Measurement Chain

Canonical Definition

A measurement chain is the complete set of signal-generation, device-under-test, sensing, conversion, calibration, timing-reference, and analysis stages that produce a reported result. The result is limited by every stage in that chain.

The microphone is only one link.

The test file, player, operating system, interface, source processing, DSP, amplifier, speaker, cabin, microphone, input converter, calibration, timing reference, and software settings can all change the result.

Full Signal Path

Test signal or program file
→ playback software and operating system
→ output device or source unit
→ factory integration / DSP / amplifier
→ loudspeaker or electronic device under test
→ acoustic environment or electrical load
→ microphone, probe, or sensor
→ preamp / interface / ADC
→ calibration and timing reference
→ analysis software and graph settings

A hidden limiter, loudness control, resampler, channel mixer, noise-cancellation system, automatic gain control, microphone input boost, or software correction can invalidate an otherwise careful test.

Five Questions Before Every Measurement

What is the stimulus? Record the file, generator, frequency range, level, channel, sample rate, and duration.
What is the device under test? Define the exact source, DSP preset, amplifier channel, driver, crossover, load, and vehicle state.
What is sensing the result? Record microphone or probe, orientation, calibration file, interface, gain, and input headroom.
What is the reference? State whether the result is relative, absolute SPL, loopback-referenced, acoustically timed, or unreferenced.
How is it displayed? Record software version, smoothing, windowing, averaging, axes, units, and detector.

Relative And Absolute Results

Result	What it can support	What it cannot claim alone
Repeatable uncalibrated trace	Before/after shape and level changes in the same setup	Absolute SPL or standards compliance
Mic frequency-response calibration	Correction of the microphone's stated response	Calibration of source, interface, amplifier, or acoustic field
Sensitivity or SPL calibration	Absolute sound-pressure reference for the unchanged chain	Automatic frequency-response or class compliance
Interface loopback calibration	Correction or documentation of the electrical input/output path	Correction of microphone, speaker, or cabin response

Headroom And Noise At Both Ends

The stimulus must be strong enough to sit above ambient noise and the measurement system's noise floor, but low enough that no source, DSP, amplifier, microphone, preamp, or ADC clips. A clean input meter does not prove the upstream source was clean, and a clean source does not prove the microphone input remained clean.

Repeat the baseline. If two captures made without changing the system do not agree closely enough for the decision you are making, fix repeatability before tuning.

Timing Reference

Magnitude-only comparisons can often be useful without absolute timing. Phase, impulse, delay, vector averaging, and trace arithmetic require a repeatable timing reference. Use a wired loopback, acoustic timing reference, or another documented method and keep it unchanged for every compared trace.

Minimum Measurement Record

date, vehicle state, doors/windows, seat position, HVAC, engine, and ambient conditions
source, mode, channel, file or generator settings, sample rate, and processing
DSP preset, routing, level, polarity, delay, crossover, EQ, limiter, and mute state
microphone or probe, serial number, orientation, calibration file, interface, gain, and sample rate
microphone position, timing reference, sweep level, input headroom, smoothing, windowing, and averaging
trace name, purpose, and the one variable changed from the baseline