5 Tips for Optimizing Audio Tests for Speed and Accuracy

Shorter test sweeps make your tests faster, but longer averaging time improves noise rejection leading to more accurate and repeatable measurements. When optimizing audio tests, how do you adjust your test signal to get fast results without sacrificing accuracy? There are 5 parameters you can tweak.

1. Frequency range: Only sweep over the frequency range that you actually need to. It’s not always essential to sweep from 20Hz to 20kHz. In many cases, 100Hz to 10kHz is sufficient because the device under test has significantly rolled off above and below these frequencies. This can offer considerable time savings especially at low frequencies where the wavelengths and averaging times are long and background noise is high.
2. Resolution: Decreasing the resolution, for example from R40 (1/12th Oct.) to R10 (1/3rd Oct.), will make the sweep time shorter. Bear in mind that the trade off is that you will have fewer data points so might miss some audio artefacts that would be detected at higher resolution.  Often you can use lower resolution at lower frequencies where the measurement curve is smooth, and higher resolution at higher frequencies where the measurement curve is more variable. This way you get the best of both worlds, faster at low frequencies and more resolution at high frequencies with little time trade off. 
3. Number of Cycles: Likewise, decreasing the number of cycles at each step will shorten the stimulus length but reduce accuracy.
4. SNR: A better signal-to-noise ratio improves measurement accuracy so consider increasing your stimulus level. You can also increase signal to noise on an acquisition step by increasing the input gain on your audio interface. Use the Max FSD function in SoundCheck to ensure that you are operating at maximum dynamic range.
5. Background noise: reducing background noise increases your measurement confidence, Consider soundproofing your test station to improve your test accuracy and repeatability.

SoundCheck’s HarmonicTrak analysis algorithm includes confidence limits, standard error, and noise tools to help you balance speed versus accuracy. These features evaluate the effect of any changes in stimulus on both test time and the measurement confidence so that you can optimize your test time while maintaining confidence that your accuracy is still within acceptable limits.

To learn more, check out our user manual or this short demonstration video.