Listen President, Steve Temme, will be presenting a paper entitled ‘Enhanced perceptual rub & buzz measurement for loudspeaker production test’. Due to Covid-19 regulations, this paper will be presented by video.
Loudspeaker Rub & Buzz faults are a problem for automotive manufacturers as they sound harsh and immediately give the perception of poor quality. There are two places such faults can occur – during speaker manufacturing and installation of the speaker in the car. A buzzing loudspeaker in a car is disappointing to a customer and is costly to replace. It is also challenging for a service center to determine exactly where the buzzing is coming from and whether it is caused by a faulty loudspeaker or bad installation. Perceptual distortion measurements are often considered the holy grail of end-of-line testing because rejecting speakers with only audible faults increases yield. Although such measurements have been around since 2011, production line adoption has been slow because until now, sensitivity to background noise has made limit-setting challenging. In this paper, a new algorithm is introduced that uses advanced technology to reduce the impact of background noise on the measurement and offer more repeatable results. This facilitates limit setting on the production line and makes it a truly viable production line metric for increasing yield. This same metric may also be used for end-of-line automotive quality control tests. Results from various algorithms will be shown, and their correlation to subjective and other non-perceptual distortion metrics explained.
Author: Steve Temme, Listen, Inc.
Presented at 2022 AES Automotive Conference, Dearborn, MI
The automotive industry’s stringent quality expectations make end-of-line quality testing on automotive speakers and drivers absolutely critical. End-of-line tests typically measure a range of parameters including frequency response, THD, and polarity. Manufacturing-introduced defects such as Rub & Buzz and Loose Particles are also measured. Reliable, automated testing has been available for decades now, and most large manufacturers rely on these software-based systems for identification and rejection of defective products. While these tests do an excellent job of identifying defective units, there is always a certain level of false rejection where units with some distortion fail even though it is completely inaudible to the human ear. From a manufacturing perspective, higher yields and therefore greater profitability is always desirable.
This has driven the development of perceptual distortion measurements – automated measurements that replicate the human hearing to detect only audible distortion defects. Such metrics increase production line yield by passing products with inaudible distortion, as the product will still sound exactly as the manufacturer intended. Perceptual methods are very simple to configure for production line use. Since they return a result in Phons, an absolute measurement that can be easily correlated to the listener’s threshold of hearing, the operator can set a fixed limit across the board, regardless of product. Naturally, the price point and quality expectations for the product may influence the level of distortion that is deemed acceptable.
Our algorithm, introduced in 2011, was the first commercial perceptual distortion metric, although in the past couple of years, other test system manufacturers have also started to offer perceptual distortion tests. It offers excellent correlation with human hearing and performs well in laboratory tests. However, like the human ear, repeatability decreases in the presence of background noise. This is not a failure of the algorithm as such, but an indication that the algorithm performs just like a human listener; when background noise is high, audible distortion is masked. This limitation restricts the value of such algorithms on the production line, as with today’s high-volume manufacturing, there is only time for one fast test sweep. If this sweep gets a different result under changing background noise conditions, limit setting becomes challenging, and repeatability and reliability is decreased. Similar algorithms from other test system manufacturers also suffer from the same problems.
This paper details efforts to create an algorithm that hears like a human in quiet conditions, e.g. in a living room or passenger automotive cabin, under the less-than-perfect conditions of a manufacturing environment where considerable and varying background noise may be present. In other words, a perceptual model that is more independent and reliable than the human ear when it comes to noisy environments. The resulting new algorithm overcomes these limitations to offer repeatable end-of-line test results, even in noisy environments. It incorporates noise reduction techniques and enhanced perceptual filters to overcome the reliability and high frequency masking issues of earlier versions. In short, the algorithm offers the performance of an ‘enhanced’ human ear – it detects distortion like an ear in a quiet environment, even when there is background noise. This makes it a viable solution for production line use.
In this paper we explain how the algorithm works, demonstrate how the results compare with earlier perceptual algorithms and show its correlation with human hearing and conventional distortion algorithms. We also compare its performance in the presence of background noise to other perceptual algorithms by adding recorded factory background noise to the signal before passing it through the algorithms.
More about Listen’s enhanced Perceptual Rub & Buzz algorithm
More about in-car measurement of impulsive distortion / Buzz, Squeak and Rattle.
