Speech Transmission Index (STI) FAQ
Comprehensive answers about STI — the objective measure of speech intelligibility. Covers the STI scale, STIPA shortcut, measurement procedures per IEC 60268-16, and strategies for improving speech clarity.
Quick Navigation
- 1. What is the Speech Transmission Index (STI)?
- 2. What do STI scores mean in practice?
- 3. What is the difference between STI and STIPA?
- 4. What is the minimum STI required for a classroom?
- 5. How can I improve STI in an existing room?
- 6. How is STI measured in the field?
- 7. What is the relationship between STI and RT60?
- 8. How does STI apply to open plan offices?
- 9. What does IEC 60268-16 require for STI measurement?
- 10. How does AcousPlan calculate STI?
What is the Speech Transmission Index (STI)?
The Speech Transmission Index (STI) is a machine-computed metric ranging from 0 to 1 that quantifies how well speech is transmitted from a talker to a listener in a given acoustic environment. Defined in IEC 60268-16:2020 §4, STI accounts for the degrading effects of reverberation, background noise, and echo on the modulation transfer function of speech. A score of 1.0 represents perfect transmission; 0.0 represents complete unintelligibility. STI considers 7 octave bands (125–8000 Hz) and 14 modulation frequencies (0.63–12.5 Hz), weighting each contribution according to its importance for speech understanding. Unlike subjective listening tests, STI is repeatable and objective. AcousPlan calculates STI automatically from your room's RT60, background noise level, and source-receiver geometry.
What do STI scores mean in practice?
STI scores map to qualitative intelligibility categories defined in IEC 60268-16:2020 Table 1: 0.00–0.30 is "bad" (less than 34% sentence intelligibility), 0.30–0.45 is "poor", 0.45–0.60 is "fair", 0.60–0.75 is "good", and 0.75–1.00 is "excellent" (above 96% sentence intelligibility). For classrooms, ANSI S12.60-2010 effectively requires STI ≥ 0.60 (good). BB93:2015 requires STI ≥ 0.60 for teaching spaces. Public address systems in transport hubs target STI ≥ 0.50 per BS 7827. Fire alarm voice systems require STI ≥ 0.50 per BS 5839-8:2013. Each 0.05 increase in STI represents a perceptible improvement in intelligibility. AcousPlan colour-codes STI results and provides specific recommendations to reach your target category.
What is the difference between STI and STIPA?
STI is the full Speech Transmission Index computed from 98 modulation transfer function values across 7 octave bands and 14 modulation frequencies. STIPA (Speech Transmission Index for Public Address) is a simplified field measurement method defined in IEC 60268-16:2020 §5, using a specially modulated test signal with only 14 modulation frequencies — two per octave band. STIPA requires approximately 15 seconds of measurement per position versus several minutes for full STI. The correlation between STIPA and full STI is typically within ±0.03 for most rooms. STIPA is the industry-standard method for commissioning PA/VA systems and verifying speech intelligibility in completed buildings. Use full STI for design-stage predictions in software and STIPA for on-site verification.
What is the minimum STI required for a classroom?
The minimum STI for a classroom is 0.60, corresponding to the "good" intelligibility category. ANSI S12.60-2010 does not directly specify STI but its RT60 and background noise criteria (0.6 s, 35 dBA) yield STI ≥ 0.60 when met. BB93:2015 Table 1.2 explicitly requires STI ≥ 0.60 for general teaching spaces. For spaces serving pupils with special educational needs and disabilities (SEND), best practice recommends STI ≥ 0.70 to compensate for auditory processing difficulties. Achieving this typically requires RT60 ≤ 0.4 s and background noise ≤ 30 dBA. Key improvement strategies include high-performance acoustic ceilings (NRC ≥ 0.90), quiet HVAC systems (NR 25), and sound-field amplification where geometry prevents adequate direct sound coverage.
How can I improve STI in an existing room?
Improving STI requires addressing its two main degradation factors: excessive reverberation and high background noise. First, reduce RT60 by adding absorption — install Class A ceiling tiles (NRC ≥ 0.90) across 80–100% of the ceiling, and add wall panels at the rear and side walls covering 30–40% of wall area. This alone can improve STI by 0.05–0.15. Second, reduce background noise — replace or attenuate HVAC diffusers to achieve NR 30 or below, seal gaps around doors and windows, and address external noise with secondary glazing if needed. Third, improve the direct-to-reverberant sound ratio by reducing the source-to-listener distance or using a sound reinforcement system designed for speech (directional loudspeakers aimed at the audience). AcousPlan's auto-solve feature recommends the most cost-effective combination of these strategies.
