Monitor Calibration: A Software Review
An In-Depth Review of Monitor Calibration Software - by Sender Burns
As you prepare your image for printing, the first and most important step you must take to ensure the accuracy of the color and tones displayed on your screen is to calibrate your monitor.
It really is that simple. People often ask me about why their prints don’t come out as expected. The first thing I ask is whether or not their monitor is calibrated. If it isn’t, I tell them to calibrate their monitors and if they still have a problem, then we can discuss the problem further. Don’t get me wrong, there are plenty of reasons why a print may not match one's expectations; however, you have to start ruling some things out. A good place to start would be to ensure that you actually know what your images look like. Read our article on Soft Proofing and discover why monitor calibration is so important.
In order to make it easier for people to determine which monitor calibration system to purchase, American Frame hired me to test a number of the current offerings on the market and see how they work in the real world. My first selection criteria was the ability to support the latest versions of Windows 7 and Mac OSX Lion as well as Windows Vista and Mac OSX Snow Leopard. There are some older devices out that don’t work with the latest operating systems, so I didn’t include them. Ultimately, I decided upon both an entry level and an advanced outfit from each of the big players in the field: Datacolor and X-Rite. Thus the lineup consisted of the Datacolor Spyder 4 Express, the Datacolor Spyder 4 Elite, the X-Rite ColorMunki Display, and the X-Rite I1 Display Pro.
If you just want to know which one I recommend for all but the most advanced users, allow me to cut to the chase: I recommend the X-Rite ColorMunki Display. It is a great combination of simplicity and power. The fun name speaks to the light hearted nature of the package. The software is by far the most well refined and easiest to use of the bunch. The software walks you through each step of the process and offers a very well made video tutorials, should you need them, for nearly each step. The results are also right on. From install to complete calibration can be achieved in less than 15 minutes. At this writing, the X-Rite ColorMunki Display retails for $199, but can be found for around $175.
Let’s begin with a basic understanding of what a monitor calibration outfit is and what it does. There are a just a few companies that make packages to calibrate monitors. The packages always consist of software and hardware in the form of an electronic device called a colorimeter or spectrophotometer that measures how color and tone are displayed on your monitor. All the packages here come with a colorimeter. A colorimeter differs from a spectrophotometer in that it can only measure transmitted light whereas a spectrophotometer can also measure reflected materials. This means that a spectrophotometer can be used to calibrate and create printer profiles, whereas a colorimeter can only calibrate and produce monitor profiles. That being said, this allows colorimeters to be tailored to the task. I wouldn’t dwell on the details of it other than to say that all the kits I am reviewing only allow you to calibrate your monitor.
Here you can see the ColorMunki positioned on a laptop for taking measurements.
All these packages have the same goal: to remove subjective human judgment from the process in an attempt to make monitor calibration truly objective. Essentially, the software displays color swatches on the screen and the device records the value of the color that is displayed on your monitor. Thus, it discovers the limits of the monitor. They also measure the brightness and tone response of the display. The software can then direct you to make changes in the brightness and contrast of the monitor as needed. Depending on the monitor used, the software can actually adjust the brightness of the monitor for you. This is a really nice feature as it makes everything truly automatic (and fast!).
The software then uses this data to create what is known as a monitor profile. This profile is placed into the appropriate folder of the operating system where it is utilized to ensure accurate color and tone.
What does it actually calibrate?
There are three main characteristics that need to be set when calibrating a monitor.
White point is the color of white. That may seem like a silly statement. Of course white is white, right? Not really though. Look around you. Look at all the many different items that we call white. Notice the slight differences in color? Which one is truly white? Well, that depends on what white you want! While there is no industry standard, there is industry consensus that a white point of 6500 Kelvin (or 65k which is its close cousin) is a nice target value. It is an ever so slightly blue white. All these packages shoot for this by default. Since all colors relate to how white looks, it has a major impact on how all the colors are displayed. Many monitors come from the factory set to overly bright and too blue. Often after calibration, people find their monitors to appear dark and yellow.
