In-depth insight into language clarity (intelligibility)(1)

In daily communication, the speech clarity of face-to-face conversations often feels natural and effortless. However, when it comes to sound reinforcement, maintaining this clarity becomes challenging. This article aims to delve into the key elements of language clarity (intelligibility) and share some practical tips to ensure that high-quality speech clarity is maintained in a variety of scenarios.

Overview of core points

Language comes from oral expression. Therefore, when amplifying speech, you should always consider how to maintain the clarity (intelligibility) of the language.

Speech generation: Sound is generated through the vibration of the vocal cords, and the volume and pitch can be changed through the control of the vocal cords. The filtering of the sound spectrum is influenced by cavities such as the pharynx, mouth, and nose.

Recording Tips: Changing the intensity of the sound can change the volume and spectrum of the sound. Even the pitch of the sound changes with the intensity of the sound. For example, shouting sounds significantly different from normal speech. Therefore, when recording, it is crucial to ensure that the peaks of the voice signal are captured, as they can rise significantly above the RMS or average volume. Similarly, consonants (k, p, s, t, etc.) appearing in the frequency range above 500hz, especially those in the frequency range of 2 kHz-4 kHz, are crucial for the clarity of non-tonal languages.

Distance and location: When the speaker is about one meter away, the clarity and naturalness of the voice are usually at their best. We will feel that his voice is natural. Standing next to or behind a person will reduce the naturalness and clarity (intelligibility). Different positions (such as side or rear) may affect the perception of sound. In fact, clear sounds can be heard from almost anywhere except when we approach the speaker with our ears or microphones.

Timbre and frequency: Different positions of the head, chest or body give the sound a unique timbre. For example, the speech spectrum at the chest position may lack key frequencies in the 2kHz-4kHz range, which can lead to reduced speech clarity. If the microphone cannot compensate for this, you should use an equalizer to correct it.

Speech as sound source

Speech is not only a tool for communication, but also a complex acoustic signal. Although language itself is universal, everyone's voice characteristics and sound quality are unique. Understanding these differences is crucial to ensuring clarity in recording and sound reinforcement.

sound level

Because people use different degrees of force when speaking, from softly whispering to loudly shouting, the volume of speech varies greatly, it is difficult to assign a fixed number to the volume of speech. The values in the table below represent the average A-weighted speech volume for adults. It is worth noting that the ability to understand speech is optimal when the speech volume corresponds to normal speech levels at a distance of 1 meter. In other words, a sound pressure level of approximately 55-65 dB is 20 Pa.

(In this case, the reference is the lowest audible sound pressure level)

Averagespeech level as a function of listening/recording distance. There is a nearly 20 decibels difference between normal speech and shouting.

peak

Also pay attention to the peak factor-the ratio between the peak level and the RMS level-which is crucial to avoid distortion in recording or pickup.

Each level shown in the table above is an average RMS level, not a peak level. Typically, the peak value is 20-23dB above the RMS level. The ratio between the peak level and the RMS level is called the peak factor. This factor is an important parameter when recording or reproducing speech in an electro-acoustic system. Please also note that loud singing measured at the lips can reach a level of 130 dB (reference 20 Pa RMS), with peak levels exceeding 150 dB(reference 20 Pa RMS).

​

Male voice, speak normally (duration 18 seconds). Average RMS: -21.5 dBFS, peak: -0.5 dBFS. Peak coefficient 11(21 dB). The red dotted line represents the RMS level.

practical advice

Choose the right microphone: Choose the right microphone for a specific application scenario to ensure high-audio detail is captured.

Make good use of equalizers: Use equalizers to compensate for missing frequencies caused by the recording environment or microphone position.

Optimize the recording environment: Avoid background noise and echoes as much as possible, use sound-absorbing materials in the space or take other acoustic treatment measures.

Pay attention to distance and position: Maintain an appropriate distance and position from the speaker to maximize the clarity and naturalness of the speech.

Follow these suggestions and have an in-depth understanding of the principles of speech generation, and you will easily ensure high-quality speech clarity in a variety of occasions. Let's look forward to the next issue and continue to explore more mysteries of language clarity!