Dual Monitor Calibration with Spyder2pro - procedures?

General Color & Graphics Monitor Calibration & Soft Proofing
1

Hello all, hoping to get some kind help & advice…

I have a dual monitor setup and need to get them both calibrated so that they are at least close to each other. I’m not doing the kind of graphics work that requires any real strict control over color.

I’d just like my two monitors to look pretty decent, or at least similar, side-by-side.

Setup:

Mac (OS 10.4.11) with an ATI Radeon 9200 and dual monitors

LCD #1
26" Samsung SyncMaster 2693HM - via DVI
(brightness/contrast control - no backlight control)

LCD #2
19" Samsung SyncMaster 914v - via VGA
(brightness/contrast control - no backlight control)

Issues:

LCD #1 - 26" Samsung is very bright out of the box. So blinding that it actually hurts the eyes. Using native settings this monitor looks very cold & blue sitting next to the LCD #2.

LCD #2 - 19" Samsung is a pretty good display with normal brightness but looks too warm & red sitting next to LCD 1.

Calibration:

I obtained a Spyder2pro to get these two in line with each. There is no procedure given at Datavision for the Spyder2 for a dual monitor setup so I’ve been “winging it” here.

I’m selecting a 1.8 gamma (Mac) and a 6500K white point for calibration. I’m also manually selecting the white and black luminance values. Then also using the RGB controls on the monitors along with the Spyder2pro software.

White Luminance

Adjusted mainly via Contrast Control but also affected by RGB Controls

Since LCD #2 is darker, I use the Colorimeter to take a reading of white and get the max white luminance value. It reads 135 but when RGB values are later adjusted, it drops quite a bit.

Over at LCD #1… since it’s so bright, white luminance is in the 375+ range. However, it’s nearly impossible to adjust this lower than 110.

I can’t get higher than 125 on LCD #2 and can’t get any lower than 110 on LCD#1, so I settled on a white luminance target of 120 for both.

Black Luminance

Adjusted mainly via Brightness Control

I do not quite understand the Spyder2pro software in this regard. When you select “manual”, it asks you to enter a value for both black & white luminance yet for whatever reason, it always ignores what I enter for black and over-rides with 0.30

Whatever… this is a value that is totally impossible to achieve on the 26" Samsung. The lowest I was able to get it was 0.38 or so. Trying to obtain the same value on BOTH monitors, I’ve had to settle for a value near 0.44.

RGB Controls

So then I proceed to carefully follow the software’s procedure. Both monitors set to luminance of 0.44 and 120 and RGB adjusted to obtain a 1.8 gamma and 6500K white point. According to the Colorvision “info” tool, I’m doing an excellent job with a delta value consistently less than 1 on both monitors.

Problem:

Why does it only look a tiny little bit better? Each monitor by itself looks fine but next to each the original problem is now flipped over as if I passed through the ideal setting. LCD #1 started off blue compared to #2 and now it’s redder compared to #2. LCD#2 started off too red and now it’s yellow/bluer than #1.

I’m on the right track since calibration made such a noticeable difference but why can’t I get them to match close enough to fool the eye?

After calibration, I’m hesitant to make any tweaks since that defeats the whole purpose of calibrating. On the other hand, the whole purpose of calibrating is to get both monitors to look the same side-by-side and after about a dozen failed attempts, I’m exhausted. :frowning:

Can I use the Colorimeter to make additional manual adjustments? If so, what? How?

At first I thought that these two monitors were just so radically different I was trying to do the impossible. But I don’t think so. I made the cold one warmer and the warm one cooler so someplace in between is where I would want to be.

I’ve even tried setting a lower white-point of 5800K on LCD #2 but that looked totally ridiculous next to a monitor calibrated at 6500K.

I don’t think there’s a problem with the USB connection or the Spyder device itself as multiple readings are consistent with each other. Also, I’ve performed the same procedure multiple times and get the same or similar results each time when all inputs are the same.

No direct sunlight, north facing windows 15 feet away, two 75 W incandescent ceiling bulbs, and direct Spyder contact with screen and LCD baffle installed. Also tried calibrating in dark room at night with same results. (not using “ambient light compensation” in either case)

Any suggestions are appreciated.

Thank-you!

2

I did some more research since my last posting and did another calibration today.

I decided to go with a 2.2 gamma and a 6500K white point and re-calibrated both monitors to those settings.

