Generic Microphone Calibration¶

This calibrates a "generic" microphone — one that doesn't need to be a precision measurement mic (e.g., an inexpensive electret used just to record verification tones in a booth) — by playing a broadband stimulus through a speaker and comparing what it records against a co-located, already-calibrated measurement microphone.

What you'll need¶

A calibrated measurement microphone (see Measurement Microphone Calibration) and its preamp, positioned alongside the generic microphone.
The generic microphone being calibrated.
A speaker to play the stimulus.
An analog to digital converter and a digital to analog converter.

Opening the workspace¶

Launch CFTSCal and select the Generic Microphone Calibration workspace.

Settings¶

The Settings panel is laid out as a small grid: one row per device (Test, Ref., Speaker), with a Channel column and a Device column.

Field	What it means
Test → Channel	Which input channel the generic microphone under test is wired to.
Test → Device	A free-form label for identifying the generic microphone you're calibrating (e.g., product ID, serial number, asset tag, etc.). Click + to add a new one to the drop-down list, or - to remove the selected one from the list (no calibrations are deleted). The drop-down after it is the generic microphone's preamp gain, in dB.
Ref. → Channel	Which input channel the reference measurement microphone is wired to.
Ref. → Device	Which calibrated measurement microphone to use as the reference, and its preamp gain. Its calibration must be set as current (right-click > Set as current in Measurement Microphone Calibration); a new reference microphone can't be added from here.
Speaker → Channel	Which output the speaker playing the stimulus is wired to.
Target folder	Which folder the new calibration is saved in. Leave it at (auto create from test device) to have a folder named after the Test → Device label. To create a new folder, right-click the Calibrations list. Each test channel remembers its own target folder.

Running the calibration¶

Click Golay or Chirp once a test device, a reference microphone and a speaker output are selected. The buttons don't check this for you, so make sure each is set first.

Golay plays a pair of complementary Golay-code sequences, several times each, and cross-correlates the recorded response against them. More robust to background noise, at the cost of taking longer.
Chirp plays a single frequency sweep. Much faster, but somewhat more sensitive to noise.

Both runs expose a Smoothing window parameter (default 10) — the width, in frequency bins, of a Hamming-weighted moving average applied to the computed sensitivity curve, which trades fine frequency detail for a less noisy curve. Set it to 0 to see the raw measurement. See Speaker Calibration for the other run parameters, which are the same.

How the calibration is computed¶

Both microphones record the same stimulus at the same time. The reference microphone's known sensitivity \(S_{cal}\) tells you the sound pressure that was actually present, so whatever the generic microphone did differently must be a property of the generic microphone:

\[ S_{exp}(f) = \frac{V_{exp}(f)}{O(f)} = \frac{V_{exp}(f) \times S_{cal}}{V_{cal}(f)} \]

Writing a quantity as a function of \(f\) (frequency, in Hz) just means it takes a different value at each frequency — which is exactly the point for a generic microphone, whose sensitivity is a curve rather than the single number \(S_{cal}\).

Or, in the dB form cftscal actually uses:

\[ S_{exp_{dB}}(f) = 20 \times log_{10}(V_{exp}) + 20 \times log_{10}(S_{cal}) - 20 \times log_{10}(V_{cal}) \]

Unlike a measurement microphone, the result is a curve rather than a single number, because a non-precision microphone's sensitivity varies substantially with frequency — that's the whole reason this workspace exists. See Calibration Math for the derivation.

The arithmetic can't distinguish the microphone from its position

The equation above attributes every difference between the two recordings to the generic microphone. If the two microphones aren't at the same distance and angle from the speaker, that positioning difference is silently folded into \(S_{exp}\) as though it were the microphone's own frequency response.

Reviewing the results¶

Generic Microphone Sensitivity plots the frequency response (in dB re 1 V_rms) of every calibration currently ticked in the list below.

Calibrations (the list) shows every calibration ever run for this workspace, with the following columns:

Column	Meaning
Name	Which generic microphone was calibrated (organizes the list; see Device below if you've filed calibrations into folders that don't match the device).
Date	When the calibration was run.
Device	The device label recorded at calibration time, independent of which folder the calibration is filed under. Usually matches Name — compare the two if you've reorganized calibrations into folders.
Input	Which input channel the generic microphone was wired to.
Gain	The generic microphone's preamp gain, in dB, that was in effect.
Microphone	Which reference measurement microphone was used.
Mic. Channel	Which input channel the reference microphone was wired to.
Speaker	Which speaker played the stimulus.
Speaker Channel	Which output channel the speaker was wired to.
Max. Freq.	The highest frequency the calibration covers.

The list works the same way as in Measurement Microphone Calibration: tick to plot; right-click for Edit note…, Export as WAV…, Set as current and Delete; double-click to rename; drag onto a folder to move. A sticky note after a calibration's name means it has a note and a warning sign means its files have a problem — hover over the row to read either. Set the calibration you want experiments to use as current (★); nothing falls back to the newest one automatically.

Sanity-checking a calibration¶

Does the response look like previous calibrations of the same device? A sudden change usually means the microphone moved relative to the speaker, or is damaged.
Are the two microphones actually co-located? If the generic and reference mics are at noticeably different distances or angles from the speaker, the computed sensitivity will reflect that mismatch, not the generic mic's real response.
Is Max. Freq. as high as you need? It's limited by both the speaker and the reference microphone's usable bandwidth.

Troubleshooting¶

Common pitfalls

Reusing a device label for a different physical microphone breaks your ability to track a specific unit over time — give each physical mic its own label.
Wrong reference microphone selected, or that microphone's own calibration is stale, propagates straight into the generic mic's measured response.
Microphones not close together is a common, easy-to-miss source of a distorted-looking response.