How to Conduct an Acoustic STI Survey

By Sarah Okonkwo · April 1, 2026

When people complain that a room “sounds bad,” they usually mean one of two things: it’s unpleasant (too live, too dead, too boomy), or it’s hard to understand speech. That second problem is where an acoustic STI survey becomes a practical, defensible tool—not a guess. STI (Speech Transmission Index) gives you a standardized way to quantify speech intelligibility in a space, whether you’re tuning a corporate boardroom, verifying a PA system in a venue, or improving a podcast studio that doubles as a remote interview room.

For audio engineers, musicians, podcasters, and home studio owners, STI is a bridge between “sounds fine to me” and measurable performance. It helps you validate mic choice and placement, speaker coverage, room treatment, background noise control, and system gain structure. If you’ve ever mixed a live panel where the audience kept shouting “we can’t hear,” or recorded dialogue that felt clear in headphones but mushy on speakers, STI is the sort of metric that keeps you from chasing your tail.

This guide walks through planning and executing an acoustic STI survey step by step, including equipment options, setup tips, reporting practices, and mistakes that commonly sabotage results.

What STI Actually Measures (and Why It’s More Than Reverb Time)

STI describes how well a speech signal survives the trip from a talker (or loudspeaker) to a listener. It accounts for:

Background noise (HVAC, traffic, projector fans, crowd noise)
Reverberation and early/late reflections that smear consonants
System distortion and non-linearities (overdriven amps, saturated DSP, clipping wireless receivers)
Frequency response and modulation loss that reduce speech clarity

STI ranges from 0 to 1. Higher is better. You’ll often see these rough interpretations used in field work:

0.00–0.30: Bad (speech is largely unintelligible)
0.30–0.45: Poor
0.45–0.60: Fair (borderline for critical spaces)
0.60–0.75: Good (typical target for many installed systems)
0.75–1.00: Excellent

You may also encounter STIPA (Speech Transmission Index for Public Address systems). STIPA is a simplified, practical test method designed for speed and repeatability in real rooms. Many modern meters and software packages use STIPA because it’s faster than full STI and widely accepted for PA verification.

When You Should Run an STI Survey

An STI survey is useful any time speech intelligibility matters and you need proof:

Venue sound checks for spoken-word events (theater dialogue, panels, announcements)
Houses of worship where sermons must be clear across seating areas
Corporate meeting rooms with conferencing and voice-lift systems
Schools (classroom voice reinforcement, lecture halls)
Broadcast/podcast rooms where voice clarity is the product (especially multi-use rooms)
Post-install commissioning for an installed sound system

Real-world example: you’re hired to improve intelligibility in a small venue. The band mix is fine, but the MC’s announcements vanish. STI measurements across audience areas can show whether the problem is excessive reverberation, uneven coverage, or noise floor—so your fixes (speaker aiming, delay alignment, EQ, treatment, gain structure) are targeted rather than random.

Planning the Survey: Define Goals, Coverage, and Conditions

1) Define Pass/Fail Criteria

Before you measure anything, decide what “good” looks like for the room. Targets depend on use case and standards your client follows. Practical targets often look like:

Conference/boardrooms: aim for STI ≥ 0.60 at listener positions
Classrooms: often targeted at STI ≥ 0.60 (higher is better for learning spaces)
PA announcements in noisy environments: STI ≥ 0.50 may be acceptable depending on constraints

2) Map the Room and Choose Measurement Points

Pick positions that represent real listeners, not just convenient locations. Use a simple grid or zones:

Front, middle, rear seating
Left/center/right coverage areas
Problem zones (under balconies, corners, near reflective back walls)

Tip: In a live venue, include at least one point near FOH (where decisions are made) and a few points where complaints happen (rear corners are common).

3) Control the Test Conditions

STI is sensitive to noise. Decide whether you’re measuring:

Quiet-room potential (empty room, HVAC set to typical operation)
Operational reality (projectors running, typical audience noise expectations, machinery operating)

Write down conditions so results are defensible: HVAC on/off, doors open/closed, occupancy, time of day, and any active equipment.

