Datacolor Spyder 5 review
Why calibrate your display
Everyone who interacts with color should work on a properly calibrated display. Period.
Your display is the starting point of your digital darkroom. It’s the only interaction between photographer and the image after its inception on the camera. An uncalibrated screen is an unknown variable. Your images or designs can vary wildly on color from one display to another, and even more during printing, where any misstep can represent a big waste of money. Calibrating your display ensures that what you see is consistent from day to day and also follows a known standard. The same logic applies not only for photographers, but also for users in other fields that require consistent color — for example, web, graphic and product design.
The resulting color profile is used by your operating system and any color managed program to adjust its output, making sure that what you see is as similar as possible to other displays, given their intrinsic differences. These profiles are also used to match colors between other color managed devices. For example, if you have custom color profiles for your LCD display and photo printer, Adobe Lightroom can understand their differences and match their output as closely as possible, or simulate the printer output on your screen.
Datacolor has been improving their sensor design for more than a decade. I’ve used all versions of it, from the original Spyder, back when the company was still called Colorvision, to this latest 5th generation. I appreciate the fact that they move fast and listen closely to the market demands. The Spyder3, for example, was the first consumer grade colorimeter that worked fine with wide gamut displays. Spyder4 focused on measurement precision and less unit-to-unit variance.
One of the biggest criticisms of the Spyder3 and Spyder4 colorimeters was that measurement performance was not so great on the darker tones. The updated Spyder5 sensor features a new 7-color matrix that is claimed to offer up to 55% improvement in tonal response, especially in the shadows. The improvement is very apparent when you compare it to previous versions and also reflects in a faster calibration time, since the updated unit can get a precise measurement in less time than its predecessors. Profiles generated with the new Spyder5 sensor have a little bit more shadow detail and less banding, comparing favorably to the X-Rite i1Display Pro, which is arguably one of the best on the market.
The colorimeter now has a more compact design with an integrated protective cap that doubles as a counterweight. When closed, this cap protects the color filters inside the unit from overexposure to light, which can degrade them and lessen the unit’s accuracy. This was a very common problem, for example, with the old X-Rite Eye One Display 2 and other colorimeters of that generation.
This updated design is elegant, compact and works well. What I didn’t like in my testing is that the cap/counterweight fits very snugly on the USB cable, and you have to pull it with more force than I’m comfortable with in order to adjust its position. My fear is that doing this multiple times can cause the cable to fail. In the end I opted to leave a reasonable loop of cable between the cap and the main unit so I don’t have to adjust it every time.
The new sensor also has an updated baffle design that uses a deeper honeycomb pattern in order to eliminate any stray light influence on the measurements. It works very well, with no discernible changes in measurements even when pointing a bright LED flashlight directly to the sensor.
Spyder 5 Express vs. Spyder 5 Pro vs. Spyder 5 Elite — Choose the best version for your needs
All three versions share the same sensor and produce color profiles of identical quality. The least expensive Express version lacks ambient light measurement, which is a feature of limited use, in my opinion.
Other than this, all differences between the three packages lie in the software, which is intentionally capped on the lesser models. This makes the Spyder5Express a great deal if you can live with its software limitations and the lack of ambient light measurement. Keep in mind that some of those extra features, especially in the Elite version, are of very limited real world utility.
You can find here a detailed table with all differences between the versions.
This is the absolute best display calibrator you can buy for a low price, especially if paired with the open source dispcalGUI and Argyll CMS package. Original software offers only one color temperature and gamma choice: 6500K and 2.2. Not coincidentally, this is the most used setting and should work well for the majority of users, including those working with design and photography. Ambient light sensor is present, but disabled, and there’s no recalibration option, so you have to perform a full calibration every time (about 5 minutes vs. 2.5 minutes).
The Spyder5Pro version is the most well rounded package. It offers the most popular choices for color temperature — 5000K, 5800K, 6500K, native and gamma 1.8, 2.0, 2.2, 2.4. Ambient light measurement and a quick recalibration option are also available.
Compare with X-Rite Colormunki Display.
The top of the range Elite version adds features more suited to the hardcore color geek and edge cases. Datacolor tries to add value to this version by adding lots of features that are, in my opinion, useful for only a small subset of users. What really matters is:
- Front projector calibration
- Custom calibration targets for video professionals, such as NTSC, PAL/SECAM, ITU-R Rec.BT.709, ITU-R Rec.BT.2020, Cineon and L-Star*
- StudioMatch, which allows you to tweak the profiles to match multiple displays connected to the same computer
Cross shop this version with the X-Rite i1Display Pro.
