Test Nothing Ear (3) 2026: Mute Zone's Verdict

Nothing has occupied a precise niche since 2021: transparent design as a differentiating argument, paired with technical specifications that exceed the entry-level segment. The Ear (3) follow this logic by targeting the 130-160 euro bracket, where the most difficult choices for the informed buyer are concentrated.

On this segment, three models structure the market in 2026:

• AirPods 4 ANC (179 euros): Correct ANC, locked Apple ecosystem, no LDAC
• Sony WF-C710N (129 euros): Proven Sony ANC, LDAC codec, shorter battery life
• Jabra Elite 8 Active (149 euros): IP68 robustness, six microphones, adaptive ANC

Nothing's promise rests on three pillars: an announced attenuation of 45 dB, LDAC support for high-resolution sources, and Bluetooth 5.4 for multipoint stability. The Mute Zone team verifies each of these pillars with measured data, not impressions.

The transparent case is Nothing's visual signature, and the Ear (3) develop it with a slightly improved finish compared to previous generations. The hinge is firm, the magnetic click is sharp, and the outer surface resists fingerprints better than expected. The dimensions remain compact: the case slips into a jeans pocket without creating awkward bulk.

The 5.2 g per earbud places the Ear (3) among the lightest in the segment. The editorial team wore these earbuds for two consecutive hours during a telework session: no notable fatigue on the pinna, but slight pressure on the left ear canal with the medium tips provided. Switching to the large tips resolved the issue, highlighting the importance of testing the three available sizes (S, M, L) before concluding a comfort defect.

On stability during movement, the Ear (3) perform adequately in brisk walking, but show their limits during running: without a stabilizing fin, they require readjustment every 15 to 20 minutes approximately. The IP54 was tested under light rain during a coastal route in Vannes: no infiltration, no audio cut. The certification covers sweat during moderate training, not immersion.

The default signature of the Ear (3) features a bass boost centered around 80-100 Hz, with moderate spillover into the lower mids (200-400 Hz) that slightly veils the definition of acoustic instruments. The mids are generally present but show a slight dip around 1.5-2 kHz, which pushes vocals back in the mix. The highs extend properly up to 12 kHz without aggressive sibilance, with a slight attenuation beyond.

Compared to the Harman 2019 in-ear curve, the Ear (3) present a bass excess of about 4 to 5 dB and a lack of presence in the upper mids. This is not a neutral signature: it is a signature that flatters electronic and hip-hop genres, but lacks transparency on acoustic jazz or a cappella vocals.

The equalizer in the Nothing X app offers five presets and a three-band parametric equalizer. By reducing the bass by 3 dB and boosting the 2-4 kHz range by 2 dB, a much more balanced signature is achieved, close to neutrality. The gain is real and measurable, not cosmetic.

On the difference between codecs, the editorial team compared SBC, AAC and LDAC on the same high-resolution source (96 kHz / 24 bits, Tidal Master). In SBC, the compression is audible on fast transients and cymbals lose definition. In AAC, the rendering is clean on an iPhone 15, without notable artifacts. In LDAC at 990 kbps, the gain is perceptible on well-produced recordings: the harmonics of acoustic strings are more resolved, and micro-dynamics gain in precision.

This LDAC gain remains conditional, however: it requires a compatible Android source, a high-resolution file, and a quiet environment to be perceived. On compressed electronic music in standard streaming, the difference between AAC and LDAC is marginal. The marketing argument is real, but its daily benefit depends entirely on the usage profile.

The figure of 45 dB is a peak value announced by Nothing, probably measured in the 100-200 Hz range where the hybrid ANC is most effective. The editorial team evaluated the effective attenuation by frequency band in three environments: TGV Paris-Rennes carriage, 30-person open space, and commercial street in Vannes.

Results by band:

• Lows (< 200 Hz): effective attenuation of 35 to 40 dB, very competitive. The TGV rumble almost completely disappears.
• Mids (200 Hz to 2 kHz): attenuation of 15 to 22 dB. Voices in the open space are reduced but remain perceptible at high ambient volume.
• Highs (> 2 kHz): attenuation of only 5 to 10 dB. Station announcements or the high frequencies of a horn pass without difficulty.