How is STI measured in the field?
Field STI measurement typically uses the STIPA method per IEC 60268-16:2020 §5. Equipment needed includes a STIPA signal generator (either a dedicated device or software playing a standardised modulated noise signal through an omnidirectional loudspeaker at representative talker level, typically 60 dBA at 1 m) and a STIPA analyser (handheld meter or software with calibrated microphone). Position the source at the talker location (e.g., the teacher's position) and measure at multiple listener positions across the audience area. Each measurement takes approximately 15 seconds. Report the minimum, average, and spatial distribution of STIPA values. Conduct measurements under representative conditions — HVAC running, room unoccupied but furnished. The measurement uncertainty is typically ±0.03 STI units. Document source level, meter calibration date, and ambient conditions.
What is the relationship between STI and RT60?
STI and RT60 are inversely correlated — as RT60 increases, STI decreases because reverberation smears the temporal envelope of speech. Per IEC 60268-16:2020 Annex J, in a diffuse field with negligible background noise, a room with RT60 of 0.4 s yields STI ≈ 0.78, while RT60 of 1.0 s yields STI ≈ 0.58, and RT60 of 2.0 s yields STI ≈ 0.42. However, this relationship is not linear, and background noise introduces an independent degradation. A room with RT60 of 0.5 s but 45 dBA background noise may have lower STI than a room with RT60 of 0.8 s and 25 dBA background noise. Both parameters must be controlled simultaneously. AcousPlan calculates STI considering both reverberation and background noise, showing you which factor is dominant.
How does STI apply to open plan offices?
In open plan offices, STI is used inversely — the goal is low STI between workstations to ensure speech privacy. ISO 3382-3:2012 §4 defines the spatial decay rate of speech (D₂,S) and A-weighted speech level at 4 m (Lp,A,S,4m), both of which correlate with STI. A well-designed open office achieves STI < 0.20 at the distraction distance (rD, where STI drops below 0.50) and STI < 0.05 at the privacy distance (rP, where STI drops below 0.20). Achieving this requires high-performance acoustic ceilings (Class A), desk screens of at least 1.2 m height, carpet flooring, and often electronic sound masking at 40–45 dBA. AcousPlan's speech privacy calculator models D₂,S and predicts STI at any distance from the source.
What does IEC 60268-16 require for STI measurement?
IEC 60268-16:2020 is the definitive standard for STI and specifies both computation and measurement procedures. Key requirements include: the full STI computation uses 7 octave bands (125–8000 Hz) with 14 modulation frequencies per band (§4.1); the STIPA simplified method uses 2 modulation frequencies per band with a specific test signal (§5.1); the source must be omnidirectional and calibrated to representative speech level (§5.2.2); a minimum of 15 seconds acquisition time for STIPA; redundancy correction factors for correlated modulation effects between bands; and weighting factors per band that reflect the speech spectrum. The standard also defines the CIS (Common Intelligibility Scale) as an alternative descriptor. Results must be reported with measurement conditions, equipment specifications, and uncertainty estimates.
How does AcousPlan calculate STI?
AcousPlan calculates STI using the modulation transfer function method defined in IEC 60268-16:2020 §4. The calculation chain works as follows: first, RT60 is computed per octave band (125–4000 Hz) using Sabine or Eyring equations from your room geometry and material selections. Second, the modulation transfer index for each octave band and modulation frequency is derived from the ratio of reverberation time to modulation period, incorporating background noise levels. Third, the effective signal-to-noise ratio is computed with truncation limits of ±15 dB. Fourth, modulation transfer indices are averaged and weighted across bands using the standard speech-weighted factors. The result is an STI value from 0 to 1 with a qualitative rating. All calculations are advisory — professional verification is recommended for safety-critical applications like fire alarm voice systems.
Try AcousPlan Calculator
Put these answers into practice. Model your room, select materials, and get instant RT60, STI, and compliance results.
Open Free CalculatorRelated FAQ Topics
Explore More
Browse all 30 acoustic FAQ hubs by category or search across topics.
Definitions for every acoustic term, from RT60 and STI to NRC and flanking.
Acoustic targets and design guidance for every room type.
Free RT60, STI, and noise criteria calculator with 5,000+ materials.