Based on the new gamma, my new achievable targets are:

Black Luminance - 0.35
White Luminance - 102

Yet again when I move an open window with a gray title bar and white background across both displays, it looks like garbage. The white window straddles the two displays… on LCD #1, it looks warm white with a yellowish gray title bar… and on the LCD #2 side, the white is cold and the gray looks pretty good.

I decided to take some readings using the Colorimeter tool.

In the “center” of each monitor-

LCD #1: 6450 K and White Luminance = 102
LCD #2: 6500 K and White Luminance = 106

Based on those readings, I’d think that these two should look pretty much the same but it practice, it’s not working out like that.

So then I took another set of readings near the area where I can straddle the two displays-

LCD #1 center right: 6170 K and White Luminance = 108
LCD #2 center left: 6670 K and White Luminance = 91

Wow… look at the differences! It’s no wonder LCD #1 looks warmer than LCD #2.

I put an open window in the center of each display and shifted my eyes back and forth… sure enough… they look very very close. The problem is for when windows and items straddle the two displays or when they are near the transition between the two that stark differences are seen.

The trick will be to either minimize this or figure out a way to calibrate these two screen areas to each other without changing the overall display so much that everything looks bad.

Now I really need some practical advice… how do I compensate for this? Do I attempt to do some kind of manual adjustments to match these edge areas? (I called them “edge” areas but the readings were taken about 2.5 inches in from the actual edge of the display area.)

I keep reading articles about how it’s possible to even match any two monitors, even LCD to CRT but from what I’m seeing between two relatively new Samsung Monitors, with how drastically the readings change across the screen, I’m left wondering how this can be possible.

3

Sparky, I am sure it is no consolation but I have 2 Dell monitors (20" and 24"). Despite numerous Spyder2Pro (2.3.5) recalibrations they look very different. Moving a white sheet betwwen monitors, you can really see the difference.

I am not sure I believe either settings and at least one of the monitors must be wrong. I can also see the correct profiles are loaded on each monitor too.

I find it difficult to believe that it cant calibrate two monitors to look the same - I am yet to manage it anyway - given that is its sole purpose.

I will be following this thread with interest and will report any success I have. I use Vista 64bit ultimate.

4

For me, the whole problem is the fact that the White Point Kelvin radically goes DOWN away from center of the 26" and radically UP away from center of the 19".

Last night I decided to make new Targets of 6000 K and 102 White Lum for the transitional areas of each.

To accomplish this, I had to add and subtract the offsets I discovered through manual readings and enter those as new targets and re-calibrate both.

So what I ended up with is something like this:

LCD #1 - 6280 K and 96 White Lum
LCD #2 - 5830 K and 117 White Lum

And taking readings in the areas near screen transition, both are:

6000 K and 102 White Lum

Believe it or not, a white sheet now looks pretty good moving between displays. I’m a a little disturbed by how different each monitor is now set but it looks pretty good for stuff near the edge.

I’ll have to give this a few days and see if my eyes agree over time.

Doing this is very difficult because it depends on where you take the readings, they seem to be consistent in one spot but as I move it around, Kelvin and White Lum go up and down. I don’t have enough experience to know how much of this is normal and how much is caused by the settings on the monitor.

Because the two Samsungs are so different, each one has one or two RGB and/or Brightness/Contrast settings at the extreme high or low. Point being, when I attempt to adjust the other monitor to match, certain settings may be at the extreme and unable to compensate. So much trial & error going back & forth between the two.

Here’s a side issue I do not understand:

When I move a photograph open in Mac Preview across the two displays, it seems the corresponding profile kicks in about one second later. I understand the concept of applying a monitor profile to a photograph via software. But what I don’t understand is HOW the profiles for the two displays can make the photo look even MORE different on each display.

Example, on the 26" Samsung, open a file with flesh tones in Preview. Slide it over to the 19" display and it looks fine on both up until halfway across. Then you can see the other color profile kick in and it gets much brighter with redder flesh tones! What the heck?

I thought the color profile was supposed to be doing the opposite and making the photo look the same on all calibrated monitors. And before anyone says that I have the two screens calibrated different, that’s true but this issue was clearly seen even when I calibrated them identically.

5

Sparky seems to be having a problem with uniformity on one or both of his displays. The 600 K temperature difference you described is pretty dramatic and you can definitely see this with the naked eye.

This might be a problem inherent in the display, but let me rule out something else first:

When you profile, are you following the instructions for adjusting your Red, Green and Blue levels using the buttons on the front of the display? If so, that might be part of the problem. One of the seldom documented differences between high end displays and these more basic models is the way in which the liquid crystals block the light. The cheaper displays are not as precise in how they do that.

