How Active Noise Cancellation (ANC) helps in games

Active noise cancellation in a gaming headset barely works at the frequencies that carry most footstep information. Footsteps become audible earlier after ANC is turned on precisely because of this. A paradox? Let's sort it out.

HATOR Hypergang 3 Wireless, one of the current models with hybrid ANC

The first part of this ANC series criticized the practice of summing active and passive noise cancellation into a single promotional figure like "−52 dB." A band-by-band breakdown removes this problem. It shows the upper limit beyond which the active system stops producing results.

Hybrid ANC in HATOR's current models is uneven across the spectrum. According to the manufacturer, the distribution is as follows:

• 50-300 Hz, 25 to 30 dB of attenuation. PC fan hum, refrigerator compressor, air conditioning, street traffic.

• 500 Hz to 1 kHz, 10 to 15 dB of attenuation. The lower midrange, which includes part of the speech frequency band.

• Above 1.5-2 kHz the active system's contribution drops to near zero. From there, isolation depends on the cup construction and ear cushions.

The roll-off at higher frequencies has a physical cause. The shorter the wavelength, the more precisely the system must match the phase, and the less time it has to compute. At 100 Hz the wavelength is about 3.4 meters, and at 5 kHz about 7 centimeters. A few-millimeter error means nothing in the first case and turns anti-noise into additional noise in the second. Every ANC system, regardless of brand, rolls off at the top. The only question is where that roll-off begins.

What actually makes noise in a gamer's room

The objection "my room is quiet" rests on habituation. A steady background stops attracting attention within minutes, because the auditory system stops flagging an unchanging signal as news. Habituation operates at the level of attention. Masking in the inner ear does not depend on attention at all.

Most of the energy in household noise sits at the low end. Fans, compressors, and street traffic produce a steady hum from 30 to 300 Hz, exactly the band where ANC works best. A mechanical keyboard does not belong on this list. Its clicks are broadband, with a large share of energy above 2 kHz, so only passive isolation handles them.

Some noise sources ramp up in sync with the workload. GPU and CPU fans spin faster exactly when the most is happening on screen. The moment when a player most needs to hear detail coincides with the moment of maximum background noise.

A smartphone sound-level meter app can measure room noise, but the reading will be low. A-weighting, which meters and apps apply by default, subtracts about 19 dB at 100 Hz and more below that. Phone microphones are not calibrated at these frequencies either. The instrument underestimates exactly the part of the spectrum that ANC removes.

Background noise in a room with a PC under minimal load

Upward spread of masking

Masking describes a situation where one sound raises the audibility threshold of another. Low-frequency noise masks higher frequencies far more effectively than high-frequency noise masks lower ones. This is a fundamental and very old finding. Mayer described it qualitatively as early as 1876, Fletcher published the first quantitative results of tonal masking in 1923, and Wegel and Lane expanded the experiments to a wider set of tones in 1924. In psychoacoustics the phenomenon is known as upward spread of masking, where loud low-frequency components interfere with hearing sounds at much higher frequencies.

Room hum in the 50-300 Hz range raises the audibility threshold in the mid and upper-mid bands, where most of the distinguishing detail of footsteps and reloading lives. ANC does not work at those frequencies, but it removes the masker. The threshold drops, and a footstep that previously fell below it becomes audible.

The effect is also nonlinear. For a signal with a lower-frequency masker, each increase in masker level raises the threshold more steeply than the same increase would at the signal's own frequency. Removing 25 dB of hum yields a larger gain in upper-frequency audibility than simple arithmetic suggests.

A headset with good passive isolation but no ANC attenuates the highs and barely touches the lows, because passive isolation below 200 Hz is almost nonexistent. In part one this is visible in the measurements of a HATOR prototype, where the test-noise line and the passive-isolation line converge almost completely at the bottom. The construction removes the frequencies that mask weakly and leaves the ones that mask strongly. The active system removes the lows, the very part that the construction leaves untouched.

