Open Loop Test Measurement

How do you measure devices that can’t be directly connected to an audio interface? In this video, we demonstrate how to measure open loop devices such as smartphones and smart speakers using SoundCheck. You’ll learn how to generate a stimulus file, play it back from the device under test, and use triggered recording with a microphone and audio interface to capture the response for analysis. This flexible technique allows accurate measurement of frequency response, distortion, THD, and Rub & Buzz – even when playback and recording are asynchronous.

Watch this Video Demo of Open Loop Testing

Learn More about Open Loop

• Open Loop Microphone Test: https://www.listeninc.com/open-loop-microphone-test-updatedtest/
• Open Loop Measurements of Smart Devices: https://www.listeninc.com/trigger-options-for-open-loop-testing/
• Trigger Options for Open Loop Testing: https://www.listeninc.com/trigger-options-for-open-loop-testing/

Video Transcript: How to Measure Open Loop Devices

Devices like smartphones and smart speakers can be challenging to test as they can’t be directly connected to conventional interfaces to make audio measurements. This means playback and recording are asynchronous.

To make accurate measurements, we use a technique called open loop testing. This lets us manually load the stimulus signal onto the device and play it using a  triggered record. This means SoundCheck starts recording the response once it hears a certain sound. This might be a level threshold, or a specific frequency, or even a chirp. Once it hears that trigger it will begin the acquisition, record the stimulus, and then analyze it as normal. 

Let’s take a look. I’m going to use a standard test sequence that comes with SoundCheck, but first let’s look at our test configuration.

We’re using SoundCheck to create the stimulus file, and we’ll load that manually onto our device – you can send it by email or download via the cloud, or even store it in the device firmware. We’ll then run the sequence, and play the test signal from the phone. When the trigger tone is detected, the software will start recording using an SCM microphone which is connected to the software via an AudioConnect 2 audio interface, and analyze the response.

Let’s go ahead and run the test.

The sequence is giving me the option to create the stimulus files ready to load onto the device, but I’ve already done that so I’ll just decline and let the sequence continue….

Now it’s listening for the trigger. I’ll play the sound…. And you can see, it went ahead and captured the response – we have our recorded time waveform here. Over here we have our fundamental, and also the second and third harmonics. Down here we have THD and Rub & Buzz, and over here we have perceptual Rub & Buzz. You can see we’ve got some distortion across the spectrum, and some Rub & Buzz at low frequencies – that’s not really a surprise – phone speakers tend to be small and inexpensive so these issues are common. Once you have the waveform captured you can run any analysis or calculation on it, so if you wanted, you could add other measurements to this test.

Now that we’ve seen it in action, let’s go back to the sequence and learn a bit more about triggered recording. We’ll start with the stimulus.

As you see, we still have a standard stimulus step from XHz to YkHz,  but when we look at it closer we can see that it has two sections. First we have a frequency log chirp that tells SoundCheck when to start recording. This is much more robust than simple level or tone triggers, especially if you’re in a production environment where you might have background noise or even pick up test tones from other lines. And you can see here we’ve set it so that this portion of the stimulus does not get analyzed. 

The second portion of the stimulus is our test signal – a standard frequency stepped sweep from 10kHz down to 100Hz and we have ‘analyze’ set to ‘yes’ for that one.

The acquisition step must also be configured for triggered record. Here, we make sure the ‘triggered record’ checkbox is checked, and then select the type of trigger that we’re using. We  used ‘level and cross-correlation’, which looks for the specific chirp that we used – that one is the most robust but we also offer simpler level and frequency triggers.

So, as you can see, open loop measurements are simple with SoundCheck. This method can be applied to any device that doesn’t have a direct connection like smart speakers, smart glasses, phones or even robots. Check out our website to learn more.