According to some new (and rather rare) benchmarks, the iPhone 5 is twice as responsive to touchscreen presses when compared to the Samsung Galaxy S4, and even faster than other Android devices. Where an iPhone 5 can respond to your touches in just 55 milliseconds, the Galaxy S4 takes at least 114 milliseconds, with the HTC One and Moto X clocking in at just over 120ms. Even the three-year-old iPhone 4 is more responsive than the Android devices at 85ms. This is one of the reasons why typing on an iPhone or using an iOS app feels much more responsive than Android.
The testing, carried out by Agawi, uses their in-house Touchmarks benchmarking software and a custom-made Touchscope hardware probe. Basically, Agawi ran a simple app on all of the phones that turned the screen white as soon as it registered a touchscreen press.
The presses were initiated by the Touchscope hardware (which looks like two metal probes controlled by an Arduino), and the response time was measured by a high-speed (240 fps) camera. Then, by watching the high-speed footage, the folks at Agawi could work out the responsiveness of each device.
While Touchmarks ostensibly tests touchscreen responsiveness, this is more a measure of the overall responsiveness of the touchscreen, the touchscreen controller, the SoC, the operating system, and the app runtime.
At a hardware level, there is probably very little difference between the touchscreen implementations on the iPhone, Galaxy S4, and other Android devices. It is possible that Apple’s hardware is better calibrated, but we suspect this isn’t the case; Samsung easily has the expertise and resources to ensure that its hardware implementation is up to snuff.
As we move towards the software stack, though, the implementation could be very different. Where the iOS keyboard and apps run much closer to the bare metal using Objective-C, Android apps run in the Dalvik (Java) virtual machine. Running Java bytecode isn’t that much slower than native code, but the difference is noticeable. This is probably what causes the huge disparity between the iPhone 5 and Galaxy S4.
While the iPhone 5S hasn’t been tested yet, we’d expect it to be just as responsive, or possibly even more so, thanks to its new A7 SoC. On the Android side of things, responsiveness is unlikely to improve.
It’s worth noting, however, that Agawi tested the responsiveness of their own app, which was executed under Dalvik; if they produced an app using native code, or tested some of Google’s own built-in apps, the difference in responsiveness would likely be less exaggerated.
For the most part, though, iOS devices will probably always be more responsive than Android. Because Apple controls the whole software/hardware stack, it can optimise every piece of the puzzle to work perfectly with each other. With Android, just as with Windows, the code must be more generic and abstracted to cater for a wide range of component choices and form factors.