Automotive Audio Testing using AES TC-AA guidelines

Introduction

This technote discusses the AES White Paper on In-car Acoustic Measurement, proposing measurement procedures for evaluating car audio system performance. It details how to follow these recommendations using GRAS microphones and AP analyzers, and explains the associated APx project files.

In-car acoustic field and microphone impact

The increasing prevalence of audio systems in vehicles has made in-car acoustics a significant area of study. Unlike traditional sound fields (free, pressure, or diffuse), the in-car acoustic environment is complex due to multiple sound sources and the mix of reflective and absorptive materials within the cabin. This necessitates careful consideration for accurate acoustic measurements.

Free, pressure, and diffuse field

A free field is an unobstructed region where sound propagates freely. In such a field, a microphone measures only the direct sound from the source.

A pressure field is characterized by uniform phase and magnitude across a surface or within a small, closed chamber, such as a wind tunnel or acoustic coupler.

A diffuse field, or random-incidence field, is one where sound arrives from all directions with equal probability and random phase. This can occur in a room with many reflective surfaces.

In-car acoustic field

The in-car acoustic field is more complex than the traditional fields. It is influenced by multiple speaker locations and a mix of absorbing and reflective materials, making it fall between a free and diffuse field. The specific characteristics vary significantly between vehicles.

Figure 1: The car cabin is a complex and non-ideal sound field made of a mix of highly reflecting materials like glass, and other absorbent materials like the ones used for the seats.

What is a measurement microphone?

Measurement microphones, defined by IEC 61094, are crucial for accurate acoustic measurements. Their specifications, including sensitivity, frequency response, and dynamic range, are critical. Using standardized microphones ensures traceable sensitivity and frequency response, leading to reliable and repeatable measurements. Microphones are designed for specific sound fields (free field, pressure field, random incidence) and come in various sizes (1", 1/2", 1/4").

Microphone influence on the sound field

A microphone's size significantly impacts measurements by interfering with the sound field. Frequency (f) and wavelength (λ) are inversely proportional (f = c/λ, where c is the speed of sound). When the wavelength is comparable to the microphone's size, interference and diffusion occur, causing pressure build-up in front of the diaphragm. Therefore, microphone selection must consider the frequency range of interest.

Figure 2: Pressure build-up when the size of the microphone is of the same order as the wavelength.

Microphone size also affects the dynamic range. Larger diaphragms generally offer lower noise floors, enabling the measurement of very low sound pressure levels. Smaller microphones can handle higher sound pressure levels. Most measurement microphones offer a dynamic range of around 120dB.

Figure 3: Typical dynamic range difference for measurement microphones of different sizes.

Types of microphones for different sound fields

Measurement microphones are designed for specific sound fields: free field, pressure field, and random incidence. A free-field microphone, for instance, has a flat frequency response when pointed directly at a sound source. Using it in a diffuse or pressure field will result in deviations from this flat response.

For non-ideal sound fields, multifield microphones are designed to provide a flatter frequency response. Figure 4 compares the frequency response of three GRAS microphones: the 1/4" multifield 46BC, the 1/4" pressure microphone 46BL-1, and the 1/4" free-field microphone 46BE. The comparison highlights how microphone type affects response in different sound fields.

Figure 4: Comparison of free-field, random, and pressure response for GRAS 46BC, 46BL-1, and 46BE.

In-vehicle Standards

The Technical Committee in Automotive Audio (TC-AA) identified a lack of consensus in automotive audio measurement standards. To address this, a group from TC-AA proposed recommendations for measurement procedures in a white paper. These recommendations cover measurement selection, test setup, test signals, and microphone placement.

The TC-AA recommends three key measurements for in-car acoustics: frequency response, max SPL (Sound Pressure Level), and impulsive distortion. Frequency response assesses spectral balance, Max SPL measures the entire audio system's capability, and impulsive distortion identifies issues like squeak, rattle, and extraneous noise.

For measurements, it is advised to restore car audio system settings to default. Using six microphones for power averaging provides a better representation of the sound field than single-point measurements. 1/4" high-sensitivity pressure microphones (like the 46BL-1) are recommended for their ability to measure high SPL and their minimal impact on measurements below 40 kHz. Their high sensitivity also allows for lower SPL measurements compared to standard microphones.

