your finished prints under lower lighting, everything will look pretty dark. Everything has to work together.

Give it a try and see what works for you. But if your prints look too dark compared to your monitor, then lower the luminance.

Lou

[_JayM_]

Hey Lou

The lowest my monitor would go with brightness set to 0 was 107cdm which I found uncomfortably dim. Looking at the screen now with it set to 120 it is still more gray than white so I'll have to experiment. Great advice - thanks.

[Guest]

Be aware that all LCD front button and/or OSD settings for luminance behave differently among brands and models.

120 cd/m2 can look different from another LCD because some brightness adjustment settings combine both contrast and white brightness like my 2004 20" G5 iMac.

Also the luminance setting for white according to an i1 tech support personnel I spoke with a while back said this number is an average of the entire overall brightness including 1/4 to 3/4 tonal regions of gray which influences the shape of the correction curve i1 creates to achieve a smoothly progressing grayramp within a 2.2 gamma target.

I can sort of confirm this by comparing the differences of my final i1Match correction curve applied to both the iMac's 118 cd/m2, .2BK at 2.2 gamma and my 21" Samsung 1100p CRT's 105 cd/m2, .1BK at 2.2 gamma. Both produce a very straight RGB curve but the iMac is slightly pulled back linearly starting from black point with a slight shephard's hook shaped gradual slope at the 200-255 region.

This suggests that whatever you set your display's black and white points at the software is going to correct according to an internal encoded standard appearance of what a grayramp should look like at any given luminance from black to white.

Follow Lou's direction on setting white making sure whatever settings you adjust on the front panel is enough to get you as close as possible to 2.2 target gamma and is COMFORTABLE looking. Then let the software only adjust for gamma and leave all other custom settings alone. The software will indicate if your front button settings are within specs.

Specs being 2.2 gamma and 6500K color temp. and whatever luminance number is comfortable for you.

[_Murat_Cuhadar_]

Hello Guys,

I'm still confused with this gamma choice in the Spectraview II calibration software. There is also the "Monitor Native Gamma" choice. Is it a good idea to choose this option besides 2.2 or any other gamma choice?

Thank you