So which version should I buy?
- Spyder5Express if you want to stick to 6500K and gamma 2.2 or use with dispcalGUI and Argyll CMS.
- Spyder5Pro is perfect for most users and worth the price difference if you don’t want to try third party software.
- Spyde5Elite if you need to tweak multiple connected displays or work with video content that requires calibration to an industry standard, like NTSC.
Calibrating your display with Spyder5
Datacolor Spyder5 works on both Mac (OS X versions 10.7 to 10.10) and Windows (versions 7, 8 and 8.1) computers. All platforms, including Linux, can use the free dispcalGUI and Argyll CMS open source package.
Software download and installation is straightforward, but requires internet access to generate an activation code that ties the hardware serial number to the software. I assume this was done to limit piracy, since the only difference between the cheapest Express and the Elite versions is software.
Spyder5 software is wizard based and intelligent enough to show and hide functions depending on your display type. For example, if you choose laptop on the setup screen, it knows that you only have brightness controls available. Some display types are also detected automatically. For example, current Apple laptops all have white LED backlight, so the software doesn’t need any user input to select it. In the same sense, using room light adjustment is recommended only for desktop users.
All screens have an interactive help sidebar. Take your time to read it. It’s good information even for savvy color management users. There’s also a “click to learn more” link with more in-depth explanations about each step.
Ambient light sensor and why I don’t use it
The ambient light sensor helps users to evaluate their room light conditions and recommend the most adequate display profiling settings for those conditions. This is done only once, during calibration, and doesn’t alter the resulting profile in any way afterward. If the sensor is connected to the computer, the Spyder Utility program can alert you about any ambient light changes so you can generate an updated profile.
In my first run, the software suggested using a 180 cd/m2 brightness based on my room light conditions, which is a very high brightness value for color critical work. This was at evening and with a 120w incandescent fixture on the ceiling. The software warns about making sure no intense light is falling directly on the sensor, but I feel that it is too sensitive. With the sensor placed on my desk, light coming from the display could also affect readings. After pointing it away from the display and the main room light, I got a more reasonable suggestion of 120 cd/m2. White point varied between 5800K and 6500K, depending on the placement of the sensor. 5800K is too warm for most uses, and I’d recommend that you ignore the suggested settings and maybe ambient light measurement altogether.
The first step in the calibration wizard is to make sure your screen is ready to be calibrated. It serves as a reminder to warm the screen for 30 minutes, reset the display controls and check your lighting conditions.
After selecting your computer type — laptop or desktop — and display type, you’re prompted to choose your calibration settings. This is where all three versions drift apart. Express only allows you to choose 2.2 gamma and 6500K as the calibration target. Pro users have more choices and can measure the room lighting conditions. Elite users have almost endless calibration options, including some broadcast video specific targets, like NTSC and Rec BT.709.
If your display has adjustable controls for each RGB color channel, turn on the option “Show RGB Sliders option in identify controls screen” in the Advanced Settings section. If not, look for preset color temperatures and set it to the closest value to your target white point, which is usually 6500K.
Next step is the calibration process itself. The sensor should rest flat on the display surface with no gaps. It helps to tilt the screen back so it doesn’t slide. It’s a good idea to dim the room lights a bit or close any nearby blinds just to make sure no intense stray light interferes with the sensor. Spyder5 software will flash a series of color patches on the screen that will be measured by the colorimeter. Full calibration takes about 5’30 minutes, while subsequent recalibrations are faster, at approximately 2’50.
After saving the profile, you can preview the results by examining a standard test image and switching between calibrated and uncalibrated views. Pro and Elite versions allow you to choose any image for this comparison, while the Express version is limited to the target provided with the software.
Pro and Elite users are presented with a final screen to compare the generated profile gamut to known RGB working spaces, like sRGB and AdobeRGB. This is useful to troubleshoot profiles and make sure that the measured gamut volume is within the expectations for that display type. Those two versions also offer a Display Analysis module that can assess gamut, tone response, brightness, contrast and white point.
Spyder5 worked flawlessly with any quality IPS display that I tested.
All ICC profiles showed little discernible banding on color gradients, with good shadow detail and tonal consistency on the darker values. This is a testament to the updated sensor quality, since measuring darker colors is more difficult, sometimes leading to color casts in the near black grays.
