Headphones and speakers get easier to pick once you read five spec lines: form factor, Bluetooth codec, noise canceling, impedance with sensitivity, and the IP code. Each line answers one question about your ears, your phone, your amplifier or the weather.
Some headline numbers mislead. Sony rates LDAC at up to 990 kbps, yet its lowest step is 330 kbps, almost the 328 kbps SBC baseline Sony quotes itself. Apple rates AirPods Pro 3 at IP57 and still limits that rating to non-water sports.
Headphone Types and Who Each One Suits
Headphone type decides isolation, comfort and portability before any other spec gets a vote. Over-ear models leave room for large drivers, big batteries and deep cushions, which is the shape of the Sony WH-1000XM6 and the Sennheiser Momentum 4. In-ear models trade that room for a seal inside the ear canal. Pick the shape first and compare specs only inside it.
- Over-ear. Cushions sit around the ear, so the seal depends on pad depth and clamp force rather than on the size of your ear canal.
- On-ear. Pads rest on the outer ear, which keeps the frame small and puts all of the clamping pressure on the ear itself.
- In-ear. Silicone or foam tips seal the canal, so isolation rises and falls with tip size, as with the Bose QuietComfort Ultra Earbuds.
- Open-ear. Hooks or clips hold the driver outside the canal, so traffic stays audible, the trade-off behind open earbuds for running.
Budget changes the shortlist and leaves the physics alone. A pair from the best wireless earbuds under $50 list and a flagship from the AirPods Pro 3 vs Sony WF-1000XM5 comparison both seal the same way. Every wireless pair on this page, cheap or not, still sends music through a Bluetooth codec, and only a codec that both devices support gets used.
Which Bluetooth Codec Your Headphones Actually Use
Bluetooth codecs compared on paper differ mainly by bitrate, but you hear only the best codec that the phone and the headphones share. The Bluetooth SIG names SBC as the codec of Classic Audio, so SBC is what plays when nothing better matches. I checked the official pages below on 5 November 2026.
| Codec | What the official source states | Source |
|---|---|---|
| SBC | Classic Audio codec, used by Sony as its 328 kbps, 44.1 kHz comparison baseline | Bluetooth SIG, Sony |
| LDAC | Up to 990 kbps, with a best effort mode that steps between 330, 660 and 990 kbps | Sony |
| LC3 | LE Audio codec that beat SBC in cited listening tests even at a 50% lower bit rate | Bluetooth SIG |
LDAC loses most of its lead on a weak link. Sony writes that the best effort mode changes the bitrate automatically with connection conditions, so a weak or crowded connection can push it down to 330 kbps. At that step LDAC carries almost exactly the data of Sony's own 328 kbps SBC reference, a detail the "3x the data" claim on the same page skips. The LDAC article covers how to force the 990 kbps step on Android.
LE Audio adds two features that go beyond sound quality. Auracast lets one source stream to an unlimited number of receivers, and LE Audio adds hearing aid support, both per the Bluetooth SIG. Check the phone and the headphones for LE Audio before paying extra for it. Lip-sync lag is a separate problem with its own fixes in Bluetooth audio delay, and it decides whether wireless headphones for gaming make sense at all. Battery figures come next, and the one Apple quotes for AirPods Pro 3 assumes noise canceling is switched on.
What Active Noise Canceling Specs Leave Out
Active noise canceling (ANC) uses microphones to pick up outside sound and plays an inverted copy to cancel it. Steady low hum, like an engine or an air conditioner, cancels best. Hybrid ANC combines microphones on the outside of the earcup or bud with microphones facing the ear.
Battery claims with ANC come with conditions. Apple lists up to 8 hours of listening for AirPods Pro 3 with ANC on, and the footnote says that figure was measured at 50% volume. I compare runtimes only when two brands state the same test volume, because a louder test setting drains the battery faster. The Sony WH-1000XM6 and the Bose QuietComfort Ultra Earbuds pages on this site look at how ANC behaves in daily use.
ANC also has a cost. It needs power, microphones and processing, so a wired pair without it plays for as long as the source does. Wired headphones skip the battery and the codec, and then the amplifier in your phone, laptop or interface becomes the limit.
Headphone Impedance vs Sensitivity
Headphone impedance vs sensitivity only makes sense as a pair. Impedance, in ohms, tells you how much voltage the source has to swing. Sensitivity tells you how loud the headphones play for a given input, and makers quote it either in dB SPL per milliwatt or in dB SPL per volt.
Those two units do not compare directly. Power equals voltage squared divided by impedance, so 1 volt pushes 31.25 mW into a 32-ohm headphone and only 4 mW into a 250-ohm one. Two models with the same dB/mW rating will therefore differ by about 9 dB at the same 1 V output, with the 250-ohm pair quieter. Convert both figures to dB per volt first.
