VAXEE AX Review

VAXEE AX: A Collaborative Gaming Mouse

Starting from 2019, VAXEE was established by a group of former Zowie staff, specializing in peripherals and an online shop. The AX represents VAXEE's second collaboration with a notable figure in esports. This time, they've teamed up with Zhengwei "alex" Bian, a former CS pro, and his brand, Outset. The Outset branding is prominently displayed on the AX, while the phrase "powered by VAXEE" is subtly placed on the back of the mouse, highlighting the collaborative effort behind the design. For clarity, I will refer to VAXEE only throughout this review.

The AX comes in variants of matte black and glossy white, which you can find exclusively at VAXEE's own shop. The matte coating is now more matte, and the stock mouse feet have a thickness of 0.6 mm. Improvements have been made to the AX, including a smoother scroll wheel, while its shape resembles a differently balanced EC2. VAXEE believes that aiming for the lightest weight isn't always the best for gaming performance, opting instead for a balanced weight of 81 g. In terms of features, the AX prioritizes on-device configurability over RGB lighting or resource-heavy software. Users can adjust settings like CPI, polling rate, lift-off distance, and button response time directly on the mouse. The mouse is built with a PixArt PMW3389 sensor and Huano switches for the main buttons, ensuring solid performance. Additionally, the AX has a soft, braided cable and a 24-step optical scroll wheel, embodying a focus on performance without unnecessary frills. Overall, while the Outset AX shares similarities with the Zygen NP-01, it stands out with its thoughtful design and performance-oriented features.

Overview

Mouse Dimensions

The this sheet created by a respected community member provides a complete list of all the measurements taken. Measurements taken with a ruler are noted to have one decimal, while those measured with a caliper have two. The term "Front height" refers to the space between the base and the main buttons. This table presents the dimensions of various right-handed ergonomic mice, including the VAXEE Outset AX.

Sustainable Packaging

The black and white box contains nothing but the well-cushioned mouse, reflecting VAXEE's dedication to minimizing waste. This minimalist approach isn't just about looks; it shows a deeper commitment to sustainability. On the AX, this philosophy is evident in every detail.

Weight Comparison

The shell rigidity has been enhanced compared to the NP-01, making the AX feel more solid. While it may not be considered ultra-lightweight, the AX is not heavy at all. In fact, it weighs about 10 g less than the Zowie EC2-B. My scale reads approximately 81 g (+/- 1 g), which is a bit higher than what VAXEE states.

Cable Quality

The cable exit is angled upwards slightly, just like on the NP-01. It features a gold-plated USB connector and an attached ferrite bead that helps minimize signal noise. With a length of 2.10 m, this cable is a bit stiffer compared to typical paracord-like cables. In terms of stiffness, I would compare it to the Razer Viper Mini cable. VAXEE emphasizes durability as the reason for opting for this more rigid cable instead of more flexible alternatives. The AX has a paracord-like, braided cable.

Foot Upgrades

Starting with the feet, the stock ones on the AX are 0.6 mm (measured), which is an upgrade from the slightly thinner 0.45 (0.4) mm feet on the NP-01. The size and shape of the AX feet match those of the NP-01 perfectly, ensuring a familiar feel. Glide performance is also perfectly fine. It's great to see VAXEE opting for the thicker feet right from the start. Additionally, small indentations next to the feet make removal a breeze. The feet are made from white-dyed pure PTFE (Teflon).

Grip Adaptation

The design choice of the AX allows for a unique experience, especially for what VAXEE refers to as "tilt grip," where the wrist is rotated to the right instead of resting flat on the pad. This can benefit any grip style, meaning a player with a claw grip might actually favor the AX over the EC2 because of this specific feature. While the AX is tailored for palm grip, offering a fuller hand feel without limiting vertical movement, the EC2 is generally better suited for claw or fingertip grips. The changes made to the AX do not drastically alter the shape but significantly impact how one can hold the mouse. The AX takes the EC2 design and enhances it by being shorter and slightly wider, with the hump still centered but the left side elevated for a bulkier feel than the more elongated EC2. The back edge has been trimmed for more palm space, the main buttons are positioned higher, and the comfort grooves have been removed. For over a decade, the EC2 has been a favorite and often imitated shape, but VAXEE's modifications with the AX aim to meet various user needs.

In my experience, the AX is quite comfortable for claw grip, especially for hands measuring 19 cm in length and 10 cm in width. It’s a medium-sized mouse that features a noticeable hump, making it a good fit for various hand sizes. While fingertip grip can work, I believe it primarily caters to palm and claw grip styles.

Sensor Performance Optimization

The VAXEE Outset AX features four pre-defined CPI steps right out of the box: 400, 800, 1600, and 3200. With a maximum tracking speed of 400 IPS, which translates to 10.16 m/s, the mouse is designed for precision. The PixArt PMW3389 sensor equips it with impressive capabilities, allowing for up to 16,000 CPI according to its specifications.