My 30ʺ Dell 3008WFP display is a worst case scenario for gray scale banding. It’s an older wide gamut display with 8bit internal LUTs that are not directly accessible to the calibration device, unlike high end LCDs and the newer Dell UltraSharp models. In practice this means that all the calibration work must be done by the ICC profile itself, manipulating the output of the graphics card before it is received by the display.
On that LCD, Spyder5 profiles are almost indistinguishable from those generated using the i1Display Pro, producing slightly harsher transitions in the blues and reds and a tiny bit more banding. The profile generated using the Spyder5 colorimeter has a little bigger gamut volume in all areas, except for cyan. In practice those are subtle differences, and I had to look hard to find them. Both devices produce excellent profiles that are equal in real world.
A more puzzling difference is that Spyder5 consistently measured white point about 200K lower than i1. This means that displays calibrated with it look slightly cooler than with X-Rite’s package. It’s hard to judge who’s right without access to a reference grade spectrophotometer. This difference seems to be consistent from unit to unit and not caused by any malfunction. It’s also software independent, since testing with third party software led to the same results. 200K is a small difference and didn’t affect print matching or overall accuracy.
Calibrating low quality TN displays was a different experience altogether.
This technology is usually found on cheap LCDs and laptops, including Apple’s MacBook Air and non-retina MacBook Pro models. TN displays are often 6-bit internally, which leads to severe color banding, especially when calibrated to anything outside their native white point. To make matters worse, TN displays are usually very blue, meaning they have a hardware white point over 7500K. The final nail in the coffin is that they shift colors depending on the viewing angle.
Calibrating a TN display certainly helps, but doesn’t solve any of the problems that are inherent with this technology. The further away the display is from your target white point and color response, the harder the colorimeter and software should work to produce a useable profile.
The calibration curves necessary to correct a typical TN display were much steeper than on a quality IPS screen. The effect is the same as when you push your curves too much in Photoshop, especially in 8 bit color mode: posterization and harsh transitions.
In practice I’ve had some really wacky and all over the place results calibrating low quality displays. Visible banding and color casts were a common occurrence, especially when I chose a 6500K target white point. When calibrating a 13ʺ MacBook Air, the resulting profile had a really small gamut volume, which led to blocked and over saturated colors.
Aiming for a native white point by selecting “Unchanged” on the calibration settings (Pro and Elite versions only) produced more useable profiles, even though they looked too blue compared to prints or high quality IPS screens. The downside is that those profiles also had severe color casts on the grayscale. All in all, I still preferred profiles with a 6500K white balance target.
Those problems are not specific to Spyder5, but related to the limitations of TN screens. i1Display Pro exhibited the same bad results, but seems to be more forgiving of lower quality panels and produced better profiles overall. On the other hand, i1Display Pro profiles had a strong red cast, while Spyder5 profiles had a more neutral rendition with a slightly green cast, which gave me a better match to my reference display. Alternative software, like dispcalGUI + Argyll CMS, ColorEyes Pro or basICColor made no difference. I tried them all. TN displays are doomed. :)
Keep in mind that I’m being really picky about it. The results are certainly better than the uncalibrated displays, but far from good enough for any critical work, in my opinion. And this holds true for Spyder5 or any other calibration device.
In order to isolate sensor and software performance, I repeated the same tests using the excellent dispcalGUI + Argyll CMS package. This is an open source display calibration software that often gives better results than the original programs. I highly recommend anyone to try it.
On high quality displays, Spyder5 profiles using dispcalGUI + Argyll CMS are almost indistinguishable from those generated using the i1Display Pro with the same software package. The difference is really small. Grayscales show a little more banding on the darker tones and that’s all.
Lower quality TN displays were still plagued by the same problems I saw when profiling with the original software packages. On the other hand, dispcalGUI + Argyll CMS can give experienced users more control of the results, if you know what to look for.
Keep in mind that this isn’t a perfect apples to apples comparison, because Argyll CMS is open source and not necessarily optimized for any of the devices, as their original software is. Measurement times lag badly behind i1Display Pro, for example, which is the opposite of what happens using the original software.
These results reinforce my conclusion that the new Datacolor sensor is a very high quality device that can compete with any other consumer grade solution in the market. I have no reservations recommending it, especially if you pair the lower cost Express package with dispcalGUI and Argyll CMS.
