Microphones

Microphone quality is critical for voice interaction quality. In public environments, background noise – music, crowds, HVAC systems, other conversations – is the most common source of speech recognition errors. Choosing the right microphone and positioning it correctly has more impact on conversation quality than almost any other hardware decision.

Microphone TypesCopy Anchor Link

Directional (cardioid) microphonesCopy Anchor Link

Directional microphones pick up sound primarily from in front of them and reject sound from the sides and rear. They are the most practical choice for most kiosk installations because they naturally attenuate ambient noise from sources that are not the user.

Recommended for: standard retail, hotel lobby, museum guide scenarios.

Beamforming microphone arraysCopy Anchor Link

Beamforming arrays use multiple microphones and digital signal processing (DSP) to focus a pickup "beam" toward the user, while actively cancelling noise from other directions. They provide the best performance in high-noise environments.

Recommended for: busy retail floors, event spaces, airports, food court environments.

Boundary (surface) microphonesCopy Anchor Link

Boundary microphones are flat and sit on a horizontal surface. They pick up audio from above the surface via reflection. Suitable for counter-top or desk-mounted kiosks where the user is close to the surface. Not ideal for free-standing kiosks.

Omnidirectional microphonesCopy Anchor Link

Omnidirectional microphones pick up sound from all directions equally. Only suitable for very quiet environments (meeting rooms, private offices). Not recommended for public kiosk deployments.

Recommended MicrophonesCopy Anchor Link

For stores, airports, and similar noisy public spaces, prefer a beamforming array with on-board DSP (noise suppression + AEC) and a limited pickup radius – the microphone should hear the user standing 30–60 cm in front of the kiosk, not the entire hall. Pair it with correct positioning below; no microphone compensates for being mounted next to a speaker or an HVAC vent.

Choose and Test in the Target EnvironmentCopy Anchor Link

There is no universally "best" kiosk microphone – microphones should be chosen and tested in the specific environment where the kiosk will operate. Every room and space has different acoustic conditions:

• Reverberation and echo – hard floors, glass walls, and high ceilings reflect sound; a microphone that performs well in a carpeted office may struggle in a marble lobby
• Noise profile – the character of the noise matters as much as its level: constant HVAC hum, background music, PA announcements, and crowd chatter each affect speech recognition differently
• Noise variability – a store at opening time and at Saturday peak are effectively two different acoustic environments; test at the loudest realistic time, not the quietest

Whenever possible, shortlist two or three candidate microphones and run comparative tests on site, at the actual installation spot, before committing – following the steps in Testing Microphone Performance below.

Guiding Users into the Pickup ZoneCopy Anchor Link

A directional or beamforming microphone only helps if the user actually stands where the beam is focused. Rather than expecting users to find the right spot themselves, design the installation so they end up there naturally:

• Floor stickers or markings – a "stand here" decal placed 30–60 cm in front of the display marks the ideal speaking position and distance
• Branded frame around the screen – a frame or overlay that visually centres the user's attention also centres their body in front of the microphone
• Physical layout cues – counters, railings, mats, or kiosk enclosure shape can funnel the user into the pickup zone without any signage
• On-screen prompts – the avatar's gaze direction and a subtle "come closer" prompt when face detection fires can pull users into range

These visual cues cost little and often improve recognition accuracy more than a microphone upgrade – the best microphone in the wrong geometry loses to a mid-range microphone with the user reliably in its beam.

Microphone PositioningCopy Anchor Link

HeightCopy Anchor Link

Position the microphone at approximately 130–150 cm from the ground (mouth height for a standing adult). In kiosk enclosures, this typically means mounting the microphone near the top of the enclosure front panel.

DistanceCopy Anchor Link

Speech recognition quality degrades with distance. The ideal pickup distance is 30–60 cm from the user's mouth. For free-standing kiosks, this means the microphone should be at or near the display surface – not recessed deep inside the enclosure.

Avoiding interferenceCopy Anchor Link

Keep the microphone away from:

• Speaker outputs (echo and feedback risk)
• HVAC vents (handling noise and wind noise)
• High-traffic paths where people walk past closely (transient noise)
• Metal surfaces that can cause resonance

Echo CancellationCopy Anchor Link

The Virbe Kiosk app uses audio echo cancellation (AEC) to prevent the TTS speaker output from being picked up by the microphone and processed as user speech. Ensure:

• The speaker and microphone are correctly identified in Windows audio settings
• The selected STT engine in the Virbe profile is configured to use the correct microphone input (check the Windows default recording device)
• The speaker volume is appropriate – very high volumes increase the AEC challenge and can cause over-cancellation

Testing Microphone PerformanceCopy Anchor Link

Before finalising the installation:

1. Use the Test conversation panel in the Conversation Editor to conduct voice tests
2. Test from the expected user position and distance
3. Test with ambient noise at the expected level (play representative background noise if possible)
4. Check whether the STT engine correctly transcribes speech at the expected accuracy level
5. Adjust microphone gain in Windows Sound settings if the input level is too low or too high (aim for peaks around -12 to -6 dBFS in Windows sound meter during normal speech)

Cookie settings