Instead it is recommended to set the monitor’s color (RGB) adjustment to the factory default, or “native”, or whatever gives it the brightest white. Then tell the monitor-profiling software to hit the aim point you’re looking for (6500 K). That means that your video card will take the brunt of your white point color adjustment instead of your panel.

A consequence of this solution is that you might see more banding in your display. What should be smooth gradations from one color to the next (blue sky for instance), are broken up by bands of color.

In general, for everyone reading this thread - the first thing to do to get two displays to match is to concentrate on the luminance first. The color is not going to match until you’re sure you have them putting out the same luminance (brightness). As Sparky has found out, for this kind of precision, it is not enough to set a value in the software and expect it to hit it. You need to go back and measure the white and verify that it is the same, making corrections if necessary. Once you know what numbers to plug in to get you the same white on both monitors, THEN profile as normal for color.

Depending on how critical your shadow detail is in all of this, you might need to do a similar test to get your blacks the same on both displays.

Our newsletter had a couple of articles on monitors a year or so ago:
http://www.colorwiki.com/wiki/Monitors_Part_One

6

Patrick-

Uniformity is absolutely terrible on the 26" Samsung. The top of the screen is visibly darker and redder than the bottom. When I crank the brightness back up to factory default, this is greatly reduced but it’s so bright my eyeballs are burning.

Yes, you are correct, I was using the RGB controls on the monitor to make the White Point adjustments. I didn’t realize that I could skip this step though.

I was initially calibrating using the “native” setting but this was also so radically different between the two that I was forced to choose 6500 for both. When you do that, you’re presented with the instructions for adjusting the RGB on the monitor. The 26" is so blue and the 19" is so yellow/red that I’m skeptical software can achieve the correction.

So what do I tell the calibration software? That the RGB controls are NOT available? Ok, I’ll have to try this and I’m actually optimistic.

My RGB settings are so extreme that I’m sure it’s contributing to the problems I’m seeing.

Thanks for the tips… I’ll report back later.

7

Frustration levels are rising… every new solution brings a new problem.

My last method of compensating for what I’m seeing on the edges seemed to make for a nice transition but both displays seemed very uneven.

So back to square one.

All RGB, Contrast, & Brightness back to factory default. New Target of 2.2 gamma and 6500 K for both and this time let the software do all the compensations.

I picked a White Lum target of something I thought both displays could achieve (102) and just let Colorvision do its thing.

Sadly, the smaller Samsung still looks very blue! I bumped up the contrast and brightness and this time ran it though again. A little better but the same problem as before. LCD #2 looks somewhat colder and windows straddling the transition do not look too good. LCD #1 looks yellowish. Even inside a 50% gray box, it seems to have a yellow cast.

Various manual Colorimeter readings across both screens confirms that a great deal of unevenness still exists. And despite the fact that I selected 6500K as my target, this changes as I move around the screen so verification is difficult.

What’s frustrating me is that LCD #2 is normally yellow and LCD #1 is normally blue… after calibration, no matter which method, I see the opposite on both! :angry:

I’m actually convinced that only the very center of LCD #2 is matching up to the calibration and everyplace else is colder. I kinda wish the Spyder software allowed for calibration from a user defined location rather than center.

8

This is getting ridiculous.

LCD #1 - 26" Samsung set to a 2.2 gamma & 6500K white-point with a White Lum of 102. I let the Spyder2pro software correct for everything via software except for the Brightness & Contrast…

The reds are so bright, they look like pink florescent paint.

This just cannot be right.

I’m also wondering what the current thoughts are on 1.8 versus 2.2 gamma on a Mac. 2.2 looks very very harsh… the “red issue” above is just one example.

I’m inclined to go back to 1.8

9

Ok, so now I’m just setting both to factory default brightness & contrast and letting the software try to achieve a 1.8 gamma and a 6500K white point.

What a joke…

LCD #2… I give it a white background and can see a blue gradient. Blue in the upper left corner which fades into pure White by the time it gets into the lower right corner. Both Bright/Contrast were at factory default of 80 and the measured White Lum was about 118.

I cranked both Bright & Contrast up to 100 and the blue cast dissapeared… the whole background was white.

So then I ran the calibration again but only this time with the Bright/Contrast both at 100. I let the software decide where the Black & White Lum should be.

But the software is compensating for something and the blue cast always returns!

