Open Loop Microphone Test (V4)
A test sequence for measuring a microphone on a device such as a smartphone or tablet where there is no direct connection between the microphone and the computer.
This sequence demonstrates the two most common microphone measurements, frequency response and sensitivity, on a microphone embedded in a recording device. Typically, when measuring a microphone the response of the device can be captured simultaneously with the stimulus. However, with devices such as voice recorders and wireless telephone forming a closed loop can be cumbersome or impossible. This sequence demonstrates how to measure such a device by recording the signal on the device under test, transferring that recording to the computer running SoundCheck and then using a Recall step to import the recorded waveform and analyze it.
This specific sequence, v4, is an improvement on the prior versions. The v1 release required that the audio file containing the recorded response waveform be manually windowed outside of SoundCheck before being analyzed. The v2 release utilized a new feature in SoundCheck 14, using values from the memory list to semi-automatically trim the waveform before analysis. The v3 release completely automated waveform editing through the use of an intersection level and windowing post processing steps. Currently the v4 release uses the new Auto Delay+ algorithm, exclusive to SC18 and beyond. Auto Delay+ is capable of detecting and accounting for delays of -0.5 seconds to any positive delay, nullifying the need for windowing steps in the sequence. If you are interested in learning more about this algorithm please refer to the Analysis section of the SoundCheck manual.
The sequence begins by the operator being prompted to begin recording on the device under test and then a stepped since sweep is generated from 10kHz through a source that has been previously calibrated to produce 1 Pascal across the sweep range. Next, the operator is prompted to transfer the recording to the computer and then load the file in to SoundCheck via a Recall step.
From here the recorded signal is sample rate converted and frequency shifted to match the stimulus and then analyzed with a HarmonicTrak analysis step, which calculates the response curve. A post processing step is used to extract the level at 1 kHz, yielding the sensitivity value. Limits are set around both the frequency response and the sensitivity. The default limits values should be adapted to your particular device.
The final display shows two graphs. The top X-Y graph displays the data at its absolute level in dBFS/Pa (since the imported recording is digital, the results will be in FS or dBFS rather than Volts or dBV). The lower graph shows the windowed recorded signal analyzed by the software. In addition, the calculated sensitivity at 1kHz and the frequency response margin is also displayed.