Why background noise also degrades directionality

Detecting a footstep is not enough. The player still needs to localize it. The mechanism responsible is described by Lord Rayleigh's duplex theory. Below roughly 1.5 kHz, direction is determined by the interaural time difference of the arriving signal. Above that, direction is determined by the interaural level difference.

ANC's operating band overlaps with the band where the time-based mechanism operates. Low-frequency noise contaminates the same frequencies that the auditory system uses to compute azimuth, so removing hum affects both detectability and directional accuracy. Part one noted that for spatial noise-cancellation systems the region around 800 Hz is the most difficult, because there the wavelength equals twice the distance between the ears. In headphones this problem does not arise, and the same region remains within ANC's working range.

ANC does not conflict with virtual surround sound. Spatial Audio, head tracking, and HRTF implementations process the signal that the headset reproduces, while ANC processes what arrives from outside. Accurate spatial-cue processing yields less benefit, however, when the cues themselves are buried in room hum before they reach the ear.

How much earlier does a footstep become audible

In the real world a point source loses 6 dB for every doubling of distance. If a game engine used the same decay curve, a 12 dB improvement in signal-to-noise ratio would mean roughly four times the detection distance. Competitive shooters rarely model physics that literally. Footstep volume is set by a designer curve, often with a hard cutoff radius. The real gain is therefore smaller than the arithmetic one, and measuring it requires testing in a specific game on a specific map.

Without measurements, three things can be stated. The gain exists, it increases detection distance, and its size depends on how loud the background was before ANC was turned on. In a quiet room at three in the morning the improvement will be minimal. Next to a PC under load, noticeable.

This dependence on background noise explains the spread of opinions on ANC in gaming headsets. A player in a country house with a quiet water-cooled build and a player in a room with a window facing a busy avenue and a blower-style air cooler get fundamentally different results from the same headset. Conflicting reviews in such cases describe different rooms rather than different headsets.

The volume you don't have to crank up

To hear a footstep buried in room noise, a player raises overall volume. Gunshots, explosions, and voice chat rise by the same amount, and the loudness gap between a footstep and a gunshot in a shooter's mix is large.

The WHO and ITU H.870 standard defines a safe weekly dose as the equivalent of 80 dB over 40 hours per week for adults and 75 dB for children. The scale is steeply nonlinear: at 90 dB, safe exposure drops to four hours per week. In its own recommendations, the WHO explicitly names noise-cancelling headphones among the ways to avoid raising volume in noisy environments.

Forty hours a week sounds generous until you count a real schedule. Two hours on weekday evenings and four on each weekend day add up to eighteen hours, and that is before music during commutes, streams in the background, and work calls in the same headset. The dose is cumulative across all sound sources, so gaming shares the budget with music and calls, even before adding every other reason a person encounters loud sound during a week.

The player gets a choice between the same footstep audibility at lower volume and greater audibility at the same volume. Only the first option concerns hearing, though attention usually goes to the second.

The second category of promises around ANC concerns concentration and fatigue. The acoustic side of this claim can be verified. The subjective side cannot. Extracting a weak signal from noise requires more processing than extracting the same signal from silence, and this follows from the same masking effect. How that feels as fatigue during a four-hour session is something each person determines individually. A specific line item about it in a headset's spec sheet is hard to imagine, yet the advantage of ANC is real.

Does the player pay in milliseconds for silence

No. Turning on ANC does not lengthen the path of game audio to the ear, because the headset runs two independent signal chains. Game audio travels from the source through the radio link or cable to the DAC and on to the driver. Anti-noise originates inside the headset and passes through its own chain of microphone, ADC, DSP, DAC, and the same driver. The second chain joins the first at the driver level, running in parallel.

