Speech Transmission Index
The Speech Transmission Index , shortly STI (of English Speech Transmission Index ) is a measure of the quality of a transmission link language transfers from the speaker to the listener. An acoustic or electro-acoustic voice signal transmission is understood as a transmission path . The measure STI describes the speech intelligibility to be expected by the listener.
Influencing factors
The voice transmission index depends on
- the volume of the speech signal for the listener
- the background noise from speaker and listener
- Echoes ( reflections with delays > 100 ms)
- the reverberation time of the listening room
- psychoacoustic masking effects
- the distortion and the frequency response of the transmission path
- the quality and directional characteristics of the sound source .
Applications
Typical applications are measurements of speech intelligibility
- Rooms without an electro-acoustic system
- Public address and evacuation systems
- of telecommunications
scale
The voice transmission index indicates the transmission quality as a numerical value in the range from 0 (= incomprehensible) to 1 (= excellent).
The typical minimum requirements for public address and evacuation systems are: ≥0.45 STI or ≥0.5 STI.
In addition to the STI, there is also the CIS scale ( Common Intelligibility Scale according to Barnett 1995, 1999), which is related to the STI as follows:
STIPA
STI-PA ( S Peech T ransmission I ndex for P ublic A ddress system ) is a simplified quick measuring method of the speech intelligibility of a transmission link that takes into account electro-acoustic and acoustic effects. A typical STI-PA application is the measurement of speech intelligibility in public address and evacuation systems. This verifies whether emergency announcements can be understood by the people present in an emergency.
Speech can be described as noise in the frequency range 125 Hz - 10 kHz, with amplitude modulation using low-frequency sinusoidal signals (0.63 Hz - 12.5 Hz). Such a speech-like signal is used as a STIPA test signal to simulate the human speaker. The measuring device measures the modulation of the received signal and compares it with the original test signal in the octave bands 125 Hz - 10 kHz. A loss of modulation depth indicates reduced speech intelligibility.
history
Initially, STIPA's speech intelligibility was determined from 12 individual combinations of octave band and modulation frequency, which covered the speech as much as possible. The measurements of the low-frequency speech components (frequency bands 125 Hz and 250 Hz) have been combined for simplification.
In 2003 NTi Audio developed the extended STIPA measurement with individual coverage of all modulation frequencies of speech. This development improved the correlation of subjective speech intelligibility with the STI reading and was standardized in the IEC - Standard 60268-16, Issue 2011th
See also
- Alcons , scale for consonant intelligibility
- Mean Opinion Score
- SINPO system , scales for the connection quality in radio
Individual evidence
- ↑ THE MEASUREMENT OF SPEECH INTELLIGIBILITY Herman JM Steeneken TNO Human Factors, Soesterberg, the Netherlands (PDF; 208 kB)
- ↑ Barnett, PW and Knight, RD ( 1995 ). "The Common Intelligibility Scale," Proc. IOA Vol 17, part 7.
- ↑ Barnett, PW ( 1999 ). "Overview of speech intelligibility" Proc. IOA Vol 21 Part 5.