Another key setting is gamma. Gamma is the setting that determines how the monitor will display the contrast of the image. It is a function of input to output. The human eye doesn’t see light in a linear fashion. A gamma function is used to map linear input data to a non-linear output that looks natural to us. If you have a background in black and white photography, it helps to think of the gamma as being similar to contrast graded printing papers. Depending on the grade of paper selected, an image printed from the same negative will have more or less contrast than when printed on a different grade of paper. Similarly, when a digital image is viewed with a Gamma of 1.8, it appears to have more contrast than when viewed on a monitor with the Gamma set to 2.2.The industry consensus is 2.2. All these packages shoot for this by default. If you are an advanced user and need to select a different gamma, then go with the Spyder4 Elite of the X-Rite I1 Display Pro.
Brightness, also referred to as Luminance, is the last criteria. The default target value of the X-Rite software is 120 cd/m². Unless you have a reason to change it, I suggest you stick with this. By default, X-Rite’s ColorMunki Display and X-Rite’s I1 Display Pro will use this as a target value. By default, the Datacolor units will not change the brightness of the display. I don’t like this. The number one complaint about prints is that they appear too dark in relation to the image on the monitor. This is largely due to the default brightness of monitors being much too bright. It has been my experience that most monitors are set at upwards of 250 cd/m² by default. Setting the brightness in the range of 120 or so will generally result in a darker monitor. A brightness of 120 cd/m² is bright enough to comfortably edit images and for general use without being too bright. It will also go a great ways towards achieving a screen to print match. I generally set my brightness to 120 cd/m² on all my monitors. On a side note, many students have told me how much nicer it is to work on a calibrated monitor as it isn’t so hard on the eyes. This may seem trivial, but if you work on the computer all day, you will appreciate the ‘kinder’ on your eyes look of a calibrated monitor with reduced brightness.
In general, these packages begin by asking what target values you want to use. Once selected, the software will prompt you to set the measurement device on your monitor. To keep it there, there is a counterweight on the wire that can be adjusted. You have to pull on the counterweight of the Spyder units to move it. I don’t like having to pull on any wire; ever. On the X-Rite products, there is a little tab that you hold down which allows you to freely slide the counterweight.
Be sure to have your monitor on for at least one half hour before trying to calibrate it to ensure that it has warmed up and reached stable color. Also, be sure not to that you don’t have any bright light hitting the monitor. Move any lamps that may be beaming onto the display. I suggest that you work in an area with reduced brightness if possible (if nothing else, don’t turn all the lights on and open up all the blinds).
I began this review with some biases based on experience. I have owned and/or used a large number of monitor calibration packages over the years. During most of this time, my go-to device has been an X-Rite i1 Pro that I used with Gretag Macbeth’s i1 Match for nearly a decade. Recently, i1Match was replaced by i1 Publish/Profiler. The X-Rite i1 Pro device is actually a spectrophotometer as opposed to all the devices in this review which are colorimeters. This means that it can measure both reflective and trans missive light. Functionally, that means it can be used to create both monitor and printer profiles. It was around $1,500 when purchased new (it has just recently been replaced by the i1 Pro 2). I like the system and I am comfortable with its use.
I have also used an X-Rite Display as well as the X-Rite Display 2 a number of times.
I also had an X-Rite PrintMunki for awhile. It is an entry level spectrophotometer system that allows for the creation of printer and monitor profiles. I never really warmed up to it. I thought the device itself was very fussy. Let’s just say; neoprene and a zipper were needlessly involved. It never sat well against the monitor. I thought the results were good, but I just never liked using it.
I have read much about the new Datacolor Spyder 4 line up and was excited to try it. For years, I used Datacolor’s/Colorvision’s OptiCal, with an early Spyder, in a school where I taught. It worked fine. I was interested in seeing the latest and greatest version.
Let’s get to it
All the colorimeters are surprisingly small. They all appear well constructed. The Spyder 4 devices appear identical to one another aside from color. Their specs imply that they are the same exact hardware, but I don’t know this with certainty. The X-Rite devices also look to be the same, but my understanding is that the hardware housed within their identical shells that vary a good deal. The I1 Display Pro is supposed to be much faster.