Two numbers from part one describe different quantities. Noise-cancellation processing in the HATOR Phoenix 2 Wireless takes 0.0075 ms, or 7.5 µs. This is the anti-noise loop's cycle time, which must stay negligible, because late anti-noise itself becomes noise. Game-audio latency in wireless mode is under 20 ms and does not depend on ANC.

Perceptible milliseconds appear at a different stage, the choice of radio channel. Bluetooth adds 100 to 300 ms depending on the codec. For competitive gaming, 2.4 GHz or a cable remain the only options. Whether ANC is on or off has no bearing on this choice.

What ANC costs in battery hours and charge cycles

Active noise cancellation draws from the same battery as the radio link and lighting. Part one cited average figures: branded headsets run 20 to 50 hours per charge, and ANC reduces that by roughly a quarter. No-name models with smaller batteries deliver 8-10 hours.

A headset rated at 30 hours with ANC on and used for four-hour daily sessions needs charging about once a week. The same headset worn for an eight-hour workday needs charging every three to four days. The difference in annual charge cycles is roughly two and a half times.

Battery cycle life rarely appears in spec sheets. The industry baseline remains 300 charge-discharge cycles. With daily charging, such a cell reaches end of life in about a year, and replacement is not covered by warranty. HATOR uses custom cells rated for 500 cycles in the Phoenix 2 Wireless and Hyperpunk 3 Wireless. Charged once a week, that outlasts the headset itself, and charged daily extends the useful life to about a year and a half instead of one.

Turning ANC off during a quiet late-night session makes sense for cycle life rather than sound. Playing while plugged in eliminates the battery question entirely, provided the model supports simultaneous charging and operation.

What Bluetooth does when the game runs over 2.4 GHz

The 100-300 ms latency rules Bluetooth out for competitive shooters, but leaves it for every other use case. The same headset rides the subway, sits on an office desk, and takes phone calls, and ANC works in all of those scenarios just as it does at home.

Hybrid models hold two channels simultaneously. The 2.4 GHz dongle carries game audio, Bluetooth holds a phone connection in parallel, and the headset mixes an incoming call into the game without reconnecting. Classic multipoint with two Bluetooth devices solves a different problem, and product descriptions name the two features differently, though they are constantly confused.

Codecs determine how many milliseconds the channel adds. SBC works everywhere and adds the most latency. AAC performs better on Apple devices. LDAC pushes bitrate the highest and pays for it in delay. For music and video the hundred-millisecond difference goes unnoticed, because the player delays the picture to match the audio. A game cannot delay the picture, so the channel requirement is fundamentally different.

Bluetooth handles the microphone differently from the speaker. As soon as the channel enables voice transmission, the profile switches to two-way, and audio quality in the headphones drops to telephone-call level. This is how the standard works, and no model circumvents it. The 2.4 GHz dongle maintains a full-duplex link without that loss, which is why in-game voice chat runs over the dongle while Bluetooth is left for music and incoming calls.

Three connection modes produce three different battery-life figures. Bluetooth is usually more efficient than 2.4 GHz, ANC takes its share in each mode, and the full set therefore contains at least four values. Manufacturers publish one, almost always the largest.

Outdoors, a problem appears that does not exist at home. The external feed-forward microphones pick up wind, and the system converts it into a low-frequency roar inside the cups. Some headsets engage a separate mode with reduced depth, others leave the switching to the user. For a player who takes the headset outside, the availability of a wind mode matters more than an extra ten decibels of rated depth.

Using the headset outside the room brings the weekly dose back into the picture. A headset that plays four hours at home and two more on the commute uses up the limit faster than the gaming time alone suggests. ANC in Bluetooth mode works toward the same result as in-game: it allows lower volume on the subway, where ambient noise reaches levels at which people push volume to the limit.

Custom battery cell with extended charge-cycle life inside a HATOR headset

Your own voice under noise cancellation

The effect named after the French otolaryngologist Étienne Lombard, who described it in 1911, consists of an involuntary increase in vocal loudness. It is triggered in two situations: when ambient noise rises, and when the level at which a speaker hears their own voice drops. The second case is what a closed headset with active noise cancellation creates. The room has become quieter, but the speaker's own voice has also stopped returning to the ears via the air path, and the reflex fires all the same.

