Do Your Eyes Agree With Our Benchmarks?
A couple of weeks ago, Tom's Hardware reader reprotected made a comment in Apple iPhone 5 Review: CPU, GPU, Battery, Wi-Fi, And Display Testing asking for more analysis of display quality not necessarily tied to the benchmark results, and we thought that was a good idea. Of course, we didn't want to give up those numbers altogether for subjective comparisons, so we're doing both.
There's a certain truth to the idea that numbers can only tell you so much. We've toyed with the idea of introducing color accuracy tests using our spectrophotometer, but it's largely non-intuitive, since the result is expressed in a distance metric for color space called delta E. Rather than opening that can of worms, we decided to go into a pitch-black room to capture a picture that reflects the performance of the LCD on its own. There are a few caveats to this, though.
A camera is able to capture the difference between two monitors. If you were to rank color gamut performance, it’d go camera, wide gamut monitor, high-quality printer, and then mainstream monitor and printer. So, depending on your output device, there are no guarantees that you'll see the difference in what we're about to present. So long as you're using a decent screen, though, your monitor should render a similar image. Further, these pictures haven't been optimized in any way. We're simply publishing them after converting RAW files to PNG, which means all six comparisons are made under the same conditions.
We want to avoid any preconception, so these pictures are unlabeled. Which tablet looks the best to you? Take a look at the pictures and form your own opinion. Scroll to the bottom of the page for the answer to which tablet is which.
In each set, the Nexus 7 is up top, rendering deeper red, blue, yellow, and magenta. You may disagree, but just look at the shot above; the blue flowers less brilliant in the second set of pictures, from the iPad mini. This supports our analysis on the previous page, particularly the 3D gamut video.