1. ColorMunki Display
From install to completion, the Colormunki was smooth sailing. If you are connected to the internet when you install the software from the disk, it checks to see if there is a newer version of software. If there is, it prompts you to download the latest version and installs that instead of the older version on the disk.
As you can see from the first window of the software displayed above, there is an easy workflow and an advanced workflow to choose from. The Easy workflow just assumes you want to use the white point of D65, a brightness value of 120cd/m², and a gamma of 2.2. In the advanced workflow, you have a few choices for white point and brightness. Unless you have a compelling reason to change these, I suggest you leave them as they are.
Under White Luminance, there is the option, “I would like ColorMunki to automatically determine the optimum luminance for my display based on my ambient light conditions where I view my printed output (Recommended for display to print matching)”. This will prompt you to move the cover of the colorimeter to the ‘covered’ position so that it can measure ambient light falling on it. As I understand it, if you are in a brightly lit room, it will set the luminance rather high so that the print when viewed under these bright conditions will closely match the monitor. Not a bad idea, but I don’t think I trust it in practice, so I left this unchecked. There are a number of reasons for this. The first is that it is better to control your lighting situation than to crank up the brightness of your monitor to deal with poor lighting. Secondly, I work on a number of computers including laptops. I like to have my display appear as consistent as possible across different machines. Lastly, if you setup up your laptop like that and then take your laptop elsewhere, a likely scenario, then it is no longer ‘optimally’ set.
Similarly, I don’t use the Flare Correct option, so I left that box unchecked. I don’t know exactly what it is doing, and I know flare will change as the light conditions in a room changes. If you can’t set up your room to reduce flare, then buy or make yourself a monitor hood.
So, I chose to stick with 120cd/m² rather than the auto-magic settings.
Notice the little icon of the movie clapboard and the question mark? This will launch a series of well made video tutorials detailing what each setting does. The windows can be resized as needed to accommodate monitors of varying resolutions and aspect ratios (NOTE: The minimum resolution is 1024x768).
The rest of it is very simple. You just place the device on the screen and let it do its thing. I would say it took around 5 minutes, but to be truthful I didn’t bother timing it. It is fast enough not to be overly burdensome. I calibrated a variety of old and new computers and found that it was able to adjust everything itself without my needing to touch the monitor. If you are using an older monitor connected via a VGA cable, you would have to make some adjustments to the monitor itself. This isn’t likely though.
The final window of the process provides you a number of images that you can see with your settings before and after calibration. This is supposed to validate that your monitor is more accurate post calibration. This is the one area that needs improvement. I would like to see objective, numerical results of the calibration process rather than essentially asking me if I think they look better before or after calibration. While I am confident that the calibration was successful, I much prefer objective validation. As I noted earlier, people often find their monitors to be significantly darker and more yellow after calibration. This leads to them questioning whether or not they did it right. If you are like me and really want the piece of mind provided by numerical results, then spend a few much dollars and get the i1 Display Pro.
Once you hit the Finish button, the software closes.
No doubt about it, the X-Rite Colomunki Display is the easiest of the bunch and the one that I would recommend for the average user. You can have the software installed and your monitor calibrated within 15 minutes without reading the instructions.
2. i1 Display Pro
The i1 Display Pro uses the same software that their higher end spectrophotometer based solutions use, namely i1 Publish. These means that you are presented with options for things that aren’t available to you, like printer profiling. This can be a little overwhelming. If you are used to i1 Match, the predecessor to this software, you may feel that this software is a step in the wrong direction and long. I know I did. However, the functionality is greater than what iMatch offered and you eventually get used to the interface (though it still needs some improvement). It also allows you to skip steps and get through things quicker once you are familiar with the software.
iPublish offers far more options than most users will ever require. Not only are there tools to calibrate your monitor to any setting imaginable, but there are also tools here to determine the overall quality of your monitor. This includes tools to analyze how the brightness and white point of your monitor varies across your screen (see the two images below which show the results from an older 17” Macbook).
It can also track the results of your monitor after calibration (see below). This would allow you to note any radical or erratic shift in the performance of your monitor. It will also allow you to note a decrease in its performance indicating its decline.