According to available measurements, the effect appears when background noise exceeds approximately 43 dB(A), and above 55 dB SPL vocal loudness rises by about 0.38 dB for every decibel of background. Restoring feedback through the headphones reduces this slope.

In practice, a player wearing an ANC headset shouts into voice chat without noticing, and hears complaints from people in the next room. The technical solution has existed for a long time and is called sidetone, the mixing of the user's own microphone signal back into the headphones. In gaming headsets it is usually buried in the companion app under a label like "microphone monitoring." The feature looks secondary until ANC is turned on, at which point it becomes mandatory.

Through bone conduction the player's own voice under closed cups sounds muffled and boomy, with emphasized lows. This shifts the self-volume estimate in the opposite direction, and without sidetone the player has no reliable reference point other than teammates' reactions.

Transparency mode and voice chat

Teammates' voices arrive through the same driver as game audio, so ANC has no effect on them. It only affects sound arriving from outside, which means noise cancellation and voice chat do not conflict with each other. What does conflict with chat is a different technology, ENC on the microphone path, which can noticeably distort vocal timbre. Part one covered this distinction separately, because the confusion between ANC and ENC remains the most common in gaming-headset descriptions.

Transparency mode handles the everyday part of the problem. It uses the same microphones but with the opposite sign: instead of cancelling outside sound, the system passes it through and compensates even for the passive isolation of the cups. Answering a question, hearing a courier, or talking to someone who approached from behind all work without removing the headset or pausing the game.

The toggle is usually mapped to a button and duplicated in the app. Nobody is going to open an app on a smartphone in the middle of a round, and transparency is needed most often at exactly that moment, so the availability of a physical button is worth checking before purchase.

Where ANC will not help

• The operating band ends around 1.5-2 kHz, so the active system does not raise audibility at the frequencies of the footsteps themselves. That job belongs to the ear cushions and cups.

• A voice in the room or a television produces an unsteady broadband signal, and adaptive filters, as part one showed, are most effective on periodic noise.

• ANC does not change how the player is heard by others. That is ENC's job on the microphone path, a separate technology with a separate task.

• Glasses, temples, hair under the cushion, a beanie, or flattened foam let background noise back in, mostly in the band where ANC has already stopped working.

• Spatial Audio, head tracking, and EQ address the positioning task, and ANC does not replace them.

• Some people feel ear pressure from the ANC. This can only be tested in person, before purchase.

Situational awareness disappears along with the hum: a doorbell, a child in the room, a courier at the door. Transparency mode exists for these situations, and part one cited the European standard EN 458, which considers excessive noise attenuation a drawback.

Why players wear passive earmuffs over headsets at tournaments

At large offline tournaments, players often sit in two layers: in-ear monitors carrying game audio, and full-size headphones on top of them. The outer layer blocks the venue, sometimes playing white noise, and in some formats players are additionally placed inside glass booths.

Crowd reactions and caster voices run ahead of the events on screen and reveal information about opponent positions. The venue is a data-leak channel that is sealed mechanically.

Crowd noise and caster speech are concentrated roughly between 500 Hz and 4 kHz, exactly where active noise cancellation weakens and drops off. A crowd's roar is also unpredictable and broadband, while adaptive filters work best on periodic signals. In a venue, passive isolation combined with masking noise handles the task, while ANC in those conditions contributes little.

A home room has the opposite profile. Its main noise is low-frequency, stationary, and predictable, making it an ideal target for an adaptive filter. The same technology produces a different outcome in two different spaces, and the reason is the noise spectrum.

Passive isolation starts where ANC ends

Above 2 kHz all the work falls to the construction. HATOR's ANC headsets use thickened cups and ear cushions made of protein leather, which holds a seal better than fabric.