High impedance alone does not mean you need an amp. A sensitive 250-ohm pair can still reach a usable level from a weak source, and 250 ohm headphones walks through that case. If the answer is an amp, an audio interface on the desk may already contain one. The same power arithmetic applies to passive speakers, where the load in ohms and the sensitivity in dB decide how many watts the amplifier must deliver.
Speaker Sensitivity and 4 vs 8 Ohm Loads
Speaker sensitivity, quoted in dB SPL at 1 watt or at 2.83 volts measured at 1 meter, tells you how loud a passive speaker plays from a given amplifier. Every 3 dB of sensitivity equals a doubling of amplifier power. So a 90 dB speaker reaches a given level with a quarter of the power an 84 dB speaker needs.
Watch the unit behind the number. 2.83 V delivers 1 W into 8 ohms, but the same 2.83 V drives 2 W into 4 ohms. A 4-ohm speaker rated at 2.83 V therefore looks 3 dB more sensitive than it would on a 1-watt rating, which is a real difference on a small amplifier.
Load matters as much. A 4-ohm speaker draws twice the current of an 8-ohm speaker at the same voltage, so check the minimum load printed on the amplifier's spec sheet first, and read 4 ohm speakers if the figures sit close. Powered studio monitors contain their own amplifiers, which removes the matching question. Car audio has its own vocabulary, from coaxial speakers to 5 way speakers. Portable Bluetooth speakers add one more spec line that home speakers skip: the IP code for dust and water.
IP Ratings for Earbuds and Speakers
IP ratings for earbuds and speakers follow IEC 60529. The first digit covers dust, the second covers water, and an X means the product was not rated against that one. Each common code below stands for a fixed test in the standard:
- IPX4. Water splashing against the enclosure from any direction causes no harmful effect in the standard spray test.
- IP5X and IP6X. IP5X means dust protected, where some dust may enter but not enough to disturb operation, and IP6X means dust-tight.
- IPX5 and IPX6. The device survives water jets from a 6.3 mm nozzle, or powerful jets from a 12.5 mm nozzle for IPX6.
- IPX7. The device survives 30 minutes with its lowest point 1 m below the water surface.
- IPX8. Continuous immersion beyond IPX7 under conditions the manufacturer must specify, so the depth and time live on the product page.
AirPods Pro 3 carry IP57 on Apple's spec page. Under IEC 60529 that code means dust protected and rated for 30-minute immersion at 1 m. Apple still limits the rating to non-water sports and writes that resistance "might decrease as a result of normal wear." The rating describes a new unit. If a product page shows no IP code at all, treat the headphones as untested against water, the situation behind the Beats Studio 3 question. Water ruins hardware, but volume damages the listener, and hearing risk has published limits.
Safe Listening Levels for Headphones and Speakers
85 dBA averaged over an eight-hour day is the NIOSH recommended exposure limit, and NIOSH halves the allowed exposure time for every 3 dBA increase above that level. That exchange rate gives 4 hours at 88 dBA, 2 hours at 91 dBA and 15 minutes at 100 dBA. NIOSH wrote the limit for workplace noise, yet the arithmetic does not care whether the sound comes from a drill or from headphones and speakers.
Young adults are the group the WHO warns about. The WHO fact sheet dated 3 March 2026 states that over 1 billion young adults are at risk of permanent, avoidable hearing loss due to unsafe listening practices. On a train I would rather pay for working ANC than turn the volume up to beat the engine noise. Hearing damage is the one fault on this page that no setting or repair reverses.
Common Headphone and Speaker Problems Mapped to Fixes
Most audio faults fall into a few repeating patterns, and each one below already has a dedicated InearSpace article with step-by-step checks.
| Symptom | Where to start |
|---|---|
| Sound lags behind video | Check which codec is active, then follow Bluetooth audio delay for your device |
| Music stops on its own | Dust, jack connectors, auto-pause and app bugs are covered in headphones keep pausing my music |
| Volume jumps without input | Dirty jacks, voice assistants and control apps are covered in headphones keep changing volume |
| Speakers play too quietly | Default device, drivers, equalizer and wiring checks are listed in speakers so quiet |
| Speakers have no bass | Connections, settings, placement and enclosure are covered in speakers have no bass |
| Left and right are swapped | Windows 10 balance settings and speaker polarity are covered in switch L and R speakers |
I checked every spec on this page against the Sony, Apple, Bluetooth SIG, NIOSH and WHO pages on 5 November 2026, and the IP digit definitions come from a published summary of IEC 60529 rather than the standard itself. Product pages change after firmware updates and new revisions, so recheck the IP code and codec list on the maker's page before you buy.