Full STI vs STIPA — When to Use Each Speech Intelligibility Measurement Method
Both Full STI and STIPA are standardized in IEC 60268-16 as valid methods for measuring speech intelligibility in electroacoustic systems. Both are widely used for voice alarm commissioning, PA system testing, and speech intelligibility assessment. Understanding the technical differences and practical implications helps engineers choose the right method for each application — and the right instrument. This guide provides the definitive comparison of Full STI and STIPA from a practical field measurement perspective.
Technical Differences
Full STI (Speech Transmission Index) uses 98 modulation frequencies — 14 per octave band across 7 octave bands (125 Hz to 8 kHz). It measures the modulation transfer function (MTF) of the complete transmission path (electroacoustic system + room acoustics) at each of these 98 frequencies, then applies a weighting matrix per IEC 60268-16 to produce the final STI value. Full STI was traditionally the definitive, most accurate measurement method but required 10–15 minutes per position. The Bedrock Elite i10 completes Full STI in under 60 seconds.
STIPA (STI for Public Address) uses 14 modulation frequencies — 2 per octave band across the same 7 octave bands. It is designed as a faster approximation of Full STI that can be measured in 15–30 seconds. STIPA results are typically very close to Full STI for well-behaved linear systems. The slight accuracy reduction compared to Full STI is generally considered acceptable for routine PA and voice alarm commissioning work.
Accuracy Comparison
For linear systems with low distortion and moderate noise, Full STI and STIPA agree to within ±0.03 STI in most cases — a difference not perceptible in practical commissioning
For systems with non-linear distortion, clipping, or harmonic distortion, Full STI can produce measurably different (and more accurate) results than STIPA
STIPA can over-estimate intelligibility in severely distorted systems — Full STI is more sensitive to distortion-induced degradation
Both methods are equally sensitive to the effects of reverberation and background noise masking
When to Use Full STI
Life-safety critical voice alarm systems where the highest measurement accuracy is required or specified
Systems with suspected or confirmed non-linear distortion, amplifier clipping, or electronic noise
Projects where the specification explicitly requires Full STI (some advanced specifications or research projects)
Borderline STIPA results (0.45–0.55) where a more definitive measurement would resolve uncertainty
Dispute resolution, arbitration, or regulatory investigations requiring maximum accuracy
When STIPA is Sufficient
Routine commissioning of voice alarm and PA systems per EN 54-16, BS 5839-8, and ISO 7240-19 — standards accept STIPA
Annual inspection of existing voice alarm systems in good working order
PA/GA system acceptance testing where project specifications permit STIPA
Situations where measurement speed is important and the system is known to be linear and well-behaved
Instruments
Bedrock Elite i10 — Full STI in under 60 seconds AND STIPA. The only instrument with both capabilities.