Equipment You’ll Need (Options for Different Budgets)

Core Measurement Tools

STI/STIPA-capable meter or analyzer (hardware meter or software with an audio interface)
Measurement microphone (calibrated omni is typical for room measurements)
Signal source for STIPA test signal (dedicated generator, software playback, or analyzer output)
Loudspeaker under test (the installed system, or a reference source if appropriate)
Tripod/mic stand with repeatable height markings

Recommended Add-Ons for Reliable Results

Acoustic calibrator (e.g., 94 dB @ 1 kHz) to verify mic sensitivity and measurement chain
SPL meter function (often built into STI meters) to document test level and background noise
Laser distance measurer to document mic locations and source distances
Camera/phone to photograph mic positions and speaker aiming

Hardware Meter vs. Software Workflow

Hardware STI/STIPA meters are fast, consistent, and less prone to routing mistakes. Great for field verification, installed sound commissioning, and repeatability.
Software-based STI measurement can be cost-effective and flexible, but depends heavily on correct audio interface gain staging, routing, clock stability, and calibration. It’s excellent for engineers already comfortable with measurement software and signal flow.

Practical comparison: If you’re doing a one-off check in a home studio, software plus a calibrated mic may be enough. If you’re documenting results for clients across multiple sites, a dedicated STIPA meter typically pays off in speed and fewer “oops” moments.

Step-by-Step: How to Conduct an Acoustic STI Survey

Step 1: Verify the System You’re Measuring

Before generating any test signal, confirm the signal chain is stable:

Correct speaker zones and DSP presets are active
No limiters are clamping unexpectedly
Wireless mics and processors aren’t clipping
Time alignment/delays are in their intended configuration

Live event scenario: If a venue uses different DSP presets for “band night” vs. “spoken word,” measure the spoken-word preset. Measuring the wrong preset is one of the fastest ways to produce useless data.

Step 2: Calibrate and Set Levels

Calibrate the measurement mic (if you have a calibrator). Record the calibration value and time.
Measure background noise at each zone (or at least representative positions). Capture LAeq over 15–30 seconds if your meter supports it.
Set the test signal level to a realistic speech level at listener positions. A common approach is to target around 65–75 dB SPL in the audience/listener area, depending on the application and ambient noise.

Tip: Don’t blast the STIPA signal just to “get a better number.” Speech intelligibility in real use depends on actual operating levels and noise conditions. Overdriving the system can also trigger protection/limiters, which may reduce STI anyway.

Step 3: Choose Source Position(s)

In many surveys, the “source” is the installed loudspeaker system. In other cases, you may evaluate talker-to-listener performance (for example, a classroom without amplification).

PA system survey: Use the system loudspeaker(s) as the source. Measure per zone or per loudspeaker group if needed.
Room-only survey: Use a reference loudspeaker at the talker position (lectern, stage center) aimed realistically.

Real-world example: In a theater, you might run STI with the center cluster only, then with under-balcony fills included, to confirm that fills improve intelligibility without adding destructive reflections or level inconsistencies.

Step 4: Microphone Placement (Listener Positions)

Consistency is everything. Place the mic where ears are:

Seated audience: 1.1–1.2 m (about 43–47 in) above floor
Standing listeners: 1.5–1.7 m (about 59–67 in) above floor

Use an omni measurement mic pointing upward (common approach) unless your specific method states otherwise. Keep the mic away from tabletops, seat backs, and walls that can create near-field reflections.

Step 5: Run Measurements and Log Everything

Play the STIPA test signal through the system at the established level.
Measure STI/STIPA at each location. Many analyzers produce a stable reading within 10–20 seconds.
Repeat if needed to confirm repeatability (especially if readings seem unstable).
Document:
- STI/STIPA value
- Test level at mic position (SPL)
- Background noise (SPL or LAeq)
- Speaker zones active
- Mic position ID and height
- Room conditions (HVAC, occupancy, doors)

Tip for efficiency: Label positions like A1–A6 (front), B1–B6 (middle), C1–C6 (rear). Photographs of each mic spot save hours when you return for a re-test after acoustic treatment or system tuning.

Step 6: Interpret the Results Like an Engineer (Not Just a Number Collector)

If STI is low, use the supporting context to narrow causes:

High background noise + okay reverb: Reduce HVAC noise, isolate doors/windows, improve SNR with better mic technique or distributed speakers.
Good noise floor but poor STI in the rear: Coverage issue—consider aiming, adding delays/fills, or fixing level distribution.
Poor STI everywhere with long decay: Room acoustics—consider absorption, diffusion strategy, and controlling flutter echoes/late reflections.
STI drops when fills are added: Possible timing/delay misalignment causing destructive modulation/echo perception; check delay times and polarity.

