To say that the ROCCAT Burst Pro Air was a disappointment would be an understatement. At 81 g, its weight did not meet market expectations, performance was subpar, and not even a wireless extender found its way into the box. Two years later, ROCCAT as a brand is history, and all former and future mice series will be continued under the Turtle Beach brand. And while the Burst II Air still has its issues, it fares much better than its predecessor.

The most immediately noticeable improvement is the weight. Even though the shape is virtually unchanged, and the only form of "holes" in the shell are two cutouts at the bottom, the Burst II Air weighs a whopping 35 g less at just 46 g. While the halved battery capacity certainly is a contributing factor, and the loss of RGB a major one, a lot of the weight reduction comes from the more efficient mechanical design, and the fact that the Burst II Air doesn't just crumble to dust when applying any kind of pressure is a testament to the engineering prowess involved. That said, some compromises necessarily had to be made, one of which being side wall thickness. As a result, the left side of my sample in particular flexes inward quite easily, which has the side buttons essentially contract. During actual usage, I didn't notice any of this, and there is no creaking, either, but for those looking for something properly "solid," the Burst II Air does not fully qualify. In addition, when shaking, a minor rattle emerges both from the scroll wheel and main buttons, which again only becomes noticeable when actively looking for it, but deserves mention nonetheless.

Though the main buttons are still marketed under the moniker of "Titan optical switch," these are no longer the same TTC optical switches used on previous Burst series releases. Instead, and somewhat ironically, ROCCAT-branded optical switches from Kailh are used, and these deliver a far firmer and snappier button response than the previous TTC switches. In addition, both pre and post-travel are low, and lateral button movement is kept to a minimum, too. What greatly diminishes how these buttons feel, however, is how when being tapped, they emit a hollow-sounding "thunk," which can get irritating when using the Burst II Air for office work in particular. To a lesser degree, the same is true for the side buttons, for which low-profile switches from CF are used. Whereas the forward button has rather low pre and post-travel, pre-travel in particular is very high on the back button, and actuation accordingly mushy. Despite this, the actuation point is even across the entirety of both buttons. For the scroll wheel, an encoder from ALPS has been opted for, which delivers very nicely defined steps allowing for controlled scrolling at manageable noise levels. The feet are made of pure PTFE with a gray backing, hence the seemingly non-white color, and glide well. A set of replacement feet is included in the box, along with a set of grip tape, both of which are welcome additions.

Turtle Beach cites 40 hours of battery life using 2.4 GHz wireless at 1000 Hz. The battery life indicator within the software is not particularly consistent or reliable, but accurate enough to at least allow me to say that 40 hours indeed is the most one can expect, and it might land closer to 35 hours. This is somewhat irritating, as the Burst II Air has the same MCU and operates in the same sensor run mode as VAXEE's wireless mice, yet only musters 40 hours at a battery capacity of 250 mAh, whereas VAXEE does 100 hours at a capacity of 400 mAh. While the optical switches of the Burst II Air draw significantly more power, this doesn't fully account for the difference in battery life nonetheless. The charging cable at least is flexible enough to allow playing while charging without feeling hindered, though charging speed is rather crawl-like at a little over 0.1 A. Some solace is provided by the fact that Bluetooth is present, allowing one to save battery life during office work at least.

Performance is solid on the Burst II Air. CPI deviation is very low, general tracking fine regardless of whether MotionSync is enabled or not, and motion delay consistently low, as smoothing is not present throughout the entire CPI range. Polling is generally stable, though at a fixed interval of at least 2 ms, there is a single off-period poll, which points towards a desynchronization drift, where the mouse and receiver operate at slightly different frequencies. In practice, this has little bearing, and similar behavior can be observed on Logitech mice, but deserves mention nevertheless. A more notable consequence is that motion delay is also increased by an amount equivalent to a single interval, which has the numbers be worse than they otherwise would be. Without MotionSync, a motion delay differential relative to the Logitech G403 (control subject) of 2 ms can be measured, which increases to 3 ms with MotionSync enabled. These numbers aren't bad, but are certainly not particularly good, either. For the record, there is a possibility that the results are negatively affected by the aggressive power saving, as the sensor enters a rest mode after just one second of idle time. Turtle Beach is currently working on a firmware update extending this to five seconds. Click latency, on the other hand, fares better: in wireless operation, 2.0 ms are averaged, which too isn't groundbreaking, but close enough to the competition still. For the record, the advertised NVIDIA Reflex support is essentially pointless, as it only works in wired mode, and the readings are off by such a degree that the function may be regarded as broken.

With a UI that looked like it came straight from 2005, countless bugs that've been sticking around probably since that time as well, and excruciatingly slow firmware update times, ROCCAT Swarm being canned had been overdue for a long time. While Swarm II still retains some elements from its predecessor, such as the way that settings are arranged in tiles, and how updates are distributed, the overall experience is much improved. Setting changes are applied quickly and without having to hit "apply," both software and firmware updates are processed much faster, and several broken or otherwise useless legacy settings have been retired. In addition, I haven't been able to find any bugs, and resource usage is under control at around 100 MB, without any of the CPU time spikes that were plaguing Swarm. Most importantly, when exiting Swarm II, no processes keep running. In short, Swarm II successfully addresses most of if not all the issues I had with Swarm.

Overall, the Burst II Air sits in that peculiar spot where one can make an equally good case for or against it. Its weight sure is impressive, but the compromises in regard to build quality following from it may not suit everyone's tastes, and while performance is solid enough, it fails to compete with many recent offerings, which are often more affordably priced, too. Battery life isn't exactly impressive, either, and the buttons leave room for improvement. Still, performance is good enough to not really matter, the software is among the better ones out there, and most importantly, those who've always wanted a wireless Burst simply don't have much of an alternative. As such, the Burst II Air clutches our Recommended award.