CPI Consistency

In this test, I'm checking if nominal CPI aligns with actual CPI. Ideally, there should be minimal difference between the two, even though it's impossible to achieve a perfect match every time. Factors that can affect this include the thickness of the mouse feet, the type of mousing surface, the sensor's mounting height not meeting specifications, and firmware. The term "CPI" (counts per inch) refers to how many counts a mouse registers when moved one inch. However, keep in mind that the variance in CPI can change from one unit to another, so results may vary for different users.

The results show a very good outcome, with minor deviations. This is particularly significant as the AX cannot freely adjust CPI steps, making CPI accuracy crucial. I've focused my testing on the four most typical CPI steps: 3200, 1600, 800, and 400.

Motion Delay Assessment

The G403 serves as the control subject for this test, as its 3366 sensor shows no noticeable smoothing throughout the entire CPI range. It gets a slight advantage by being tested first. The focus of this examination is on sensor smoothing, which refers to the averaging of motion data over multiple capture frames to minimize jitter when using higher CPI settings, ultimately leading to increased motion delay. The aim is to achieve minimal smoothing. However, this test won't capture any extra sources of input delay, as "motion delay" includes all types of sensor lag. Since it's impossible to measure motion delay accurately in absolute terms, comparisons with a control subject are essential for assessing performance.

I'm examining two xCounts plots—one at 3200 CPI and another at 1600 CPI—to see if there's any smoothing, which would show up as "kinks." Typically, the plot at 3200 CPI would exhibit these "kinks" due to the 3389 having 32 frames of smoothing at and above 1900 CPI, leading to an additional motion delay of about 4 ms at the slowest speeds. As you can observe, this holds true here, although the kinks are only faintly visible. Additionally, we can note that there is very clean tracking, with no significant SPI timing jitter to point out.

To determine motion delay, I examine xSum plots created at 3200 and 1600 CPI. Notably, at 1600 CPI, there’s no noticeable difference in motion delay. However, at 3200 CPI, a differential in motion delay of about 3.5 ms appears. This aligns with expectations for a 3389 sensor, which features 32 frames of smoothing when operating at or above 1900 CPI. The leftmost line indicates the sensor that experiences less motion delay.

Speed-Related Accuracy Variance Explained

When discussing "acceleration," people usually refer to speed-related accuracy variance (or short SRAV). The key point is that the cursor does not move the same distance when the mouse is moved the same physical distance at varying speeds. This concept can be tested easily by comparing it to a control mouse that is recognized for its low SRAV, specifically the G403. As seen in the plot, the cursor paths show no displacement between the two, confirming that the SRAV is indeed very low. It's important to note that this isn't about the mouse having a fixed amount of positive or negative acceleration.

Speed Control Excellence

At no point does the sensor show any signs of malfunctioning at my maximum speed of 4.5 m/s, which falls within the announced PCS range. Perfect Control Speed (or PCS for short) refers to the highest speed at which you can move the mouse and its sensor without any issues. It's fascinating how this capability works!

Polling Rate Options

Stable and nice are the three polling rates available: 1000 Hz, 500 Hz, and 125 Hz.

Sensor Performance Evaluation

At the highest tested CPI of 3200 on the AX, there is no movement of the sensor lens. Additionally, no jitter was detected at any of the CPI steps evaluated. Finally, as observed, there are no issues related to angle snapping (the unintended straightening of linear movement) or sensor lens rattle. This test effectively highlights these potential issues.

Click Latency Analysis

The measured value for click latency is not the absolute figure. It has been noted that, using the 2 ms button response time setting, the click latency was roughly +1.2 ms when compared to the SteelSeries Ikari, which serves as the baseline at 0 ms. It's important to highlight that the AX's internals could not be examined, leading to the assumption that its click latency is identical to the Logitech G100s. Click latency, which is influenced by debouncing, typically introduces a delay that can also include any processing delays, now referred to as click latency. Since most computer mice utilize mechanical switches for their buttons, debouncing becomes necessary to prevent accidental double clicks. For further insights, comparison data is drawn from this thread and my own experiments, utilizing qsxcv's program.

Build Quality Assessment

Slam clicks can occasionally happen when the mouse is dropped from a significant height and set to a 2 ms button response time. Squeezing the sides activates the side buttons, but it takes a lot of force compared to the NP-01. There are no creaks when applying lateral pressure, and the mouse doesn't rattle at all when shaken. Overall, the build quality is solid.

Button Performance

The buttons on the AX utilize a set of Huano switches (pink plunger, 10 M) that require an actuation force of 60 g, making the stiffness light. Lateral movement between the buttons and the main mouse body is minimal, even though they are visually distinct from the shell. However, the buttons do produce a noticeable sound, which might be off-putting for some users. The post-travel distance is moderate, but pre-travel is low, ensuring a firm and snappy response. Overall, the main buttons on the AX are very good.

The use of Huano switches (3M) enhances the experience. Actuation is simple, thanks to the well-placed buttons that allow for easy thumb rolling. The pressure point remains consistent across the entire button, contributing to a solid feel. With minimal pre and post-travel, button response is both quick and firm. Overall, the side buttons are fantastic.