It will let you compare the accuracy of your calibrated screen to specific industry color comparison and show to you to what degree they vary (below).
You can decide if you want your profile saved an ICC version 2 or 4 compliant format.
I am sure there are a number of users for whom all these features are very helpful. For the average user, these may be more interesting than useful. I confirmed that the brightness of the display on my older Macbook Pro is inconsistent from area to area. I already knew this to be true. I am not going to throw it away after receiving numerical evidence of this. However, if I could test the screen of several laptops before buying them, the results would likely influence my purchasing decision. It would be great if those who review monitors start including the results of these tests in their reviews.
Here is another view of the resulting profile data. Other than looking pretty, I really don’t see the point of this. Without the ability to compare the gamut to another, I think it’s just silly. You can use the Colorsync Utility on Macs to compare two different profiles. That’s much more useful.
While this may not initially look like there is much value in seeing this, over the years, I have found that a very nonlinear looking curve can indicate that something is wrong. When that has happened in the past, I have tried things like updating the graphics card driver and ensuring that the monitor was reset to defaults. That usually resolved the issue.
The biggest reason to purchase the i1 Display Pro over the ColorMunki Display is pictured above. It is the piece of mind that the confirmation window offers. While the introductory kits provide you with a before and after image, this and the Spyder4 Elite are the only two that actually provide you with numerical confirmation of the results. I do wish it indicated the information about gamma here as well.
Above are the results from calibrating an older LCD. It validates that I was able to achieve the desired white point and a luminance level that I am happy with. Notice that it wasn’t able to achieve my target luminance of 120cd/m². That is indication that my monitor may be near the end of its useful life for critical editing.
The other feature that I really like about the i1 Display Pro was the ability to use a larger set of color swatches. This should allow a more accurate profile to be generated. If you aren’t in a rush, you can select a much larger sample size and let it do its thing while you have a sandwich. Below are the results of using a large sample size.
If you can get past the complex interface, you will be rewarded with a very powerful monitor calibration and testing outfit.
3. The Spyder 4 Express
The Spyder 4 Express using a step by step process with a series of checkboxes to get you started. Like the Colormunki, by default the program will try to achieve a white point of 6500k and a 2.2 gamma. The difference is that with the Colormunki, you have a few choices for the white point. The only option for the Express is that you can opt to use the native white point of the monitor (it is a ‘hidden feature’ available in the Preferences window). Using the native white point of an LCD is supposed to allow the greatest color range from a monitor. However, it may not give you a pleasing white. I always elect to use 6500K/D65 for my white point.
In terms of brightness control, the software asks the user to ‘Adjust the Brightness to the level at which you are comfortable’. This is asking for trouble. It is introducing subjectivity into what should be an objective process. It is very likely that the novice user has been staring at overly bright monitors for their whole life. ‘Comfortable’ is likely much too bright. The other issue is that you can’t set multiple computers to the same target values. I like having all of the computers I work with set to the same values so that images appear the same across all the computers.
One nice feature of the Spyder Express software is that there are optional netbook controls in the form of a little floating palette that pops up with a ‘Back’ and ‘Next’ button (see image below). This allows you to control the application even if part of the screen is cut off due to the netbook’s screen being a lower resolution than the window of the program. Of course, it would be preferable to just allow the program window to be sized dynamically, but that isn’t the case.
I read that it could be used to calibrate Netbooks with resolutions as low as 1024x600. Given that this is the highest resolution of my older Asus EEEPC, I decided to give it a shot. The lowest part of the screen wasn’t visible, but the netbook controls allows you to navigate things okay.
The software is pretty straightforward and it walks you through the whole process. Yet, after completing the calibration the results were terrible. There is a “before” and after “setup” similar to that of the ColorMunki Display (see below). The “before” results looked much better than the “after” results. I ran it again multiple times. Still, terrible results.
The problem is, I knew the results were terrible, but the software didn’t provide me with any numerical evidence to back this up. It doesn’t say anything like, “Sorry, there seems to be a problem, please contact tech support.” Instead, you have to be able to recognize that there is a problem visually. Since, we are using these packages to get objective results, I wish there was some type of objective measurement to the results it was gibing me to better understand the problem.