The microphone array should be positioned in an "H-shaped" configuration, with dimensions optimized to represent typical driver head positions (5th, 50th, and 95th percentiles).

Figure 5: Pictures of the microphone array and its dimensions, from the AES White Paper.

Figure 6: Figure explaining how to place the microphone array, from AES White Paper.

Stimulus and Listening Levels

Test signals, provided as .wav files, include pink noise for frequency response and max SPL (-12dBFS from 20Hz to 20kHz), and a logarithmic sinusoidal chirp for impulsive distortion. These signals can be played via Bluetooth, CD, memory stick, or auxiliary line-in.

Measurements can be performed at various listening levels, with a reference of 80 dBA. It is recommended to start at 50 dBA and increase in 10 dBA steps until the audio system's maximum capacity is reached.

Equipment for in-car testing

This section details the equipment needed for in-car testing, focusing on GRAS microphones and Audio Precision analyzers.

The microphones

GRAS offers two suitable microphones: the 46BL-1, a high-sensitivity 1/4" pressure microphone with a low noise floor, and the 46BC, a multifield microphone ideal for in-vehicle acoustics. Both have a small footprint and high sensitivity comparable to 1/2" microphones.

The GRAS PR0004 AutoArray

To facilitate the precise placement of six microphones as per AES recommendations, GRAS offers the PR0004 AutoArray AES Configuration. This system includes a 6-microphone array, a seat mount, and an adjustable-angle adapter for easy and accurate positioning. Components like a pole, ball joint, magnetic surface for an inclinometer, and ball level attachment ensure precise placement in terms of position, height, and pitch angle. Signals from the six microphones are channeled through a 7-pin LEMO connector, simplifying setup and reducing cable clutter.

Figure 7: Maximum and minimum responses of the GRAS 46BC and 46BL-1 in a non-ideal sound field.

Figure 8: The PR0004 AutoArray AES configuration.

GRAS also offers the PR0003 AutoArray Cross Configuration for different testing methodologies.

The power modules and the analyzer

Various analyzer setups are suggested:

• Multi-channel analyzers (e.g., APx585, APx586): These feature eight analog input channels for simultaneous acquisition. Since they lack built-in power supply, two GRAS 12BB 4ch CCP power modules are needed. This setup is recommended for its ease of use and speed.
• 2-channel analyzers (e.g., APx515, APx516, APx52x, APx555): A switcher, like the AP SWR 2755B-U, is recommended to connect the six microphones. Alternatively, manual plugging/unplugging can be done, but it increases the risk of errors. For APx515, APx516, APx52x, or APx555, a 2ch CCP power module (GRAS 12BE) is needed between the analyzer and the switcher.
• APx517 analyzer: This analyzer has a built-in CCP supply, so no separate power module is needed.
• APx500 Flex Key with ASIO devices: This can be used with or without a CCP Power Module.

Figure 9: Setup with APx58x.

Figure 10: Setup with APx555.

Figure 11: Setup with APx517.

Figure 12: APx Flex and ASIO Devices.

These setups are recommendations; other GRAS or AP equipment can be used if they comply with AES recommendations.

APx Project File

This application note includes four project files corresponding to the described setups. These files are adaptable and meant to be modified based on specific equipment and desired results. The project files can be unlocked by clicking 'Unlock' and entering the password '123'.

Project File Structure

The project files are designed for open-loop measurements and are used in sequence mode. They guide the operator through the process, indicating which test signals to play via Bluetooth, CarPlay, Android Auto, or AUX input.

Project files for two-channel analyzers have two sequences: 'AES In-Car Acoustic Measurements' and 'AES In-Car Acoustic Measurements with Switcher'.

Project files for APx585 or 586 have three sequences: 'Mic Calibration (TEDS)', 'Mic Calibration (External Calibrator)', and 'AES In-car Acoustic Measurements'.

Figure 13: Prompt for 2-channel analyzers (left) and for APx58x (right).

Mic Calibration (TEDS) Sequence

TEDS (Transducer Electronic Data Sheet) stores microphone sensitivity information on a chip. The TEDS calibration sequence reads this data to set the microphone sensitivity in the Signal Path Setup.

Mic Calibration (External Calibration) Sequence

This sequence prompts the operator to calibrate each microphone using a sound calibrator (e.g., 42AG) set to 114dB at 250Hz.

