Which Monitor Calibrator is best?
September 1, 2008
Written by Rob Griffith
August is usually a good month for me in that, because of holidays etc, I have less site visits for training or consultancy and so I can actually clear the back log of testing and research jobs I’ve been promising myself I’ll do all year. One such job is trying to come up with an answer to the above question, one I get asked very frequently. To get an answer I used my LaCie 324 monitor and the BlueEye Pro software that comes with it, and a Spyder3, a i1 Display 2 and an i1 Pro. I used the BlueEye software and not the Datacolor or X-Rite software in each case to make sure I was only testing the calibration hardware and not the very different approaches the different software takes to monitor calibration.
Each calibration and profile was then tested in three ways. Firstly with the BlueEye Pro’s own test and report function, secondly with BabelColor’s Patch Tool and then lastly with some visual evaluation in Photoshop and a series of test files.
The results from the first test were very interesting. Any built-in reporting tool in a monitor calibration system typically displays some RGB colours, calculates what the profile that has just been created indicates should be the colour reproduced (in the Lab colour space) and then measures the colour actually output. It can then calculate a Delta E value – a measurement of colour difference – and hence give you and indication of how accurate the profile is (for those technically minded I used the Delta E 94 equation). The results from this first test are below:
| Device | Average dE94 | Maximum dE94 |
| Spyder3 | 0.8 | 6.3 |
| i1 Display 2 | 0.6 | 1.2 |
| i1 Pro | 0.7 | 1.5 |
As you can see the Spyder3 produced the worst figures, and the i1 Display 2 the best. Also the report includes a graph of the colour gamut of the profile and that of the Syder3 was significantly smaller.
The weakness of the internal test and report function with any monitor calibration software is that it is only testing the profile against itself, not any external standard. So for the next test I used some software from BabelColor called PatchTool and tested the profiles ability to display a series of 72 colours in the ISO Coated v2 CMYK press colour space. The results are below.
| Device | Average dE94 | Maximum dE94 |
| Spyder3 | 1.02 | 5.94 |
| i1 Display 2 | 1.29 | 7.82 |
| i1 Pro | 1.66 | 11 |
This time the Spyder3 got the better result. In each case the max dE values were typically in the dark neutral colours.
The last test I did was based around the good old mark one eyeball, all the numbers in the world are useless if when you actually start looking at images you are not happy with the results. For each profile I looked at three files in Photoshop – a gradation from black to white, concentric squares of very dark blacks varying from around 25, 25, 25 RGB to 0,0,0 RGB, and my own Colour Collective Test Form image. With the i1 Display 2 calibration active all three looked OK. The gradation was generally neutral with only a hint of colour creeping in here and there and not very much banding. I could see most of the black squares and the test image looked good. The i1 Pro results were slightly worse with some reasonably obvious colour in the gradation and noticeable steps in the shadows. I could see less of the squares and there was an obvious green cast. That green cast could also be seen on the shadows of the test form. The Spyder3 profile looked by far the worst. In the gradation there were marked steps in the shadows and a severe green cast. Only a few black squares were distinguishable and they were very green, and test form looked a bit over saturated and one of the model’s black dresses looked dark green.
Of course these results are based on one monitor, one calibration application and three individual sensors and the results on other monitors and sensors could be different. You might be surprised that the i1 Pro did so badly but spectrophotometers can have trouble reading very low levels of emitted light, and are more prone to variation, and also to be fair my own i1 Pro is a Rev A – the first version – and later versions were improved. Another factor you have to evaluate when comparing calibration systems is the ease of use of the included software. Here the X-Rite systems are also a clear winner. The DataColor software I find confusing to use sometimes whereas the Eye-One Match software gives users a nice and easy step by step interface.
And the winner is…
So, which Monitor Calibrator is best? I would have to say the i1 Display 2.
Comments
One Response to “Which Monitor Calibrator is best?”
Got something to say?
[...] For more information on our testing results with calibrators, check out this article here. [...]