I could tell that things were way out of whack; however, my concern is that the novice wouldn’t be able to tell that things were off and would accept the poor results that I initially obtained.
I wrote Datacolor tech support and sent them screen grabs of my results.
They responded, “First of all, please be aware that Netbooks are not really made for serious photo editing. The reason is their low quality. For example, you can see in the gamut graph, they hardly can reproduce 50% of the Adobe RGB color gamut. So please do not expect "miracles" from your netbook's display, rather get a better quality display, then give it another try. However, please make sure that your netbook has the very latest video card drivers installed. Also let us know what kind of video card is being used in that computer, that way we can let you know how to deactivate video card tools which could have a bad effect on the calibrated view of that netbook display.”
I am well aware that netbooks aren't designed for serious photo editing; however, the program should allow me to see colors accurately within the limited gamut of the device. This isn't the case though. The color was much more accurate without using the profile that the Spyder software generated. Gradients were not even linear after calibration. I was using the latest driver for my video card. This is the driver that was in place when I calibrated my monitor. I wrote back and told them as much.
“This is a problem forced by the graphic drivers. Their control panels need longer to boot than the SpyderUtility start-up which gives them the chance to overwrite the data we just wrote to the lookuptable of the graphic adapter. Depending on your graphic card you can disable the control panel. For Intel-Cards it's a program called either - hkcmd.exe - igfxtray.exe or - igfxpers.exe.
All these programs will be started via the registry, so you'll need a startupmanager to disable them, or you can use the windows-tool "msconfig". Once you have disabled the control panel, reboot your computer, than run another calibration with the white point set to native.
To find the hidden feature to change white point to native in Spyder4Express, on top in the program window, go to: Spyder4Express - Preferences. In the window which is now appearing, select "LCD native" and confirm that window with OK.”
I don’t think the novice, for whom this package is designed, wants to go through all of this. I didn’t want to go through all this. A novice may not even understand what/how to do what they suggested.
I took their advice and disabled all the Intel utilities in the startup manager. I then restarted and recalibrated using the LCD native white point as suggested. The results were much better than before, but there was definitely a warm color cast in the darker tones. So, I ran it again not using the LCD Native white point so that it would try to achieve a white point of 6500k. The results were much better; however, there is now a very subtle green cast visible in the mid-tones. I am sure of this. However, there is no objective way to ensure this, happens, as this outfit doesn’t provide you with numerical indications of its success (or lack thereof).
The final window of the process provides you with this Profile Overview. It tells you how much of a particular color space your monitor is capable of reproducing. Here, we see the monitor can display 60% of the AdobeRGB color space. While it is interesting to note this and it may be of value if you are trying to determine which of your monitors has a larger gamut, I would find it much more useful to see the numerical results of the calibration process. Namely, was the desired white point and brightness achieved? That would provide me with some piece of mind.
In short, I don’t trust it. As its name implies, this is supposed to be a fast process. It wasn’t. In fact I spent more time fussing with this than any other of the outfits (and any other monitor calibration package that I have ever used).
4. Spyder4 Elite
The Spyder4 Elite is two steps above the Spyder4 Express in the Datacolor lineup. There is a Sypder4 Pro package that slots in between. The Spyder4 Elite software is very similar to that of the Spyder4 Express. The Elite has many more options though. For example, you can select the white point, gamma, and brightness target values.
Additionally, there are an incredible number of advanced features such as the ability to check your calibration (see image above). This is a very quick process. If you don’t need to recalibrate, then you can exit the program. If you do, you can initiate it right after the check.
Here, you can see where I was asked to manually adjust the monitor brightness in order to achieve my desired brightness value. After I adjusted the brightness, I had to hit the Update button at the bottom of the screen. I also calibrated the same laptop using X-Rite’s ColorMunki Display. The ColorMunki was able to automatically adjust the brightness.
Notice the button is partially cut off. That is how it appeared on the 15” Windows laptop that I was calibrating. You can’t adjust the size of the window within the program. The software should allow the windows to be readily resized to fit.
The results window is similar to all the others in that you can toggle between a “before” and ‘after’ view.