Spyder5 Elite vs. i1Display Pro
Datacolor software feels less buggy than X-Rite’s i1Profiler. The interface may be older and look dated, but I appreciate the fact that it doesn’t install any daemon other than Spyder Utility, which can be easily disabled. i1Display Pro installs a daemon that is constantly running to detect if the sensor is connected, even when the software or tray utility are not in use.
The latest i1Profiler version 1.6.1 has lots of bugs. For example, some of the ICC v4 profiles I generated looked muddy and with blocked shadows, while the same settings in earlier software versions worked just fine. In Firefox, using an ICC v4 display profile from i1Profiler results in corrupted colors for images tagged with ICC v4 profiles, while Spyder5 v4 profiles worked just fine. The i1Profiler Tray app doesn’t run at all on my system running OS X Yosemite 10.10.3.
Hardware wise, both colorimeters offer comparable performance, with an edge for i1Display Pro, especially on lower quality displays. Room lighting measurement is better with the i1, exhibiting less interference from strong light sources and sensor positioning than Spyder5.
Measurement times with original software are comparable, but i1 is faster in dispcalGUI + Argyll CMS, which can be an advantage for some users, at least until Argyll CMS is optimized for the new Datacolor sensor.
i1Display Pro (EODIS3) currently retails less than the comparable Spyder5 Elite, making it a better choice overall, in my opinion.
Spyder5Pro vs. Colormunki Display
Colormunki Display (CMUNDIS) has a great sensor that offers the same calibration quality as i1Display Pro, but with longer measurements times and a simplified software design. Upgrades between both versions are not possible, but Colormunki is still compatible with Argyll CMS and works great with it. Datacolor used to offer software upgrades in the past, but this option is not currently available for Spyder5, and we don’t know if they have any plans for it in the future.
Software interface is beginner friendly, more akin to Spyder5Express, but it has an advanced mode with more choices for gamma and white point, matching Spyder5Pro in that respect. It lacks the more advanced display analysis features present in Spyder5Pro and the gamut comparison tool, which is really useful to catch any abnormalities with the generated profile.
Raw sensor performance is on par with i1Display Pro, meaning that it is a little better than Spyder5. Both devices are so close that there’s not much to be said about it.
Sticking to the original software, Spyder5 Pro is a better deal than Colormunki Display. It offers comparable profile quality, but better and less buggy software. If you want to work with Argyll CMS, though, Colormunki Display is the best choice. It offers faster measurement times and slightly better profile quality.
Since sensor performance is comparable, it all boils down to the software features offered at each price point.
Spyder5Express is hard to beat at its price. No colorimeter from any other manufacturer, including the X-Rite Colormunki Smile (seems to be a reworked Eye One Display 2 — old technology), can compete with it on pure sensor performance. Using it with the free dispcalGUI + Argyll CMS package removes all software limitations and provides a quality calibration solution from beginner to advanced users.
At an intermediate price point, Spyder5Pro competes favorably with the X-Rite Colormunki Display, beating it on features, calibration speed and software quality, but losing a little bit on raw sensor performance. The Colormunki Display is less expensive and works better with Argyll CMS, though, making it a better choice in my opinion.
The top of the line Spyder5Elite version is a tough sell. X-Rite’s i1Display Pro offers comparable software features and slightly better sensor performance, lacking only the multiple monitor tweaking features available in Spyder5. It also works faster with Argyll CMS and is compatible with most of the high end displays that offer direct hardware calibration, in case you upgrade in the future. That would be my recommendation for high end users.
Users of high end displays that support hardware calibration — Eizo ColorEdge, NEC SpectraView II and Dell PremierColor, for example — should stick with the calibration devices that are directly supported by the manufacturer proprietary calibration software. Those closed solutions yield much better results by directly accessing and adjusting the hardware response, instead of only profiling it and adjusting the color response on the graphics card level. For most manufacturers, this means buying an i1Display Pro and their software package.
PS: What's with all those cryptic product names? Is it Spyder5 or Spyder 5, with a space? Why not Spyder 5 Elite, Pro and Express, instead of everything in all caps and without any spaces? X-Rite is also guilty: i1Profiler instead of i1 Profiler, i1Display Pro instead of the more natural i1 Display Pro. How can color management be simpler if the product names themselves sound complicated?
Do you have any questions? Feel free to write me in the comments and I’ll be happy to help.