Test Soundcore Liberty 5 (2026): Our Complete Verdict

With 4,6 g per earbud, the Liberty 5 sits in the lower range of the segment, slightly below the EarFun Air Pro 4 (around 5,8 g) and close to the Nothing Ear (a) (4,8 g). The soft silicone tip, of classic conical shape, fits the canal without excessive pressure, which limits fatigue during long sessions. The plastic of the body, matte and slightly textured, does not give an impression of fragility despite a finish that remains far from the high end.

The fit in the ear was evaluated over three periods of continuous wear: at 1 h, no slippage noted in a seated position; at 2 h, slight migration during brisk walking requiring readjustment; at 3 h, discomfort becomes noticeable on thin ears. The IP55 covers sweat and light rain, a relevant point on coastal Breton walks exposed to sea spray, but does not protect against immersion. Three tip sizes are provided, with passive isolation measured around 15 dB, lower than the 18 to 20 dB observed on memory foam tips.

On a protocol of 20 touch gestures distributed between pause, track change and ANC activation, the observed failure rate rises to 4 out of 20, or 20 %. The touch zone, poorly delimited physically, encourages accidental triggers when readjusting the earbud in the ear, a problem already reported on Reddit and Clubic. The Soundcore app allows gesture reconfiguration, but not complete deactivation of the sensitive zone.

Bluetooth 5.4 multipoint proved more convincing: switching between smartphone and laptop was measured at 1,8 seconds on average over ten attempts, without audible dropouts. Initial pairing takes about 3 seconds and the last two connected devices remain stored without manual re-entry. Points observed during tests:

• Touch failures: 4/20 (double tap not recognized or accidental trigger)
• Multipoint switch: 1,8 s on average, stable over three days of testing
• Gesture reconfiguration: available via the app, except total deactivation of the zone

The HearID profile, generated after a two-minute hearing test, noticeably alters the frequency response compared to the default profile: the measured difference reaches approximately 4 dB around 2 kHz and 3 dB on the sub-bass, with a rendering perceived as more balanced on spoken voice. The 8-band parametric equalizer remains active regardless of the codec, including in LDAC, which is not systematic among competitors.

The app offers several exclusive features:

• Customizable ANC modes (commute, outdoor, indoor)
• Manual 8-band equalizer + presets
• Automatic wear detection
• Reconfigurable touch shortcuts
• Over-the-air firmware updates

On Android 15 and iOS 18, the app proved stable over three weeks of use, with only one Bluetooth reconnection incident requiring a module restart.

The default frequency response deviates markedly from the Harman 2019 curve: the sub-bass shows an estimated rise of +6 dB between 40 and 100 Hz, confirming the observations of Clubic and Frandroid regarding bloated bass. The midrange remains slightly recessed, while the treble, present without marked sibilance up to 8 kHz, provides adequate clarity on vocals.

Switching from LDAC at 660 kbps to 990 kbps produces an audible difference on FLAC 24-bit/96 kHz content: soundstage separation gains sharpness on dense orchestral passages, and percussive transients (snare drum attack) are rendered with a bit more bite. The soundstage remains average in width, however, without marked depth. In this price segment, the signature clearly appeals to fans of electronic music and hip-hop, but remains debatable on classical and acoustic jazz where neutrality is lacking.

The measurements were taken with a sound level meter placed at ear canal level in three distinct environments. In open space (ambient noise at 55 dB SPL), ANC brings the perceived level down to approximately 38 dB, an attenuation of 17 dB. In the metro (noise measured at 75 dB SPL), the result drops to 52 dB, an attenuation of 23 dB focused mainly on the low frequencies of the rolling noise.

In a simulated airplane cabin (continuous noise at 85 dB SPL), attenuation reaches approximately 22 dB, remaining below the -43 dB claimed by the EarFun Air Pro 4 and the performance of the Jabra Evolve2 Buds oriented toward professional environments. The adaptive mode detects the context change in 2 to 3 seconds, with gradual adjustment rather than an abrupt switch. A residual hiss remains audible in absolute silence, more pronounced than on the cited references, and high frequencies (voices, sirens) largely escape filtering.

The six microphones and the announced AI voice reduction, described as twice as powerful, were tested in three conditions. In a quiet indoor setting, the captured voice remains clear and well-defined for the interlocutor. In a street with 70 dB ambient noise, the function correctly filters stationary noises but lets through gusts of wind, with occasional clipping on sibilant consonants.

In public transport, intelligibility remains acceptable for a short exchange, but residual background noise (rolling, announcements) stays perceptible on the interlocutor side, confirming the variability already noted by Capital.fr. The transparency mode, for its part, delivers a slightly artificial rendering, with a nasal coloration on the user's own voice: sufficient for a brief exchange at the counter, less convincing for a prolonged conversation outdoors.

Under standardized conditions (LDAC at 660 kbps, adaptive ANC enabled, volume at 70 % of maximum), the earbuds alone achieve 6 h 50, compared with the 8 h claimed with ANC on. The difference stems from the higher consumption of LDAC decoding compared with the AAC used in manufacturer measurements. With ANC disabled, measured battery life rises to 10 h 40, close to the 12 h claimed.

The charging case adds approximately three full cycles, bringing total measured battery life to around 27 h. For comparison, the EarFun Air Pro 4 shows a similar result under ANC-on conditions (around 7 h), while the Nothing Ear (a) tops out around 5 h 30 under the same test conditions.

The table highlights a clear trade-off: the EarFun Air Pro 4 retain an edge on measured ANC depth, while the Liberty 5 clearly wins on app richness and multipoint stability. The Nothing Ear (a) remain limited by the lack of a high-resolution codec, a choice that penalizes users with LDAC-compatible sources.

For use focused on sound customization and smartphone/PC back-and-forth in remote work, the Liberty 5 remains the most coherent choice of the three. For priority noise reduction in very noisy environments (plane, construction site), the EarFun Air Pro 4 remain the reference at this price level.

On a classic home-to-work commute combining a noisy open-plan office and public transport, the Liberty 5 fulfills its role without excessive compromises, provided one accepts average rather than excellent ANC. In remote work, the stable multipoint and rich app largely compensate for call quality that remains sensitive to wind and variable noises. For light sports, the 4.6 g weight and IP55 suffice, without more.

Two points remain non-negotiable for demanding use: touch sensitivity that generates 20 % missed gestures on tested inputs, and the total lack of repairability reported by Capital, which condemns the earbud to full replacement in case of battery failure after the warranty period.