Spatial Averaging
Canonical Definition
Spatial averaging combines measurements from multiple defined positions to estimate the response of a listening region rather than one point. The result depends on microphone geometry, weighting, level alignment, averaging math, and whether phase is retained.
Why Average Positions?
- reduce the influence of one fragile microphone location
- represent normal head movement around a primary seat
- balance more than one listening position
- identify response features that remain consistent across an area
- separate local comb filtering from broader system trends
An average can improve the decision, but only when the positions and weights match the real listening goal.
Averaging Methods
| Method | What is combined | Phase retained? | Typical use |
|---|---|---|---|
| RMS magnitude average | Linear magnitudes or powers converted from dB | No | General spatial response and EQ target work |
| dB average | Displayed dB values | No | Alternative magnitude summary; result differs from RMS average |
| Vector or complex average | Magnitude and phase | Yes | Coherent arrays or aligned measurements with stable relative timing |
| Moving-microphone average | Continuous samples across a path or volume | No useful fixed-position phase | Fast broad tonal average for a region |
Level Alignment And Weighting
Measurements from different positions may have different overall levels because of source distance and geometry. If the goal is to average response shape, level-align the traces over a documented frequency span first. If the goal is to preserve real seat-to-seat level differences, do not remove them.
Weights can make the primary driver-seat region count more than a passenger or rear seat. Record the weight for every position; an unlabeled average hides the design priority.
Keep The Spread
Never publish only the average. Retain the individual traces or at least a span underlay. A smooth average can hide a severe dip, peak, or level imbalance at one important position.
If every position has a different deep null, the average may look smooth while no listener actually hears the displayed curve.
Suggested Position Sets
Driver head region
Use a documented centre point plus small left, right, forward, rearward, and vertical offsets that represent realistic head movement.
Two-seat compromise
Use matched position patterns around driver and front passenger, then assign explicit weights based on the project goal.
Coverage investigation
Measure a grid or arc rather than averaging immediately. The pattern reveals lobing and directivity that one average would hide.
When Not To Average
- when aligning impulse arrival or relative phase without stable timing
- when diagnosing one exact microphone or competition sensor position
- when a worst-case seat must meet a defined limit
- when the individual traces use different processing, smoothing, windows, or levels
Measurement Record
- coordinate or description of every microphone position
- weight assigned to each position
- level-alignment range and target, if used
- RMS, dB, vector, or moving-microphone averaging method
- timing-reference and alignment method for vector averages
- individual traces, average, and span or worst-case positions