My only guess is that it ALWAYS over-rides whatever I enter for Black Lum. If I decide 0.30 is not achievable and enter 0.44, too bad… the Spyder2pro over-rides me and enters 0.30

Can somebody explain why the Spyder2pro allows you to enter your values for Black and White Luminance but then totally ignores your Black Luminance setting and enters in 0.30??

I think this is the cause of the new problem… it’s trying very hard to get that Black Lum down to 0.30 and in order to do that, it’s reducing the brightness so much that this blue cast appears.

So unless somebody has something to suggest, I give up. :angry:

10

Due to the fact that this Mac has only ONE good USB (1.1) port, I’ve been running my Spyder through a SELF-powered Belkin USB 1.1 hub. I didn’t think that would be such a problem since the hub is using it’s own power supply. Plus, I thought I was getting consistent results from colorimeter reading to reading.

Anyway, after my last failed Spyder2 calibration marathon, I discovered my powered hub has flaked out on me… I cannot get it into “self-powered” mode. It just sucks USB power from the computer which naturally can’t handle it and kills the port.

Anyway, I have a brand new Four Port USB 2.0 Adapter Card coming this Friday. Short of getting out my soldering iron to repair the old USB port on this mother-board, I can’t do any more calibration until then.

I imagine the Spyder2 will run better and even somewhat faster on a USB 2.0 adapter card anyway. At least that’s my hope.

11

It’s certainly a good idea to make sure you’ve got sufficient power to your colorimeter. The manufacturers often recommend that you plug directly into your computer, not through a hub. Since you only have one USB port, you might not have a choice but to run it through the best powered hub you can find.

My main concern is your uniformity problem. If you can visibly see a difference in color in different corners of your screen, it’s going to be impossible to get these screens to match. There aren’t any combination of settings I know of that will fix such a uniformity problem. Sometimes this is a matter of the age of the display, and the components that a particular brand chooses to build their displays with.

12

Hi Pat-

Tomorrow I’ll be putting a brand new 4-port USB 2.0 adapter card in this machine so it will not be an issue.

The 19" Samsung is about 2 years old and I can only see the blue gradient AFTER calibration. Once I reset to factory default, there is no noticeable difference from corner to corner.

The problem is when I try to calibrate to “some settings” I can achieve on both monitors, the gradient appears.

You see, the issue seems to be achieving the same Kelvin Temp AND White Luminance on BOTH monitors simultaneously. If I had to pick only one to achieve, I’m thinking about just trying to get both to 6500 K and forgetting about white luminance. The 19" Samsung just doesn’t have the same level of brightness as the 26"… by a factor of 3 or 4.

It’s all very weird and maybe I should just start fresh tomorrow with the new USB ports.

13

I finally got it!! They match!!

I switched to ColorEyes Display Pro software with the Spyder2 sensor but this procedure should work with any software. ColorEyes seemed to be much more powerful, taking much of the guesswork out of the process. One thing ColorEyes can do is that it lets you place the sensor ANYWHERE on the monitor for the calibration. With my varying tones, this was a huge advantage.

Here’s my procedure so others should hopefully be able to follow:

1. Decide which LCD monitor is weaker. In my case this was easy. The 19" Samsung cannot come anywhere near the brightness of the 26".

2. Do an automatic calibration of the weaker monitor first. I set this monitor to FACTORY default Brightness & Contrast and reset all other color options on the monitor.

I selected 6500 K with a 1.8 gamma. White & Black Luminance set to be determined automatically. This again yielded a darker blue area in the upper-left corner.

Solution: scrap the 6500 K setting and just go for “Native” white-point.

I set the Spyder sensor off-center… halfway to the screen edge towards the top and the left. After that, I ran the automatic calibration and then ran the “Gray Balance” option. I followed the directions to correct the grays and ran the auto-calibrate again.

I ended up with a nice looking display on this weaker screen.

Native white-point is around 5500 K and this was the problem all along. White Luminance ended up around 118.

3. Now I switched to the crazy-bright 26" Samsung. First I reset all monitor settings to FACTORY default. Like I said, this monitor is so bright it hurts my eyes so I made some initial manual adjustments to brightness & contrast. I first left Contrast at factory and took the Brightness way down.

Then I set-up for calibration. Under “White Point Target”… there is a handy tool that lets you “Measure with Sensor” to obtain your target points. Since I’m trying to match two monitors, I used this option. I placed the sensor on the opposite monitor, took a reading and it automatically remembered those settings for calibration. You then manually enter in your White Luminance target which is also obtained during that same reading and also the previous monitor’s calibration.

