Wednesday, July 3, 2013

The Science and Marketing of Sound Quality

To my surprise, this morning an audio friend tweeted a link to an article I recently wrote for our company's  internal newsletter  entitled, "The Science and Marketing of Sound Quality."  My article can be found on a new Harman Innovation website  launched today that features articles on current and future disruptive technology that will impact consumers' infotainment experiences. Check it out.

My article focuses on a longstanding pet peeve of mine (first mentioned in this blog posting): The lack of  perceptually meaningful loudspeaker and headphone specifications in our industry.  While consumer surveys repeatedly report sound quality to be a driving factor in their audio equipment purchases, consumers lack the necessary tools and information to identify the good sounding products from the duds.

This is particularly true for loudspeakers and headphones where the typical throw-away "10 Hz to 40 kHz" specification provided by the manufacturer is utterly useless. This specification only guarantees that the product makes sound, with no guarantee that the sound is good.  While the science exists today to accurately quantify and predict the perceived the sound quality of  loudspeakers (and hopefully, soon headphones), the audio industry continues to drag its heels into the 21st century,  and not routinely provide this information to consumers.

A rare exception is JBL Professional who provides comprehensive detailed measurements on studio/broadcast monitors like the new JBL M2 Master Reference shown below. Inspecting the measured frequency response curves shown  below, you can easily recognize the loudspeaker sounds exceptionally neutral and accurate based on the shape (flat, smooth, and extended)  Based on this set of measurements, we can predict how a listener would rate the sound quality of the loudspeaker in a controlled listening test, with 86% accuracy. The only pertinent information not shown in this graph is how loud the loudspeaker will play before producing audible distortion (trust me, this loudspeaker will play very loud! )

Perceptually meaningful loudspeaker specifications like these have been available for almost 30 years! Yet,  these specifications are currently not part of any professional and consumer loudspeaker standard. Such a standard would go a long way towards improving the quality and consistency of recorded and reproduced sound. Audio consumers want to hear the truth. We need to provide better information and audio specifications so they can find it.

JBL M2 Master Reference Monitor provides true reference sound quality that is clearly indicated by its technical measurements shown below. 
The spatially-averaged frequency response curves of the JBL M2  (from top to bottom) for the listening window (green), the first reflections (red), and the total radiated sound power.  At the bottom are shown directivity indices of the sound power (dotted blue) and first reflections (dotted red). These measurements tell us that the quality of the direct and reflected sounds produced by the loudspeaker will be very accurate and neutral over a relatively wide listening area.


  1. After reading Bose's obit in the NYT and then thinking about Harman within the industry ,I came upon your blog.
    I'm a layman, I have a basic understanding of audio acoustic but music and sound has been the forefront of my life. I live and breath music 18 hours a day and I thrive on the introduction and marketing of new products to the public.
    Sound and marketing is all built om perception.Beats head phones terrible, but the public doesn't understand. It's like food ,most people eat junk food because they've been told to eat it because they don't know no better.
    This is what your up against.
    If Microsoft can attempt to have all divisions work as one ,why can't Harman?
    It's very sad that Harman acts like many separate company's instead of one, the divisions and products should integrate. ( to be continued it's late,love the paper and blog)

    1. Hi Peter
      Good idea. You may not be aware of this but the most recent reorganization of Harman into Lifestyle and Harman Professional divisions addresses your suggestion of having brands all under the same roof to focus their engineering and marketing resources on the same purpose.

      The Lifestyle Division includes all of the brands used in consumer and automotive audio products. The Harman Professional includes all the brands in professional products.

      The idea behind Lifestyle was to have a more consistent focus and message about each brand across consumer and automotive products. I see a lot of advantages in this organization, and so do our customers.

  2. Hello,
    first of all thank you for your great blog as well as your efforts together with Floyd Toole to rationalise the understanding and hearing of speaker performance. I don't understand though how do you expect others to show and use their measurement specifications when in the whole Harman Group you have to pick a high end Pro speaker that gives those data, why not also in the other series as a first step? It takes more birds for the spring to come.
    Best regards,

    1. You make excellent points. Harman should lead by example if we want our industry to adopt more meaningful loudspeaker performance standards and specifications.

      Our consumer loudspeaker products are measured the same way as our Professional speakers as illustrated with the JBL LSR series. So we have the measurements but haven't yet incorporated them into product brochures and literature on the web sites.

      One of the challenges is educating consumers about what the curves mean so they understand that the product sounds good. I think this can be easily done with the right training materials.

  3. Dr. Olive, I agree completely but I don't know if your soundboard from blogspot is big enough to make the kind of impact it should. Perhaps editorials in the british audio magazines would help generate some much needed discussion of the topic.

    Sound Stage has put some effort into creating a database of NRC tested speakers here:

    1. Good idea. How do we get audio magazines to write editorials to advocate scientific meaningful loudspeaker measurements?

      Sound Stage is rather the exception and has indeed set a very high standard in reviewing speakers based on the NRC loudspeaker measurements

    2. Have you spoken with the marketing department at Harman International about starting a campaign to positively reinforce magazines who'd publish scientifically meaningful tests and editorials, maybe along with helping the publications set up measuring environments where they could independently tests products ?

  4. I think we need something even better then these loudspeaker measurements. We need polar plots, horizontally and vertically.

    Having a horn with controlled direcitivy (CD) down to 800 Hz isn't bad, but it's not great either. That means the response below will not have CD, being much wider and the reflected energy from the room will be colored unless broadband treatment is utilized at the reflection points. So even this speaker, has it's limits and which a polar response would reveal.

    Still a nice speaker, but also with a hefty price tag....

    1. I'm not a big fan of polar plots. Polar plots are limited to a single frequency so unless you have multiple plot taken at every 1/12-octave frequency you don't have sufficient frequency resolution to tell you what you need to know. Another problem with polar plots is they have no spatial averaging so you see acoustical interference at a single point in space, which may not be perceptually relevant.

      We have calculated both horizontal and vertical directivity indices for M2 and they are almost identical.

      Broadband absorption is necessary for any speaker if you want to treat the reflected sounds as neutrally as possible.

      $20k gets you a pair of these speakers including 4 channels of Crown I-Tech 5000 and DSP for speaker crossovers and room correction. If you look at what other professional companies charge for a speaker like this, it's very reasonable.

  5. Wow the M2 Master Reference has perhaps the flattest on-axis frequency response I've ever seen.

    1. You could be right. I've not seen anything flatter myself.

  6. Great post! I have been working with sound since the 90's, and there is some really great information here. Keep up the good work!

  7. Dr. Olive, I don't know if you are still reading this blog but I would like to thank you as a voice of sanity in a sea of audiophile woo-woo madness and the Harman/JBL engineers for making the LSR305 to prove one doesn't even need to spend over $300 to enjoy incredible sound.

    The sheer amount of lies, misinformation or outright scams and the misguided fools who believe that while perpetrating pseudoscience in their religion of placebo in consumer audio circles is sickening and has greatly harmed the market as a whole.