Monitor Calibration Variables

[venchka]

I rescued another orphan monitor yesterday. A 19" model to replace the previously resuced orphan 17".

I set it up after work. Connected the video & power cables. It worked! After letting it warm up while I watched N.C.I.S., I commenced to calibrate the new arrival. The prevous 17" monitor only had two presets for color temperature and no RGB sliders. I had been using 6500K, Gamma 2.2 (PC) and had recently turned down the brightness a lot.
Much to my surprise, the new 19" monitor had RGB sliders. Life is good. I turned down the brightness & contrast about halfway on the scale. I set about calibrating to the same 6500K temperature as before. I was able to get the color temp. spot on with a luminance of 60 cd/m2.
I also found Ken Lee's notes on monitor brightness. Ken suggested aiming for a monitor luminance of 80 cd/m2.
Working my way through the calibration ordeal one more time, I discovered a second preset: 5000K, Gamma 2.2. Remembering that 5000K is close to daylight & much more normal than 6500K, I went for it. BINGO! I arrived at a finished profile of 5020K and a luminance of 70 cd/m2.
Questions:
1. Are these figures acceptable? 5020K, Gamma 2.2 (for a windows PC) and luminance of 70 cd/m2?
2. Should leave well enough alone?
3. What are the customary parameters for calibrating a monitor? I'm beginning to see that telling a rookie like me to calibrate my monitor is a bit like telling someone they need a car for the Indy 500. There are a few variables to contend with.

I haven't printed yet. I used up all my green ink recently. I did browse a few photos that I have printed recently and they look like my prints using the previous monitor and calibration variables.

Thanks for any light you can shed on this vexing exercise.

[Peter De Smidt]

In their book Real World Color Management, Bruce Fraser, Chris Murphy, and Fred Bunting say:

"The three of us, by separate paths, have come to the same recommended calibration settings: calibrate your monitor to a white point of 6500 K and a gamma of 2.2 . " p 133.

[venchka]

Ok. Thanks. I can do that. Too bad there isn't an undo button for monitor calibration.

So, by dumb luck, I arrived at the correct solution on the first try?

[Ken Lee]

For the record, I have a Mac, and all Macs ship with native support for 2 monitors. Each monitor is independently adjusted and calibrated.

One of my monitors is set to the settings that Venchka mentions, and I use it only when printing. It appears dark, compared to ordinary monitors. It mimics the brightness of a sheet of white paper in a room lit by standard illumination.

My other monitor is set brighter, to standard "computer" levels. It is much brighter than the light reflecting off a sheet of paper.

You might find this brief article helpful. See My Printer is Too Dark (http://www.kenleegallery.com/html/tech/index.php#dark).

To learn more, see CHROMiX (http://www.chromix.com).

[venchka]

Ken,
Your article is what got me to thinking yesterday. Thanks for the very helpful information. As much as I would love to keep both monitors and set them up as you describe, I just don't have the space. Unless I move my scanner or place a monitor on top of the scanner....this isn't going to work. Besides, my only use of the one monitor that I do use is for editing photos. That's it. No other "normal" computer use. I can live with a dedicated photo editing set up for now.
Assuming I can get back to where I started last night, 6500K & 60 cd/m2, is that luminance bright enough? Will inceasing the brightness of the monitor get me closer to the 80 cd/m2 luminance you mention as an optimum setting? I print on Moab Entrada Natural Rag. I'll take a meter reading of the paper by window light. I doubt that it will be as bright as some of the uber-white slick papers under room light. Or maybe I will stop worrying about it. My prints look fine to me.

ps: Both articles that Ken links to suggest 90 cd/m2 for LCD monitors. Obviously, my setting last night of 60 cd/m2 is too low.

[Peter De Smidt]

On a monitor, the "brightness" adjustment controls how dark the darks are. The "contrast" adjustment controls how bright the lighter tones are. You want to set the brightness adjustment according to your profile software's recommendation. In my case (Spyder3elite), I set it so that I can just distinguish the 4 dark squares from each other. You then control the brightness in it's standard meaning by adjusting the "contrast" adjustment. That gets you in the ball park. You then adjust the RGB channels and measure with your specrto to get as close to ideal as possible. Once you do that, you create a profile.

How bright your monitor should be depends on the level of illumination in you room, and the illumination that you'll view your print in. If your prints consistently look darker than your monitor, then you should lower the brightness of the monitor by adjusting it's contrast adjustment. If prints look too light, then adjust your monitor to be brighter. After any change to the color settings on your monitor, you'll need to re-profile it.

Sorry for the confusing terminology. It wasn't my idea. ;)

[venchka]

Peter,
I hear you. That's something else I learned in the last 24 hours.

[Nathan Potter]

OK, handy discussion. I've always wondered, how do people actually measure the screen brightness in cd/m^2?

Nate Potter, Austin TX.

[venchka]

I didn't have to measure. The calibration software did it for me.
For what it's worth, the blank white "paper" in the print module of Lightroom measured EV 3.7 with my Gossen Luna-Pro SBC in incident mode.

[venchka]

Done. 6500K. According to the Spyder2 software, luminance is ~63 cd/m2. Not exactly sure why it is that low. Perhaps because this is an old LCD monitor? Brightness & contrast dialed back to minimums to preserve white & black.

Thanks for all of the great help!

[Peter De Smidt]

Yeah, monitors get dimmer as they get older. I have an old Sony CRT on my scanner's mac. It's dim and blue. Nonetheless, using the color controls and a profile get it looking pretty good.

When you do the color controls setup with the Spyder software, were you able to match all three channels near the top of the target box, you know on the bar graph that show the levels of the red, green and blue channels? If not, are any of your color channels on your monitor at max level? If they aren't, I'd move one channel up to max on the monitor, and the others up the same amount. (For example, suppose your highest channel is blue at 95%. Move it up 5 points and then move the red and green adjustments up 5 points.) After that, check to see how the channels compare in Datacolor's software. Tweak as needed and re-profile.

If one of them is at the max, I'd make a few prints and see how they compare to your screen. If it's workable, I'd leave well enough alone. If not, you can adjust the brightness (colloquial use) in the settings for your video card. Brighten everything up a bit and then re-profile. Don't do this unless you really need to though.

[venchka]

Peter,

I have the R-G-B bars level and in the narrow horizontal target window that the Spyder software provides. The red & green are at about 43-45. The blue is about 88-89 as I recall-it was late last night.

[Peter De Smidt]

Hi Wayne,

I'd make some prints and see what you think. The bars should be level in the box, but there's a difference between being low in the box, i.e. darker, or high in the box. If you find that your prints are lighter in your viewing light than you'd like, then I'd raise each of your channels on your monitor a few points to get the bars higher up in the narrow target window in the Datacolor software. But you have a good reference point already. I'd spend some time making prints.

[venchka]

Aye! Printing is the only way to know.