Figure 14: Prompt when using external calibration.

For two-channel analyzers, TEDS calibration is the preferred method due to the time required for external calibration when only two microphones can operate simultaneously.

AES In-Car Acoustic Measurements Sequence

The AES In-Car Acoustic Measurements sequence involves three tests: Frequency response, Impulsive Distortion, and Max SPL. Measurements can be performed at various levels from 50 dBA to 100 dBA.

Measurement Steps

• Noise (RMS) 80 dBA: Measures RMS noise level and computes power average. The operator plays the 'PinkNoise 20-20000Hz' signal and adjusts the car audio system to 80 dBA.
• LEQA 80 dBA: Measures RMS noise level and computes power average over 30 seconds using A-weighting and filters.
• Response 80 dBA: Measures and power averages the frequency response over 30 seconds using the pink noise signal.
• Impulsive Distortion 80 dBA: Uses a logarithmic sinusoidal chirp signal to measure impulsive distortion metrics like HOHD, peak ratio, and Rub and Buzz Crest Factor.

Figure 15: Noise (RMS) 80 dBA result.

Figure 16: Octave spectrum for the six microphones (left) and power averaged (right).

Figure 17: The ID HOHD (left) and ID Peak ratio (right) plotted against the RMS Level.

MAX SPL Measurements

The sequence also includes measurements for maximum sound pressure level:

• MAX SPL: Measures the RMS level after increasing the audio system to its maximum. Filters at 20 Hz and 20 kHz, and a C weight are applied before power averaging.
• MAX SPL Spectrum: Measures the frequency response at maximum SPL, similar to 'Response 80 dBA'.

Figure 18: Octave spectrum for the six microphones (left) and power averaged (right) at max SPL.

The project files generate a detailed report that can be saved in various formats (PDF, Text File, Excel, HTML).

Figure 19: Example of the first page of a Sequence Report.

Open-loop vs. Closed-loop

The sequences are designed for open-loop measurements but can be modified for closed-loop measurements, for example, by connecting the analyzer directly to the vehicle's audio system via Bluetooth or a balanced cable.

Note for two-channel analyzers

For two-channel analyzers, measurements can be performed with or without a switcher. Using a switcher automates connection changes, while manual connections require following on-screen prompts.

Figure 20: Example of a prompt indicating the connections to be done.

Figure 21: The sequence steps when using a switcher.

Note for ASIO devices

Project files for ASIO devices, such as RME interfaces, can serve as a base for building custom project files. These files can be unlocked with the password '123' and modified.

Conclusion

The AES white paper on in-car acoustic measurement is a foundational step towards establishing industry consensus. The TC-AA group's recommendations for frequency response, impulsive distortion, and max SPL measurements, along with specific microphone placement and test signals, provide a framework for consistent testing.

This technote presented four setups using various analyzers and microphones (GRAS 46BL-1 and 46BC), along with associated APx project files. While setups and project files are suggestions, multichannel solutions are recommended for efficiency and accuracy. Two-channel project files are also available for users with existing APx analyzers.

AES continues to provide recommendations that may evolve as car manufacturers adopt them, aiming to establish a formal AES standard. The combination of GRAS PR0004 AutoArray, GRAS microphones, and Audio Precision analyzers offers a comprehensive solution for accurate automotive audio system testing, ensuring consistency, precision, and repeatability throughout the development process.

References

1. AES Technical Committee Automotive Audio
2. How to Match a Measurement Microphone to a Sound Field
3. GRAS Microphone Guide
4. GRAS 46BC - Product Information
5. GRAS 46BL-1 - Product Information
6. GRAS 46BE - Product Information
7. AES White Paper In-Car Acoustics Measurements
8. Dynamic Range of a microphone – GRAS Article
9. PR0004 AutoArray AES Configuration - Product Information
10. APx58x Analyzers - Product Information
11. GRAS 12BA, BB, BE - Product Information
12. SWR 2755B Switcher - Product Information
13. AP Analyzers - Product Information
14. APx517 Analyzers - Product Information
15. APx500 Flex - Product Information
16. What is TEDS - GRAS Blog
17. GRAS 42AG - Product Information
18. AP Application Note Loudspeaker Rub and Buzz Measurements