i1Display Pro vs. Spyder 4 and Spyder 4 quick review
So I’ve had the Spyder 4 Elite since early January and I’ve had a few questions about it. The most common question asked is with regards to how it fairs against the i1Display Pro. In order to do a comparison of just the hardware I need a platform that supports both the i1Display Pro and the Spyder 4 otherwise the results you’re comparing will include the software that comes with each respective package as well. Fortunately Argyll CMS, an excellent open source calibration package, recently added support for the Spyder 4 which now gives me a platform on which to make a direct comparison of the Spyder 4 hardware to the i1Display Pro.
The following paragraph is for the nerds; if you’re not a color nerd you can skip to the next paragraph. Using Argyll CMS I created a CCSS file using an Eye One Pro and then proceeded to create two ICC profiles, one using the i1Display Pro, and one using the S4. For those not familiar with what a CCSS file is, here’s a quick primer. It stands for Colorimeter Correction Spectral Set. The i1Display Pro was the first to support this feature and the Spyder 4 also supports CCSS files. Both the i1Display Pro and the Spyder4 are measured for their individual response at the factory. This information is stored in the puck itself. A CCSS file contains spectral measurements (which are done done with a spectrophotometer) for a specific display. These spectral measurements can then be compared to the factory spectral characterization of the sensor. The advantage gained by doing this is that because we are using known references, we can obtain much better agreement in measurements from instrument to instrument than we could with older sensors like the i1Display 2 and the Spyder 3.
Comparing hardware only: The first thing I noticed is that the integration times on the Spyder 4 (“integration time” is the amount of time it takes to measure a patch of color on screen and report the results back to the software) are extremely slow for shadows. While better performance speed wise would be nice, this isn’t something I consider to be a big deal. Prior to the i1Display Pro, it’s probably not even something I would have mentioned however the i1Display Pro is so incredibly fast that the difference is impossible to ignore. In terms of evaluating the profiles generated using each profile, the i1Display Pro definitely has the advantage when it comes to measuring chroma in the shadows. Specifically if you use -K1.0 with dispcal in Argyll the i1Display Pro produces a much more neutral result where as the S4 has visible color casts. Before I go any further I should explain what -K1.0 means. The default configuration for Argyll is to gradually reduce the amount of chroma correction being applied (while still correcting for luminance) as it gets closer to “pure” black (it is impossible for current LCD displays to produce a truly pure black but for the purposes of this article, assume the blackest black that the monitor can produce). Before it actually gets to the darkest black, it’s actually applying no chroma correction at all. This is how most other calibration packages work as well (although Spectraview II DOES have an option that allows you to make Chroma corrections in the shadows). The reason black is often left uncorrected for chroma is because at 0,0,0 you can’t decrease 0 anymore so in order to make a color correction you must do so by positively biasing the blacks. It gets the job done but it also has another side effect which is that it increases your black point luminance, thus also decreasing overall contrast and this may be considered undesirable to some. So, getting back to the review, -K1.0 applies full chroma correction to black and in doing so it can often reveal certain limitations about a sensor which was the case with the Spyder 4. With -K0.0 (no chroma correction on for black and shadows) the performance of the Spyder 4 was much better, enough so that I would need more time to evaluate both sensors to get a better sense of where the differences are.
So is that a deal breaker for the Spyder 4 sensor by itself? Here’s the deal, like Argyll’s default (-K0.0) most packages do [i]not[/i] attempt to color correct black and the shadows. I know Spectraview II DOES have an option that allows you to do this and I actually use that option but again, most folks won’t be doing this. With that in mind (if you build profiles that do not correct black) performance is a lot closer. If you don’t care about the bundled software and you just need the sensor itself to run with 3rd party software, X-Rite doesn’t have an inexpensive puck (not that the i1Display Pro is particularly expensive considering what it does at $249) that offers the same performance as the i1Display Pro for use with other packages so once Color Eyes Display Pro, and I’ll assume BasICColor Display and Spectraview adopt support for the Spyder4 it becomes a question of whether or not the i1Display Pro is $130 better? I’m not sure of that yet. I think the safe bet is the i1Display Pro though.
Comparing the Spyder 4 Elite and i1Display Pro bundled software: As can be expected the first release of the Spyder 4 Elite version 4.5 software contained some bugs. Specifically the profiles it produced contained color casts, and the luminance measurements were incorrect as prior to measuring white luminance, the software would normalize the video LUTs resulting in a measurement that didn’t correspond to the target white point. Fortunately DataColor was very quick to respond and fixed all of these issues relatively quickly (they were much faster at addressing issues than X-Rite was when the i1Display Pro was first released). Fortunately these bugs have been fixed as of version 4.5.4. Doing a direct comparison of the i1Display Pro and the Spyder 4 using their bundled software (i1Profiler version 1.2.0, and Spyder 4 Elite version 4.5.4 respectively ), the i1Display Pro is definitely the leader both in terms of the quality of the profile it produces, features, and usability.