Steve Temme discusses the importance of detecting manufacturing-induced defects such as Rub & Buzz and Loose Particles during end-of-line testing, and explains the various algorithms that are used. He compares conventional and perceptual metrics for the measurement of Rub & Buzz, including Listen’s new enhanced Perceptual Rub & Buzz algorithm, and discusses why it can be beneficial to use both conventional and perceptual measurements in tandem.
Not all distortion is audible. Perceptual Rub & Buzz measurements identify only those devices with audible Rub & Buzz defects, rather than all Rub & Buzz defects. Basing end-of-line pass/fail decisions on this metric increases yield while maintaining customer satisfaction.
Our full ePRB launch seminar goes in-depth with algorithm explanation, testing demonstrations, use cases, and more.
For all things Enhanced Perceptual Rub & Buzz including listening examples, published articles, and a sequence to try ePRB for yourself, check out our Enhanced Perceptual Rub & Buzz page.
Video Script:
While conventional Rub & Buzz measurements identify faulty products and provide valuable information about the health of your production line, they can sometimes result in devices with inaudible distortion artifacts being rejected. To maximize yields, perceptual distortion metrics enable the rejection of only devices that actually sound bad.
SoundCheck is the only audio test system with a perceptual Rub & Buzz metric that offers excellent correlation with subjective perception and sufficient noise immunity for production line use.
We led the way with research in this area here at Listen. In 2011, we launched the first Rub & Buzz algorithm built on a model of human hearing . This original model was well received due to its excellent correlation with subjective listening tests. However, like the human ear, performance was less repeatable in the presence of background noise. Many hours of original research over the past few years have resulted in our new Enhanced Perceptual Rub & Buzz algorithm. This combines our original methods with proprietary noise suppression and a refined perceptual model to exceed the performance of the human ear. In other words, it replicates the hearing characteristics of a human ear in a quiet environment, even in the presence of background noise.
This means that it offers the excellent noise immunity needed for a wide range of manufacturing environments.
Let me show you what I mean.
These graphs show 3 speakers, a good, bad and borderline loudspeaker, each measured 10 times using 3 different perceptual Rub & Buzz algorithms; our original 2011 version, our new ePRB algorithm, and a perceptual algorithm from another audio measurement company. You can see that while all three indicate the relative magnitude of the distortion, in both our original algorithm and the ‘other’ algorithm, the variation in the repeated measurements for each speaker is inconsistent enough that it would be problematic on the production line. Our new algorithm, the middle graph, clearly shows much greater repeatability between measurements. This offers a high level of confidence in the results and makes it easier to set limits.
Correlation to subjective tests is also improved through the use of more comprehensive masking curves that include additional factors to more accurately replicate the human ear’s behavior and reveal increased detail in the ear’s highly sensitive 500Hz – 2kHz range.
Perceptual distortion metrics are a valuable end-of-line test addition as they increase yield by passing products with inaudible distortion. That said, in most cases, it is desirable to also measure normalized Rub & Buzz and Loose particles, as these are convenient ways of monitoring your production line for early warning of any problems that could eventually lead to a returned product from a customer.
All three distortion measurements, along with a whole host of other end-of-line parameters can be made in SoundCheck simultaneously, using the same stepped sine sweep stimulus signal. In other words, there is no increase in test time when you add perceptual metrics to your end-of-line test. Check out our website for more detailed information, demo sequences to try this out, and more.
In this short video, Devin Vaillancourt gives a brief overview of SoundCheck 20 new features including the new enhanced Perceptual Rub & Buzz, features for multichannel testing including the new multichannel RTA, new features for communications testing, POLQA analysis, signal generator upgrades and more.
This sequence demonstrates how SC20’s new enhanced Perceptual Rub & Buzz algorithm compares to normalized Rub & Buzz and subjective listening. Running this sequence on a batch of good and bad buzzing loudspeakers should help identify by measurement, audibly defective units and where to set production limits. Starting the sequence, the user is asked at what test level to play a stepped sine sweep (Stweep™) in 1/12th octaves, from 20 kHz to 50 Hz. The user is encouraged to try different test levels and change sweep parameters to find the optimum settings to catch buzzing loudspeakers. The loudspeaker is measured via two channels of the audio interface. A calibrated reference microphone is connected to one of the channels and an impedance reference built into the SC Amp or AmpConnect is connected to the other. A HarmonicTrak™ Analysis step analyzes the recorded waveform from the reference microphone, and displays both the enhanced Perceptual Rub & Buzz and normalized Rub & Buzz graphs.
