CLEAR^TM Distortion Measurement: Perceptual Rub & Buzz Test

Introduction to the NEW CLEAR^TM Perceptual Rub & Buzz Measurement Method

The new CLEAR (Cepstral Loudness Enhanced Algorithm for Rub & Buzz) algorithm from Listen offers true perceptual Rub & Buzz analysis for production line applications. It uses a simplified auditory perceptual model to measure the loudness of Rub & Buzz distortion in phons rather than the more traditional dB SPL and % distortion units. These better identify whether distortion due to manufacturing defects can be heard by the listener than conventional measurements. In addition to a result which corresponds more accurately to the human ear, this new test method also offers two significant advantages for use on the production line. It is less sensitive to transient background noises than traditional methods, therefore is reliable in noisy environments, and it is much simpler to set limits than when using conventional distortion measurements.

Perceptual Rub & Buzz using the new CLEAR™ algorithm shows audible distortion more clearly. Traditional Rub & Buzz measurements do not take into account the insensitivity of the human ear to low and high frequencies, therefore it is more difficult to identify problem areas and set limits on a production line. This speaker clearly shows audible distortion below resonance (<100 Hz) and at cone breakup (750 Hz). Both curves are shown on a 60 dB scale for easier comparison even though the units are different (Perceptual distortion in Phons and traditional distortion in %).

Perceptual Versus Conventional Rub & Buzz

Conventional Rub & Buzz detection has been widely used on the production line since Listen introduced it in SoundCheck Version 1 back in 1996. It offers excellent identification of Rub and Buzz defects caused by manufacturing problems, and will continue to do so. In recent years, some manufacturers have moved to a defect detection model where they prefer only speakers with audible faults to fail QC checks. This is because if only speakers with audible faults rather than any faults are rejected then yields are higher. Perceptual Rub & Buzz offers a means of identifying and precisely quantifying this with all the benefits and reliability of an automated test system.

The CLEAR^TM Rub & Buzz Detection Algorithm

Listen’s CLEAR^TM Rub & Buzz detection algorithm uses true perceptual analysis to ‘hear’ any faults in the speaker. It offers many advantages over other ‘perceptual’ Rub & Buzz analysis systems:

• Low Cost: The CLEARTM algorithm to add-on to SoundCheck costs just $500. Other test equipment manufacturers charge up to $10,000 for their Rub & Buzz software.
• True Perceptual Rub & Buzz: The CLEARTM Algorithm is a true perceptual Rub & Buzz algorithm. Based on well-proven psychoacoustic principles, it accurately replicates the human ear using mathematical models found in MP3 encoders that mimic the way that both the ear and the brain interpret sound. This results in close to 100% correlation to the human ear.
• Less sensitive to transient background noise: A significant advantage of our Perceptual Rub & Buzz algorithm is that it is very insensitive to transient background noise – tests show that it offers far more consistent results with high background noise levels than other Rub & Buzz measurement methods. This makes it ideal for noisy factory environments.
• Flexible: The CLEARTM Rub & Buzz detection system is extremely flexible. While it can of course be configured for a simple pass/fail result, it can also offer detailed results including defect analysis and offers a calibrated loudness value rather than simply a comparison to a reference.
• Better Correlation to Human Ear: Testing carried out by an independent laboratory shows excellent correlation to the human ear.

How The CLEAR^TM Rub & Buzz Detection Algorithm Works

Unlike other Perceptual Rub & Buzz detection systems that simply claim 'unmatched correlation' based on 'patented technology', Listen's Perceptual Rub & Buzz distortion measurement is based on well-proven psychoacoustic principles. It accurately replicates the human ear using mathematical models found in MP3 encoders that mimic the way that both the ear and the brain interpret sound. Since these methods are so well established and documented, we are confident being open with our users about our test methodology. The flow chart below offers a simplified representation of the CLEAR^TM algorithm.

 Download the CLEAR Flyer

 

A special offer for VoiceCoil Readers

Tell us that you learned about the CLEAR algorithm in a VoiceCoil Advertisment and receive 20% ($100) off the cost of the CLEAR Module (part number 2030). Discount is on part number 2030 only; SoundCheck version 9.1 or later is required.

Would you like to learn more about the research and development behind this algorithm? We have written 2 AES papers on the subject. The first (presented in 2009) entitled 'Practical Measurement of Loudspeaker Distortion Using a Simplified Auditory Perceptual Model' explains how the model was derived from the ITU PEAQ standard, and second (presented in 2010) entitled ' Practical Implementation of Perceptual Rub & Buzz Distortion and Experimental Results' describes the modification of this model to used common audio test signals, describes its development for production line use and presents experimental results. To receive copies of these papers, fill out the short form below. You may also request a demo download (with the CLEAR module enabled) if you wish to try the software.