The click feel of the buttons is perfectly fine. At the bottom of the mouse, you can find several buttons designed for adjusting CPI, polling rate, and debounce delay.

Scroll Wheel Performance

The loudness of the scroll wheel is simply unacceptable. I'll readily concede that the implementation on the AX is probably the best one we've seen, and the tactility is indeed nice. However, the scroll wheel remains as noisy as ever. These days, multiple manufacturers have shown that it’s entirely possible to have distinct steps while keeping noise levels in check. Unfortunately, VAXEE continues to cling to this dreadful scroll wheel. The middle (scroll wheel) click demands an excessive amount of force for activation, turning it into a hassle to use at all. An optical encoder is in play, and another Huano switch is utilized for this function. The scroll wheel is poor.

Coating Durability

All in all, the materials used are good. After cleaning, there are hardly any signs of wear. It’s easy to clean, even though it easily attracts fingerprints, dirt, and similar marks. The rest of the mouse has a smooth matte coating, similar to that of many Zowie mice, while the main buttons are made of slightly rougher plastic.

Screw Challenges in Disassembly

Eventually, I had to give up due to the damage inflicted on the screw head from my numerous attempts. Not a single screwdriver in my collection was effective, particularly on the second tri-angle screw, which refused to budge. The first one, however, was manageable with a Torx T5. Two of these tricky screws are situated beneath the bottom skate, complicating the disassembly process. VAXEE later confirmed that the internal components of the AX are completely the same as those in the NP-01, meaning the insights from that review are still relevant. Much like on the NP-01, disassembly is complicated by the fact that tri-angle screws are used.

The main PCB is quite sturdy, secured with four screws. It houses the MCU, which is a Cypress CY7C64356-48LTXC. For more information, including the datasheet, check out this. Below the buttons, individual button switches are enclosed to enhance button consistency. The scroll wheel features a covered design, with a spring mounted vertically above the wheel click switch, helping to maintain reset accuracy and preventing movement when the mouse is inverted. Three PCBs connect to the main PCB via thin ribbon cables. Typically, these ribbon cables are clipped into their connectors, but in this instance, they seem impossible to unclamp. Overall, the design is impressively intricate.

I can't find any significant flaws in the PCB's overall quality and soldering.

Conclusion

The wrist rotation, known as "tilt grip," plays a crucial role in the design, impacting all grip styles. This design choice caters to players who prefer a more dynamic hand position while gaming. The AX is bulkier due to its higher left side but still allows for unrestricted vertical movement, making it versatile for different playstyles. Palm grippers, in particular, might find the fuller feel comfortable while benefiting from the extra space towards the back. By revisiting the classic Zowie EC2 shape, VAXEE introduces the AX, a compressed version that is both shorter and wider. This evolution reflects the popularity of the EC2 over the last decade, noted not just by its prevalence among professional players but also by the numerous clones that have emerged. The AX offers a fresh take on a beloved design, merging familiarity with innovation.

The appeal of the AX could definitely improve with some adjustments, especially since the wheel click is now even stiffer than on the NP-01, making it tough to use at times. It's puzzling why VAXEE continues to opt for this noisy scroll wheel, given that other brands have shown that a tactile yet quiet scroll wheel is achievable. While the scroll wheel has become a bit smoother, it remains excessively loud. Moving on to the main features, the main and side buttons still deliver a satisfying and firm click, maintaining that pleasant click sensation. Configuration remains user-friendly with no software needed; CPI, polling rate, button response time, and lift-off distance can all be adjusted directly on the mouse. The PixArt PMW3389 sensor continues to perform outstandingly, and the click latency can still be set to impressively low values. Despite the shape being the main change compared to the NP-01, everything else has seen improvements that enhance the overall experience.

The coating has been updated, yet I didn't find any significant grip enhancement compared to the NP-01, which I actually prefer. The cable remains unchanged, still a bit stiffer than typical paracord-like cables, but it's claimed to be more durable. Consequently, the weight is slightly elevated at 81 g, striking a balance that isn't particularly high or low. Concerns about shell rigidity on the NP-01 have been tackled: pressing the side buttons by squeezing the sides is still an option, but it needs a lot of force, making it less of a problem. The AX now features 0.6 mm feet by default, improving glide on less-than-perfect pads and providing nearly flawless CPI deviation, which is a welcome change from the NP-01's excessively thin stock feet that caused poor glide and above-average CPI issues due to a lack of fine-adjustment. Improvements have indeed been made.

The AX might not be my top choice for office tasks, but when it comes to gaming, it really shines. It's impressive how VAXEE is addressing the flaws in their designs with the AX. While some might argue that every new product should offer something groundbreaking, I believe there’s real value in sticking to what works. Innovation isn't everything for gaming mice. Companies like Zowie and VAXEE have faced their fair share of criticism for not pushing boundaries enough. Ultimately, it's about delivering reliable products that perform as expected, rather than constantly reinventing the wheel with each release.