Spyder Tune allows you to tweak the White Point, Gamma, and Brightness of your monitor post calibration. I don’t understand why anyone would want to use this or why the options are available. Why create an objective profile than change how the monitor looks afterward? This may serve a need for someone, but I can’t wrap my mind around it.
Like the X-Rite i1 Display Pro, you can see the results presented in number of different ways. Here, you can see the results in numerical form. I like that the Gamma is included here. Unfortunately, the White Point is presented only in hard to comprehend CIE xy values with Kelvin scale value.
Just like the Spyder4 Express, you can see the profile as a 2D graph to compare it to other standard color spaces. In Elite, you can also compare it to other profiles on your computer.
I ran the program on a number of computers without any problems. I was pretty happy with its performance and feature set. Then I decided to run it on a three year old MacBook Pro. First, it presented me with an error which seemed to indicate that it couldn’t automatically adjust the brightness. I would hit OK, but it was stuck in an infinite loop of errors. I had to force quit the program to get out of it.
Then, when I ran it again, I got this error. I sent a message to Datacolor about these problems.
“Please always plug the Spyder into an USB port in the back of your computer (close to the keyboard/mouse). Please do not use USB-Hubs and/or USB-extension cords as this could result into problems with the power supply.
Also check that you have turned off all screen savers and energy savers during the calibration. Then please make sure that the Spyder lays flat on the screen in the center where the patches are displayed and try to prevent direct bright light (i.e. sunlight, desk lamps, etc.) from hitting the screen.
Also double check and make sure that you have turned off the auto brightness adjustment of your Mac. Also give it a try and turn the ambient light feature in Spyder4Elite off, then give it a try and let us know how it went.”
None of the suggestions applied to me. In fact, I stated in the problem ticket that it was directly plugged into the laptop. I tried it again after a restart. It worked fine.
In testing all these devices, I made use of well over 9 different computers running a variety of different operating systems including OSX 10.6.8 thru OSX 10.7.2 and Windows Vista 32 bit thru Windows 7 64bit.I didn’t have a single problem with the X-Rite packages. I had numerous problems with the Datacolor devices. In fact, in preparing this review, I had to contact Datacolor tech support no less than 5 times (none of which resulted in satisfactory answers). If you are a very advanced user who is comfortable with some trouble shooting, then this may not be a big deal to you. However, if you are a novice, these troubles will likely turn you off to monitor calibration all together and that would be very unfortunate.
Monitor calibration should be a relatively simple procedure. These devices have now been around well over 10 years. If the general imaging public is going to integrate monitor calibration into their standard workflow (which they should/need to do), then these things have to work well, and work simply. I didn’t find this to be the case with the Datacolor products. Keep in mind that I was really looking forward to working with their new products. I have read other very favorable reviews. I can only speak to my experiences. I didn’t find the results to be consistent. As I write this, I am looking at two different monitors calibrated with Datacolor products set to the same white point. The one definitely has a slight green cast to it. Another thing that doesn’t sit right with me is that it says my 5 year old budget netbook is capable of reproducing 67% of the sRGB spectrum but my 3 year old Macbook Pro is only capable of producing 62% of the sRGB spectrum. That just doesn’t add up.
The X-Rite ColorMunki Display gives a great combination of simplicity, ease of use, and effectiveness, at a reasonable price. For most people who just want to ensure that they are seeing accurate color on their monitor, this fits the bill perfectly. For those who want more from a monitor calibration package such as numerical confirmation, the ability to track the condition of the monitor, the ability to calibrate dual monitors, the ability to select different target values for white point, gamma, and brightness, then the i1 Display Pro is the way to go.
About the Author
Based in the Midwest, Seder Burns is an artist and professor of art. His primary medium is photography; though he also works in a variety of other mediums. He has been teaching courses in digital art and photography since 1999. He is currently a Lecturer of New Media at the University of Toledo in Ohio. An undergraduate alumnus of U-M, he received his M.F.A. in Digital Arts in 2009 and M.Ed. in Career and Technical Education in 2006 from Bowling Green State University. Since 2004, he has worked as a consultant on issues of digital color management, workflow, and printing.