Let’s first take a look at the profiles it produces. When viewing a smooth gradient going from black (0,0,0) to white (255,255,255) such as the following example:
…there is a considerable amount of visible banding present in the Spyder 4 Elite generated profile. By comparison the profile generated by i1Profiler (used by the i1Display Pro) is very smooth.
In terms of features, I was hoping to see a bit more from DataColor this time around. As it stands, Spyder 4 Elite v4.5 can only produce matrix profiles, it is not capable of v2 or v4 LUT based profiles. It also has no ability to measure the color temperature or luminance of a 2nd monitor. I find this a very useful feature to when calibrating two monitors of a different make and model (or even when calibrating two of the SAME monitors). On the plus side, its interface for visually adjusting a monitor to match a 2nd monitor is probably the most advanced of all the packages and this can potentially be more valuable in certain situations as a color temp match between two different monitors (for various reasons having to do with the spectral power distribution of the backlight in different monitors) doesn’t always equate to a perceived visual match and in those situations you must make adjustments by eye. In terms of measurement speed, while it’s an improvement over the Spyder 3, compared to the i1Display Pro it seems slow (again, I don’t consider that to be a deal breaker, if it took all night and produced a better result I’d use it). It does contain a very nice suite of features for evaluating other aspects of monitor performance such as luminance, color temperature, and contrast uniformity across the entire screen. This is a very nice feature for tracking the performance of your monitor over time for example: if you took measurements every 6 months and 2.5 years down the road you see a significant change in color temperature uniformity on one side of the screen it can help you identify a potential problem so you can take action. In addition, it can also help you to confirm a change in performance, and to validate visual observations. Such information is extremely useful in getting manufacturers to move forward with warranty action, instead of giving you the run around which can sometimes be the case.
In terms of usability there’s a few issues I took with the Spyder 4 Elite’s software. Most notably the way it deals with measuring luminance is extremely tedious. It first asks you to turn you monitor’s brightness up to 100%. Perhaps there is a legitimate reason for this, but I’m not aware of any other package on the market that does this (and I own all of the major packages that are typically used by professional photographers). The problem with this is that in the next step, it asks you to adjust the monitor to the target luminance. Once a display has been set to maximum luminance, it can potentially take a full 2 to 5 minutes before the display’s luminance stabilizes again at the lower level. Unfortunately most photographers are unaware of this and as a result you may not hit your target luminance. What’s even more frustrating though is the whole process of measuring luminance and getting your monitor calibrated to your target luminance. Eye One Match, Monaco Profiler, Color Eyes Display Pro, BasICColor Display, i1Profiler and even Argyll CMS (which has a command line interface) constantly measure, and update your luminance with a display on screen in real time. All you have to do is change your monitor’s luminance until the software tells you that your monitor’s luminance matches your target luminance at which point you click a single button to move on to the next step. The Spyder 4 Elite software by comparison requires you to press a button that says “update” every single time you make a change to your monitor’s luminance, a workflow that I find to be extremely tedious. Finally when the profile is completed, there is no obvious change to the screen (aside from the fact that the patches stop changing color) that indicates that measuring is complete. The text on a button changes to say “Finish” which you have to click prior to moving on to the next step. Again, this is unnecessary. Why not just immediately prompt the user to enter a name for the profile? This serves as an obvious indicator to the user of a change in state and reduces button clicks. I’d LOVE to see all these improvements in the next update of the software.
Overall, as an entire package (software and hardware), my current recommendation for those in the market for a new calibration and profiling package will go to the i1Display Pro for non-Spectraview supported NEC monitors, and Eizo Color Edge monitors (for NEC PA series (and other Spectraview supported monitors) I recommend Spectraview II with an i1Display Pro or SpectraSensor Pro and for Eizo CG series monitors I also recommend an i1Display Pro in conjunction with Eizo’s ColorNavigator software, or BasICColor Display v5). That said, I do think the Spyder 4 Elite has a lot of potential as a stand alone package. One nice “feature” that is not often talked about is customer service and I’ve found DataColor’s customer service to be quite good. The time it took them to update and release new versions of software as new issues were discovered after the release of the Spyder 4 was quite good indeed (and I say this as a former software engineer).
Stay tuned. As time goes on we’ll be exercising the Spyder 4 Elite package, especially when DataColor releases new updates. We assume that it won’t be long before Integrated Color’s Color Eyes Display Pro, NEC’s Spectraview II, and BasICColor Display/Spectraview Profiler all adopt support for the Spyder 4.
If you found this article useful, please share it with your friends on FaceBook and Google+ by clicking their respective buttons below.