The mounting method matters as much as the material. Part one noted that in the budget segment cushions are often simply stretched over the cup, producing a loose fit around the entire perimeter. Perimeter-locking mounts cost more to manufacture. This is one of the few parameters that can be evaluated by hand in a store.

Cushion foam compresses over time, and with it the effectiveness of the entire system drops, including the active part. Ear cushions are consumables, and replacing them every year to eighteen months preserves rated performance. Whether replacement cushions are sold separately from the headset is better checked before purchase, because two years later the option may no longer exist.

Protein-leather ear cushions with perimeter-locking mount

Ear-cushion materials and how they change the sound

Isolation forms at the point of contact between the cushion and the head. The material affects it through two properties: how well the surface blocks air, and how closely the foam follows the contour of the skull.

Protein leather holds a seal better than other options thanks to its non-porous surface. The trade-off is heat. The material does not breathe, moisture accumulates under the cups, and sweat and skin oils gradually degrade the coating, so after a year or two the surface begins to peel at the edges and along the folds. In a warm room during summer the difference from fabric is noticeable within half an hour.

Fabric and velour breathe and stay cool, but they pass air and sound along with it. The losses start at the bottom, exactly where the active system has already done its work, and compound with it.

Hybrid cushions split these properties across different surfaces. The outer ring that contacts the head is made of synthetic leather and holds the seal. The inner surface that touches the ear is made of fabric. Isolation is created by the ring. The inner surface plays almost no part in it, so the loss is small and thermal comfort rises noticeably. This trade-off makes sense for players who spend more than a single match in the headset.

Foam handles the second half of the problem. Standard polyurethane is cheap and springy, but it does not follow the contours of a jaw or glasses temples. Memory foam conforms to irregularities and closes the gaps that polyurethane leaves open, but it compresses faster and stiffens in the cold. Gel inserts remove some heat and add weight, which the neck feels over a long session.

Cushion thickness changes more than isolation. It sets the distance from the driver to the ear and the air volume in front of the diaphragm, both of which influence the frequency response. Replacing stock cushions with third-party ones, whether thicker or thinner, shifts the tuning described in the section on bass and, along with it, the threshold at which the upper midrange gets masked. A headset is tuned for a specific cushion, so aftermarket options change comfort and sound at the same time.

For the active system, a poor seal costs more than it does for the passive one. The feedback microphone inside the cup works with the acoustic load that the filter was calibrated against at the factory. A gap near a glasses temple changes that load, the filter receives a different response than it expects, and part of its work misses the target. Passive isolation in the same situation simply lets a bit more sound through and loses exactly as much as the gap opened, whereas active isolation can lose more.

The time for replacement shows several signs. The foam stops springing back after the headset is removed and stays flat, the surface cracks near the mount, and the sound thins out in the low end because the lost seal takes the bass first. ANC hiss in silence also becomes more noticeable, as the system has to work against higher residual noise.

Alcohol and harsh cleaning agents destroy protein-leather coatings, so the surface should be wiped with a damp cloth and allowed to dry. Storing the headset in a closed case immediately after a session traps moisture inside, and this is a leading reason the coating degrades ahead of schedule.

IEMs vs. over-ear headsets for laptop gaming

A laptop puts a player in acoustically different conditions. Its fans sit much closer to the head than tower fans and spin many times faster.

The noise spectrum shifts upward as a result. A fan's blade-passing frequency is the product of RPM and blade count, so a 120 mm tower fan with nine blades at 1,200 RPM produces about 180 Hz, while a laptop turbine at 5,000 RPM with several dozen blades lands in the range of several kilohertz. The first figure sits in the center of ANC's operating band. The second sits above its upper limit.

Active noise cancellation on a laptop therefore delivers less than on a desktop, even at the same background-noise level. The bulk of the work falls to passive isolation: cups, cushions, and fit quality.