SoundCheck 20, Listen’s flagship audio analysis software, introduces a new state-of-the-art perceptual distortion algorithm for end-of-line testing – Enhanced Perceptual Rub & Buzz. It also includes a host of new features for multichannel and communications testing, increasing its capabilities for audio measurement of smart speakers, infotainment systems, headphones and hearables, and other voice-activated and multichannel devices.
The new enhanced Rub & Buzz algorithm is a reliable and repeatable perceptual method for detecting audible Rub & Buzz faults on the production line. This offers manufactures the option to increase yield by rejecting only those devices with audible distortion defects. Based on the sound psychoacoustic principles used in Listen’s industry-standard 2011 perceptual algorithm, new research offers refinements that improve listener correlation and new, patent-pending, noise reduction technology offers unrivaled repeatability and reliability for end-of-line use. In fact,this algorithm performs like a human ear under normal listening conditions with the background noise of a manufacturing environment! SoundCheck measures perceptual Rub & Buzz using the same stepped sine wave, simultaneously with other end-of-line measurements, enabling this valuable analysis to be added to production line tests with no extension in test time.
Two of SoundCheck’s virtual instruments, the multichannel RTA and the Signal Generator, have been significantly upgraded. The multichannel Real Time Analyzer now includes expanded functionality for real time observation of audio signals. It displays multiple channels simultaneously, and offers real time calculation and display of channel, subtraction, maximum, minimum and power average. Applications for this new feature include real-time observation of active noise cancellation, automotive industry standard 6-microphone array infotainment measurements, validating headphone seal on a test head prior to measurement by comparing left and right earphone response, and more. RTA curves are now seamlessly integrated with SoundCheck’s standard graphs and can be dragged and dropped into any graph for easy comparison to limits and reference curves. This is useful for quick comparisons to reference standards or golden units, tuning automotive infotainment systems, and more.
An overhauled signal generator brings additional functionality and simplified operation, particularly when using wav files. In addition to standard audio stimuli (sine, white noise, pink noise, etc.) with user-selected sampling rates and resolution, it can play any wav file, any memory list file, and any complex waveform created by the stimulus editor. The level can be referenced to Peak, dB or RMS. The signal can be equalized in real time, and custom EQ curves applied, which is useful when using a non-flat source such as a mouth simulator. A specific portion of a waveform can be selected to play, either by selecting a start and stop time in seconds, or by examining the waveform, and signals can play for a fixed duration, a fixed number of times, or in a continuous loop. Outputs from 2 or more signal generators can be mixed on the same channel, and waveforms from multiple signal generators can be synchronized, or delayed for phase control during playback.
The optional POLQA (Perceptual Objective Listening Quality Analysis) module brings Opticom’s POLQA 3 algorithm, implementing the ITU-T P.863 standard, right into SoundCheck, where it can be used for perceptual measurements of speech degradation in communications applications ranging from telephones to smart devices. Used just like any other analysis step in a sequence, this is a fast and cost-effective alternative to a panel of human listeners. In addition to speech degradation measurements, it may also be used to assess the impact of noise reduction algorithms, evaluate bluetooth degradation due to packet loss, or to analyze distortions introduced into the audio path.
Several new features expand the software’s post-processing capabilities for testing communications devices. These include batch processing of external files, a silence stimulus step, and RMS level versus time and histogram post-processing steps. A pre-written test sequence for categorizing Doubletalk performance to two international standards, ITU-T P.502 and ETSI TS 126 132 (part #3114), from communications expert John Bareham, is also available. Other enhancements include a stimulus editor upgrade, the ability to export multiple SoundCheck recorded waveforms to a single multichannel wav file and a hardware reset option for quickly changing audio interfaces or restoring hardware to a known configuration.
Listen continues to lead the way in perceptual distortion measurements with our new Enhanced Perceptual Rub & Buzz algorithm. This new product, based on years of original research, offers unrivaled repeatability, reliability and correlation to human hearing, making it an unbeatable solution for end-of-line perceptual audio test.
Our launch event will include live demonstrations, results, technical explanations and more. Be the first to know. Sign up below.
Listen 凭着新的增强型感知摩擦音算法(Enhanced Perceptual Rub & Buzz algorithm)继续引领著感知失真測量。這種基於多年原始研究的新產品提供了無與倫比的可重複性、可靠性和与人类听觉的相关性，使它成為生產線末端感知音頻測試的無與倫比的解決方案。
This event has passed. Watch the Recording.
1-617-556-4104
LISTEN INC.
580 HARRISON AVE. SUITE 3W
BOSTON, MA 02118