I ran the auto-calibration on the 26" monitor and for some reason it looked a little warmer (more yellow) than the 19". It also had a warmer band only across the top third of the monitor making the white Mac menu-bar look a bit red in the edges.

4. Tweak & adjust. I took some readings on both screens and although the color temperature on both was nearly spot-on 5500 K, the White Luminance was quite different.

My solution was to make more manual adjustments to Brightness & Contrast on the 26" and run auto-calibration yet again. I only had to do this a few more times before BOTH the Color Temperature AND White Luminance MATCHED on both… when this occurred… they matched! Side by side, the same desktop photo FINALLY looked the same on both! Sure, there is some slight variance from top to bottom, but holy cow, a white window straddling both displays looks like the same window now. :smiley:

Notes: I think I may have eventually figured this out with the Spyder software too but not with as good of a result. ColorEyes Display Pro allowed me to do auto-calibration from any location on the screen. This was a huge advantage as I could focus in on transitional areas between banding and thus effectively eliminate the banding.

Ok, if you’ve read this whole thread and stayed awake, you may be confused by all the various problems I’ve described… please ask any questions if you have them… I believe I have this all figured out now.

14

Very good!

You reminded me of a few things I should have thought of earlier:

  • On an LCD of dubious quality, it’s a good idea to leave the RGB controls at their factory defaults, and just allow the profiling software to do the color adjusting to the white point temperature you’re aiming for.
    http://www.colorwiki.com/wiki/Monitors_Part_Two

  • The ColorEyes Display Pro software is a great tool to try. It works with all the usual colorimeters out there today, has just about all the features you could want, is easy to use and is of very good quality. You can also try it for free. They usually have a demo you can download from their website that will run full-functional for something like 10 days.
    http://www.integrated-color.com/productdemo/demo.html

15

I do agree, factory default is the best way to go.

But in my case I tried factory default many times previously. The actual problem was that my 6500 K white-point could never be achieved with satisfaction on the weaker display no matter what. I had to just let the weaker monitor get calibrated to its “Native” white-point of 5500 K and use that as the new target for both.

The ColorEyes trial is for 10 free usages. It’s a far superior tool to the Spyder software which runs the user through too many hoops.

Again, like I said before… the ColorEyes calibration window can be manually placed anywhere onscreen. Unlike the Spyder2 software which only allows calibration from dead center.

And even after perfect calibration, it gives you additional tools to tweak the grayscale and white-point even further.

But the biggest advantage to ColorEyes Display Pro is the user feedback. You are presented with screens and settings you can play with all day long… then just hit the calibrate button and walk away. Very easy.

The Spyder2Pro software requires you to interact with it every step of the way… and then user confidence gets pretty low if you think you’ve incorrectly changed a monitor setting.

The fact that I messed with this in the Spyder2Pro software for two weeks and finally had it nailed in a couple hours in ColorEyes speaks for itself.

16

Patrick-

Is there a way to use these sensors to assist in getting the focus or sharpness adjustments made properly?

If not, is there a test pattern out there which can be used in this regard?

Thanks!

17

FYI for anyone looking for various Monitor Test Patterns…

http://www.lagom.nl/lcd-test/

18

I read somewhere that having a certain single video cards driving multiple displays can be a problem for the Mac since it’s using a single color look up table for both so if I followed the theory correctly when you profile 1 display it affects the other. I believe you would have to go with separate video cards that have separate LUTs.

19

I’m not experiencing this issue and both of my monitors are driven by one card. Each monitor has it’s own ICC profile and can be calibrated independently. After profiling, the colorimeter verifies each as correct. Macs have been pretty good about sorting out this color stuff for quite a few years…

Perhaps it’s an issue with cheaper cards.

In fact, you might be thinking of Windows XP which has a dual monitor limitation but it has less to do with the video card than the OS, AFAIK. Here’s some stuff on workarounds for XP…

northlight-images.co.uk/arti … ation.html

21stcenturyshoebox.com/cm/multimonxp.html

My biggest problem is that my two monitors are so wildly different from each other right out of the box. So it was difficult finding common settings that they both could reasonably reach.

20

Looking at the post where some users could not calibrate two different monitors . . . the problem is some older video cards have only one Look Up Table (LUT). Consequently, the calibration software calibrates the LUT for either the left or right monitor.

The answer is to get a modern video card with dual LUTs. I suspect all the newer video cards have dual LUTs.