Compared to the Sony WF-1000XM5 (segment reference), the Ear (3) are competitive on the lows but show a gap of 6 to 8 dB on the mids. Compared to the AirPods 4 ANC, the Ear (3) are on par on the lows and take the advantage on the lower mids.

The transparency mode of the Ear (3) reproduces the external ambiance with correct naturalness: nearby human voices are rendered without the metallic distortion found on some competitors. The perceived latency between real sound and reproduced sound is under 20 ms, which avoids the annoying lip-sync effect during face-to-face conversation.

On call quality, the editorial team conducted Google Meet and Teams calls from a noisy open space (background measured at 68 dB SPL) and from the street. Interlocutors reported intelligible voice in both cases, with effective background noise reduction on low frequencies. However, in strong wind (coastal conditions, gusts at 40 km/h), microphone capture degrades noticeably: gusts create audible artifacts that harm intelligibility.

The multipoint on two simultaneous devices (MacBook Pro and iPhone 15) proved stable throughout the test. The switch delay during an incoming call on the second device is 2 to 3 seconds, without disconnection. A brief audio cut of about 0.5 second systematically accompanies the switch, which is within segment norms but worth noting.

The Mute Zone protocol requires a calibrated volume of 75 dB SPL, ANC enabled, LDAC codec, ambient temperature of 20 degrees. Under these conditions, we measured 5 h 05 min of autonomy per earbud, which is 25 minutes below the manufacturer's announcement (5.5 h). The difference is partly explained by the simultaneous use of LDAC, which is more energy-intensive than SBC or AAC.

By disabling ANC and switching to AAC, the measured autonomy reaches 9 h 40 min, close to the announced 10 hours. Fast charging is functional: 10 minutes in the case provide approximately 1 h 20 min of listening with ANC enabled. The case offers 2.5 additional full cycles, for a total autonomy of approximately 17 h 45 min ANC on in real use.

For a Paris-New York flight (approximately 8 h 30 min), the Ear (3) do not cover the journey on a single charge with ANC enabled. A pass in the case midway through the flight is necessary. This is not a deal-breaker, but the Sony WF-1000XM5 (autonomy measured around 7 h ANC on) remains more comfortable for this type of use.

The Nothing X application (available on iOS and Android) offers a solid set of features for the segment:

• Parametric equalizer with three bands and five presets (Balanced, More Bass, More Treble, Voice, Custom)
• Customization of touch gestures (press and hold, double tap, triple tap)
• Adjustment of transparency level (three levels)
• In-ear detection with automatic pause
• OTA firmware updates

Bluetooth 5.4 stability is good under normal conditions. We observed no dropouts in multipoint during an eight-hour session. Effective range is approximately 12 meters indoors without obstacles, which is standard. Beyond 15 meters with a wall, micro-dropouts appear.

The measured latency in standard mode is around 180 ms, which makes video slightly out of sync. Activating the low-latency mode from the application reduces latency to 65 ms, acceptable for video and casual gaming. The Sony Headphones Connect application remains more complete (10-band equalizer, DSEE Extreme, zone-based ANC control), and Jabra Sound+ offers finer sound customization. Nothing X covers the essentials without reaching that level of depth.

The table reveals a nuanced reality. The Ear (3) dominate on low-frequency attenuation and on measured latency in low-latency mode. The AirPods 4 ANC take the advantage on mids and call quality, but at 179 euros and without LDAC, they address a different profile (exclusive Apple users). The Sony WF-C710N is the most direct competitor: 20 euros cheaper, identical LDAC, slightly superior autonomy, but a richer application and less refined design.

The buyer profile for the Ear (3) becomes clear: someone who values design, wants LDAC without paying the price of the WF-1000XM5, and uses Android. For an Apple user or someone who prioritizes call quality, the AirPods 4 ANC deserve consideration despite their higher price.