Complete Audio Test Solutions for a New Age
Testing audio where there´s no direct access to the microphone or speaker presents new challenges to R&D and production test engineers alike.
For the traditional loudspeaker or microphone we can make our basic tests of frequency response and distortion etc, by connecting to the transducers themselves (or at least their analogue interfaces). We term this system ´closed loop´.
However, with audio devices such as Bluetooth earbuds, the signal path for the test stimulus or microphone input will be digital. Or, more head scratching, would be attempting to test the frequency response of a smart device such as an Amazon Echo speaker. Here the test stimuli have to be streamed from the Cloud. And that´s not to mention testing the MEMs microphone array embedded in the product. Apart from the practicalities of connections and stimuli-response access , the input and output signals will be asynchronous. That is to say originating from different clock sources, with potentially different sampling rates. Therefore frequency, phase and delay differences have to be handled before any analysis can take place.
Fortunately AcSoft´s partner, Listen have understood and been solving these problems for many years. As a consequence, they are actually making these otherwise complex tests not only achievable but also fast and accurate, for both R&D and manufacturing environments.
Listen´s test software, Soundcheck, already incorporates the necessary algorithms for integration into test sequences that solve these issues. An example is automatically correcting for clock differences by re-sampling and frequency shifting. Whilst for phase and delay, correlation techniques are employed to correctly align these otherwise asynchronous stimuli and responses.
Furthermore, Listen offers a choice of compatible hardware interfaces (audio A/D, MEMs, Bluetooth etc..), whilst Soundcheck´s simple but comprehensive signal path management, makes sense of the calibration in the exchange of data between the analogue and/or digital domains. Put together this all means that any traditional analysis from the closed loop method is achievable in these asynchronous "open loop" testing modes. Read more at Listen on their open loop solutions here.
HOWEVER, to design and develop the best performing products, we need to test them with the help of the real thing - humans. Well almost - actually simulated body parts in this case. To achieve this AcSoft partners the Listen solutions with GRAS head, ear and voice simulators for complete audio test systems.
What´s so special about ear simulators and why?
The performance of devices such as headphones or earbuds can only be sensibly assessed if tested in-situ, i.e. on a head or in an ear. So we use simulators (rather than the real thing). A measurement microphone is integrated into a mechanical construction that presents the same "acoustic load" as would the human ear to the audio device. We know too well how the performance of headphones change when lifted from our head or an earbud is not nicely seated in our ear. Those are the effects of the acoustic load changing.
And of course if it´s a closed headphone we´re testing, we definitely want a cheek. Never mind a pinna and an ear canal. More about GRAS ear simulators.
Likewise, whether testing a headset microphone or the voice control of a smart speaker, this will require a test loudspeaker as sound source. This should have the same characteristics as that of the human mouth, such as this simulator from GRAS, pictured. This mimics the dispersion of its less cyber-esque counterpart. And as for the ear simulator, the mouth simulator is told how to behave by international standards!
Meaningful Audio Testing
Key to meaningful testing is accuracy, repeatibility AND adapting to changing market expectations! That is achievable because GRAS simulator designs not only conform to international standards, but have been designed from the ground-up for more realistic performance and to meet real-world demands. GRAS innovations include developing an unnervingly realistic ear with of a choice of material grades for softness to suit - the ´anthropometric´ear. In practical terms, it increases the measurement realism and repeatibility.
A more recent example is the introduction by GRAS of high frequency ear simulators. These exceed the requirements of existing international standards, to measure up to and beyond 20kHz. These all support the engineers in their ongoing quest to improve, optimise and test the audio quality of our devices, in both R&D and manufacturing scenarios.
A variety of testing platforms are available. In addition to the complete head and torso above, there are headphone only and ´table top´ single ear and cheek text fixtures. There are voice simulators and even compact production line anechoic chambers for testing smaller devices such as mobiles and tablets.
The Perfect Partnership
In summary, audio testing has required new solutions for an IOT world of devices that increasingly incorporates audio with a signal path that wanders through the Cloud. Fortunately AcSoft partner the advanced electro-acoustics test and analysis solutions from Listen, with the perfectly pèrforming body parts of GRAS to provide the complete audio test systems for this new age. Please feel free to call us to discuss your applications - we´re sure we can help.