Studio scenario: A podcast room may have low noise but still measure mediocre STI if there are strong early reflections off a desk, untreated side walls, or a glass door near the mic position. Even small changes—desk mat, gobos, side absorption—can improve perceived clarity and measured intelligibility.

Practical Tips That Make STI Surveys Go Smoother

Measure at realistic operating settings. Use the same EQ, compression, and DSP routing used during actual events or sessions.
Stabilize the room. Close doors, pause HVAC changes, and keep people from talking during measurements.
Use multiple runs in problem zones. If a corner reads inconsistently, that’s a clue: reflections or noise fluctuations may be dominating.
Check speech chain headroom. If the PA processing is squashing the test signal or clipping, fix gain structure first.
Pair STI with a quick RT60/EDT snapshot if your tool allows. It helps explain why intelligibility is suffering (or why changes worked).

Common Mistakes to Avoid

Measuring too loud and calling it “excellent intelligibility.” If the system only achieves that number at unrealistic levels, it’s not meaningful.
Ignoring background noise measurements. STI without documented noise floor is missing half the story.
Bad mic placement (too close to walls, on tables, near reflective surfaces) creating artificially low or inconsistent readings.
Testing the wrong system configuration (wrong DSP preset, wrong zones active, wrong mic processing in the chain).
Not repeating questionable readings. One odd value can be a fluke—or it can reveal a real acoustic problem. Repeat to confirm.
Failing to document conditions. If you can’t describe the room state, you can’t compare “before/after” reliably.

What to Do After the Survey: Turning Results Into Improvements

Once you’ve mapped STI across the space, the next step is to choose fixes that match the root cause. Common improvement paths include:

Improve SNR: quieter HVAC, noise control, better mic technique, closer miking, more appropriate microphone polar patterns, distributed speakers instead of a single loud source.
Improve coverage: adjust speaker aiming, add front fills/under-balcony fills, verify delay and polarity, smooth level distribution across zones.
Control reflections/reverb: add absorption at first reflection points, treat rear wall slap, reduce flutter echo, consider ceiling clouds in reflective rooms.
Optimize processing: avoid excessive compression that raises noise floor; use EQ to improve speech band clarity (typically 1–4 kHz) without harshness; confirm limiters aren’t constantly engaged.

Actionable workflow: Make one change at a time, then re-measure the same points. STI surveys are most powerful as a before/after verification tool.

FAQ: Acoustic STI Surveys

What’s the difference between STI and STIPA?

STI is the full Speech Transmission Index method. STIPA is a streamlined version designed for faster measurements in typical PA and room scenarios. STIPA is widely used for installed sound verification because it’s practical and repeatable.

Do I need a dedicated STI meter, or can I do this with software?

You can do it with software if it supports STI/STIPA and you can calibrate your measurement chain reliably. Dedicated meters are usually faster, more foolproof in the field, and easier to standardize across projects.

How many measurement points should I take?

For small rooms, 6–12 points often gives a solid picture. For venues, 12–30+ points may be appropriate, especially if there are balconies, fills, or multiple seating zones. Focus on real listener locations and known problem areas.

Should the room be empty or full of people during the test?

Measure under conditions that match how the space is used. An empty room may overstate reverberation compared to a seated audience. If you can’t measure with a crowd, document the condition and consider a follow-up measurement during a rehearsal or with simulated absorption (seat pads/soft goods) where practical.

What STI value should I aim for in a podcast or voiceover room?

Podcast rooms are close-mic environments, so STI isn’t always the primary metric, but intelligibility still matters for playback in the room and remote monitoring setups. If you’re using loudspeakers for playback/monitoring or conferencing, aiming for STI ≥ 0.60 in listener positions is a sensible starting point—then prioritize low noise floor and controlled early reflections for the actual recordings.

Can EQ alone fix poor STI?

Sometimes EQ helps, but it’s rarely the whole fix. If reverberation, reflections, noise, or coverage issues are the main cause, EQ won’t solve intelligibility by itself. STI surveys help you avoid over-EQ’ing when the real problem is acoustic or geometric.

Next Steps

Plan your measurement points, document the room conditions, and run a baseline STI/STIPA pass before you change anything. Then apply targeted improvements—noise control, coverage optimization, acoustic treatment, or DSP adjustments—and re-test the same locations. That before/after comparison is what turns an STI survey from a number into a decision-making tool you can stand behind.

For more practical measurement workflows, studio acoustics tips, and real-world gear guidance, explore more guides on sonusgearflow.com.