Smart Speaker Testing - Speech and Hearing
Voice-operated devices such as smart speakers have taken households by storm. With consumer satisfaction closely linked to how well smart speakers understand voice commands; speech intelligibility and overall sound quality are crucial factors in meeting today's customer demands.
Realistic Testing of Smart Speakers
An often-used method to measure response accuracy rate in voice-operated devices is called "RAR". The idea is to replay a recorded voice from a speaker in order to evaluate if and how often a voice command is recognized. However, as the directional character and frequency range of a human voice is different, this method can produce misleading results.
Smart speaker testing standards and requirements
In a room, good acoustics require a design that is fit for purpose, such as easy communication and a high degree of intelligibility in office spaces or long reverberation times in concert halls. Issues with acoustics are normally caused by sound being reflected too much, too little or in the wrong direction.
Speech and hearing simulation
To achieve the best result, Brüel & Kjær offers market-leading Mouth and Ear simulators, which provide highly accurate means to secure quality in testing of voice-operated products.Brüel & Kjær’s High-frequency Head and Torso Simulator complies with key features defined in the ITU-T P.58 standard for Objective Measurement Apparatus.
High-Frequency HATS Type 5128
Capable of both issuing voice commands and measuring smart speaker response (full frequency range up to 20 kHz) makes fully automated testing of smart speakers and other voice operated devices possible.
Mouth Simulator Type 4227
Brüel & Kjaer's Type 4227 Simulator is a high-performance artificial mouth, which simulates the mode of diffusion of human speech.
The compactness and robustness of this artificial mouth is ideal for use in an R&D laboratory or on a test bench in production. Its design quality - compliant with ITU-T P.51 test standards - guarantees reliable and reproducible measurements, even over long periods.
To reproduce a realistic human voice setup for smart speaker testing, acoustic speech dispersion comes into effect. Dispersion and attenuation of amplitude effects speech level in various angle and distance.
Brüel & Kjær’s High-frequency Head and Torso Simulator complies with key features defined in the ITU-T P.58 standard for Objective Measurement Apparatus, which defines attenuation values in relation to the Mouth Reference Point (MRP).*
*A position 25mm in front of the Lip Reference Plane (LRP) and defined in terms of dB attenuation relative to 65.3dB SPL at a distance of 500 mm in front of the MRP or 89.3 dB SPL at the MRP. Points are located on a circle centred at MRP and distributed in the horizontal plane at 0°, ±15°, ±30° and ±90° as well as in the vertical plane at ±15° and ±30°.