DataColor SpyderX Colorimeter
Finally, X-Rite has some competition
For the last several years the X-Rite i1Display Pro has pretty much been the only game in town when it comes to affordable, serious display calibration. Sure, DataColor offered the Spyder range of colorimeters at a substantially lower price, but their performance has not been competitive with the X-Rite alternative. DataColor has now released the next generation of Spyder colorimeters, the SpyderX. X-Rite now has some serious competition. To be clear, the SpyderX performs substantially better than all previous Spyder instruments.
Colorimeter performance is defined by three criteria: color accuracy, low-light sensitivity, and speed. Of these color accuracy is the least important, because color errors can always be corrected by a better instrument. However, speed and low-light sensitivity are baked into the colorimeter's design and nothing can be done about either without reengineering the device.
For color accuracy I simply measured 11-point grayscale and the primary and secondary colors with each meter. Next, I repeated the measurements after correcting a Colorimetry Research CR-100 colorimeter with the Colorimetry Research CR-300 reference spectroradiometer. Finally I compared the x and y values of each set of measurements. This defines the color error.
To evaluate speed I imported a custom list of colors into ChromaPure's Raw data module, which were just the grayscale in 1% increments from 100% down to 5%. Next, I measured using each meter. Finally, I took note of how much time was required to measure the entire list of 95 grayscale values.
To evaluate low-light sensitivity I measured subsequently lower grayscale values starting at 8% with each meter stepping down one 8-bit code value at a time. For this test I used the Accupel 6000 test pattern generator and the 2018 LG OLED. Finally, I noted the lowest point at which the meter would continue to return reasonably consistent luminance measurements. For this test I did not consider color performance.
As you can see, the SpyderX's grayscale accuracy is substantially better than the Spyder5. However, this result is a little misleading. Because the SpyderX benefits from substantially improved low-light sensitivity, the bulk of the Spyder5 color error is in the 5% reading. At 10% and above the Spyder5 is actually slightly more accurate than the SpyderX. With respect to color accuracy, the two meters measured the same. The SpyderX's accuracy is comparable to the i1Display Pro.
The low-light sensitivity of the SpyderX is more than 10 times better than the Spyder5. However, it must be noted that the Spyderx's low-light sensitivity is still considerably inferior to the X-Rite i1Display Pro, which can read as low as about 0.0025 nits in its optimum mode. In most cases this difference won't matter because 0.028 is more than adequate to measure 5% grayscale. The only instance in which this will matter is measuring the black level of a very high contrast flat panel (It won't matter for projectors because you can always measure off the lens and simply place the meter as close to the lens as necessary to get an accurate black level reading.) The SpyderX will measure the black level of any flat panel whose contrast is up to 5,300:1 (assuming a peak white of 150 nits). Contrast above that all the way up to about 50,000:1 the i1Display Pro will be a better choice. Contrast on OLEDs is unmeasurable by any instrument.
The SpyderX is more than twice as fast as the Spyder5 and is comparable to the i1Display Pro. This is no small matter, especially whenever you capture a large sequence of readings.
The SpyderX represents a substantial improvement in the Spyder family of colorimeters and now offers comparable performance to the i1Display Pro in every respect except low-light sensitivity at a non-trivially lower price. It is highly recommended.
I performed some additional tests for color accuracy on different display types and tested repeatability.
As you can see, the average xy error is between 0.003-0.006, and the average dE error is 1.3-2.0. As before, these errors are comparable to what we see with the i1 Display Pro.
One interesting feature of the SpyderX that I discovered involved its optimal operating mode. The instrument includes four modes
I discovered that the most accurate color measurements were obtained by using the Wide LED mode regardless of the display type. For example, I got more accurate readings using this mode with the plasma, despite the fact that the plasma was neither a LED nor wide gamut.
One area of performance that I neglected to measure the first time around was repeatability. This is an important area of performance because without it you cannot trust any individual reading if the readings change substantially each time you measure. This test includes 10 consecutive measurements of 10% grayscale, which is a good stress test for repeatability.
These are very good results. At 20% grayscale and above the repeatability is nearly perfect.
In general, I am impressed with the SpyderX, especially considering its price point.
I have only two reservations. First, its low-light performance, though adequate for virtually all calibration tasks, is still not good enough to measure black level on many modern displays. This result is tempered somewhat by the fact that the black level of the very popular OLED display type is unmeasurable in any case.
Second, the utility of the SpyderX is somewhat limited for those who have projectors because it has no illuminance mode, so all projector readings must be taken off the screen. Sometimes it is useful to take readings directly from the projector lens to get more light into the sensor. The i1 Display Pro includes a diffuser that is put into the light path whenever taking lens readings. The SpyderX does not. Users may be able to use the meter in this way by employing an aftermarket diffuser, but I have not tested this.