AES-2id: Averaging Strategies and Spatial Sampling
TL;DR
AES-2id describes two forms of averaging: synchronous averaging at a single position (to improve SNR) and spatial averaging across multiple positions (to characterize room behavior). Synchronous averaging of N identical measurements at the same position improves SNR by 10·log10(N) dB — 4 averages add 6 dB. Spatial averaging across positions reduces uncertainty in spatially-varying parameters. For convolution reverb, a single representative position may suffice. For ISO 3382 compliance, the number of positions follows the survey/engineering/precision grades.
Synchronous Averaging
Repeating the measurement at the same position and coherently averaging the impulse responses reduces uncorrelated noise while preserving the deterministic room response:
SNR improvement = 10 · log₁₀(N) dB
| Averages (N) | SNR gain | Measurement time (10s sweep) |
|---|---|---|
| 1 | 0 dB | ~15 s |
| 2 | +3 dB | ~30 s |
| 4 | +6 dB | ~60 s |
| 8 | +9 dB | ~2 min |
| 16 | +12 dB | ~4 min |
Synchronous averaging requires that the room impulse response is time-invariant between measurements. Movement of people, doors, or temperature changes can decorrelate the response and reduce the averaging benefit. For swept sine measurements, time variance primarily affects high frequencies where SNR may already be limited.
Spatial Sampling
Room acoustic parameters vary with position. T30 is relatively spatially uniform (standard deviation 3-5%), while C80 and G vary significantly (standard deviation 1-3 dB). Spatial averaging across multiple receiver positions provides a representative characterization of the room.
Grid Design
AES-2id recommends distributing measurement positions on a regular or quasi-random grid that covers the audience area. Key considerations:
- Avoid placing receivers on nodal lines of room modes (especially in rectangular rooms at low frequencies)
- Spacing between receivers should exceed half a wavelength at the lowest frequency of interest (~2 m for 80 Hz)
- Include positions at various distances from the source to capture the distance-dependent behavior
- Include positions near boundaries if listeners sit near walls
Application-Specific Recommendations
- Room characterization (ISO 3382): Follow the survey (6), engineering (12), or precision (20+) position counts
- System alignment: Measure at key audience positions — front, center, rear, sides, under balcony
- Convolution reverb: 1-3 representative positions may suffice; binaural capture (HRTF or dummy head) provides spatial realism
- STI verification: Per IEC 60268-16, measure at all listener positions within each PA zone
Post-Processing Averaged Results
Spatial averaging of parameters (not impulse responses) is the standard approach. Averaging impulse responses directly would create interference patterns and is not recommended except for coherent sources in special cases.
SonaVyx Multi-Capture
The IR tool supports both synchronous averaging (multiple sweeps at one position) and multi-position capture through the session system. Results are stored per position with automatic spatial averaging. Export all positions as individual WAV files per AES-2id format guidelines.
Try It Now
Open this measurement tool in your browser — free, no download required.
Last updated: March 19, 2026