Page 13 - Summer2017

of the smartphone. Although recent smartphone models often incorporate two or more microphones for noise re- duction and/or beamforming for the purpose of enhanc- ing speech signals, it is common to find a primary, omni- directional microphone that may be most useful for typical acoustical measurements, such as the overall sound level.
The directionality of the microphone is an important param- eter to consider when assessing the suitability of a particu- lar smartphone as a measurement device. Dedicated sound level meters (SLMs) employ measurement microphones that are designed to exhibit a directional sensitivity that is as omnidirectional as possible. A microphone that exhibits a cardioid, supercardioid, or some other nonuniform direc- tional pattern will be of limited value for overall sound level measurements because of its decreased sensitivity to sounds coming from certain directions. It is important to note that the shape of the smartphone body will itself have some im- pact on the directional behavior of its embedded omnidirec- tional microphone(s), especially at higher frequencies.
Beyond the built-in microphone, there are several other mechanisms available for getting signals into a typical smart- phone. These include headset microphone input, Bluetooth and Wi-Fi wireless communications protocols, and electri- cal ports that can support either standardized (e.g., USB Au- dio Class driver) or proprietary protocols for data transmis- sion. The audio signal path is currently best supported by standard protocols for Bluetooth and USB, but this does not preclude the existence of, or potential for, other proprietary solutions for acquiring acoustic signals with a smartphone.
The simplest mechanism for connecting an external mea- surement microphone to a smartphone is via the common headset jack. Most smartphones include a headset jack, which, in addition to serving as a headphone jack for au- dio output, supports a single microphone input. A typical headset microphone may have similar characteristics to the built-in microphone of the smartphone, but the headset jack makes it possible to connect a higher quality measure- ment microphone without the need for additional adapters or power sources. This means that using a microphone con- nected to the headset jack represents the most affordable and portable means for replacing the built-in microphone as the primary input source for higher quality measurements.
Getting Signals Out of the Device
For certain kinds of measurements, such as measuring the frequency response of a sound reinforcement system or the
impulse response of a listening room, getting signals out of the smartphone can be just as important as the inputs. Smartphones typically have very small built-in loudspeakers with a limited frequency range or limited power output, but they also have a headphone jack through which signals can be transmitted to a power amplification system or directly to a device under test (DUT). In addition to built-in ana- log outputs, smartphones offer the same alternative signal paths for output as for input. That is, signals can potentially be transmitted over Bluetooth, Wi-Fi, USB, or some other proprietary interface.
Hardware Considerations
In addition to the directional behavior of the microphone, as mentioned previously, the frequency response, dynamic range, sensitivity, and other characteristics of the various hardware components within the signal path can significant- ly impact the accuracy and/or precision of a measurement. The presence of automatic gain control (AGC), for example, which is commonly used to optimize the acquisition of speech signals for telephony, may significantly diminish the accuracy of an overall sound level measurement, even though it may not adversely affect a measurement designed solely to identify specific frequency components of an acoustic signal.
Operating System Considerations
Once a signal of interest passes through the relevant hard- ware components into the digital domain, the OS of the smartphone takes ownership. Again, this new step in the sig- nal path can potentially affect the viability of a measurement. One way for OS software to directly affect the potential for quality measurements is to offer mechanisms for control- ling certain behaviors of the hardware components within the signal path. For example, if a hardware component, such as the analog-to-digital converter (ADC), employs its own form of AGC, then it may be possible for the OS to provide an application programming interface (API) to third-party apps to allow them to disable it for the sake of measurement accuracy. As another example, the OS may provide an API for adjusting the analog gain of the microphone input signal prior to digitization.
OS-level signal processing may also affect signal integrity. When third-party application software (an app) requests an input signal at a sample rate that is different from the hard- ware sample rate, the OS may resort to sample rate conver- sion, which requires some form of filtering that may have an impact on the frequency content of the signal. OS software
Summer 2017 | Acoustics Today | 11