In-ear monitors have an advantage in this scenario that over-ear headsets do not. A properly fitted ear tip seals the ear canal and provides high isolation precisely in the upper range, where the laptop is loudest. The ANC loop in IEMs is shorter, because the feedback microphone sits a few millimeters from the eardrum, making it easier for the filter to keep up with the phase. IEMs do not conflict with glasses temples and do not heat the ears, which during a long laptop session matters as much as the sound.

HATOR Hyperpunk Truepods Hybrid ANC

The downsides come in three areas. The air volume in an IEM is small, so ANC's headroom in the low end is less than in an over-ear cup. The occlusion effect, which part one covered, makes the player's own voice and chewing unnaturally loud, and combined with the Lombard reflex this throws off voice-chat volume estimation even further. The microphone in IEMs sits on a cable or in the housing, far from the mouth and close to the laptop, so without ENC the teammates hear the same fans that the player just escaped.

Ear-tip size in IEMs matters more than any rated cancellation depth. Switching from a stock size to the correct one provides more isolation than the difference between two models with different spec-sheet figures. Foam tips expand inside the canal and seal it more tightly than silicone, but they tolerate frequent removal less well and need replacing sooner.

The choice between the two form factors for laptop use comes down to whether voice communication is involved. For solo gaming and travel, IEMs win in isolation and comfort. For team play, an over-ear headset with a boom microphone remains more practical, and part of its disadvantage at high frequencies is compensated by cup thickness and fresh cushion condition.

Bass vs. footsteps

Masking works regardless of where the masker comes from. ANC removes room hum. But if the player then enables a profile with boosted lows, the masker returns, this time inside the cup and at whatever volume the player sets.

This explains the long-running dispute over "gaming" sound signatures. A profile with heavy bass delivers impressive explosions and sells well in stores, but it raises the audibility threshold right where the detail lives. A profile with cut lows removes the problem at the cost of a fatiguing long session and everything except a competitive shooter sounding thin.

In HATOR headsets the lows are present but do not enter the band where they begin to mask the upper midrange. Together with ANC this keeps both sources of low-frequency masking under control, the room and the player's own sound profile. The room half of the problem cannot be solved with EQ.

The same applies to presets in the app. Profiles labeled "cinema" or "bass" are designed for a different use case, and enabling them before a competitive session brings the masker back by the player's own hand.

What to look for in the specs

Points to check in the specification before purchasing.

• System type. A hybrid layout with feed-forward and feedback microphones in each cup has become the standard. A feed-forward-only layout is simpler and cheaper.

• How depth is presented. A band-by-band breakdown is more informative than a single number. A single large number often sums active and passive cancellation.

• Upper limit of ANC operation. A figure around 2 kHz looks good. Figures at 6-8 kHz describe passive isolation.

• Cushion mounting and material, plus availability of replacements for sale.

• Rated battery cycle life. 300 cycles is the industry baseline.

• Battery life separately with ANC on and off. A single figure in the spec is most likely stated without ANC.

• Presence of a physical ANC/transparency toggle button.

• ANC operation across all connection modes, and whether the headset can be used while charging.

• Sidetone availability, i.e. microphone monitoring.

• Wired operation on a dead battery, at least without ANC.

ANC or no ANC?

For a gamer, ANC pays for itself in proportion to how much noise was in the room before it was turned on. In a quiet apartment with a quiet PC, the same money is better spent on ear cushions and fit, because the audibility threshold is already low. In a room with a window facing a busy street, air conditioning, and air-cooled PC fans under load, the active system delivers what no cup construction can.

A headset will change in two or three years, ear cushions sooner, and the user's hearing will stay the same. Of everything described above, only volume has a cumulative effect, the volume at which thousands of hours pass. ANC removes the reason people raise it, but noise cancellation by itself does not regulate playback volume. Managing the headroom it creates is up to the player. Better to manage it wisely, because hearing does not recover.