JBL Hearing the Truth: Testing
In the second part of our JBL Hearing the Truth series, we dive into the methods JBL uses when testing their loudspeakers for the quality of their construction and fidelity. There are a lot of elements that go into testing different variables from listener tests to rigorous performance tests. Sean Olive, the President of the Audio Engineering Society, talked us through how JBL sets up blind evaluative testing systems so they can determine which factors of a speaker’s fidelity are most pleasing to the trained ear but also the novice listener. Peter Chaikin also showed us the JBL power test room at both their offices in California as well as their manufacturing facility in Mexico. There are plenty of other testing methods utilized so be sure to check out the video to see how JBL incorporates scientific measurements into their product evaluations. If you haven’t seen our first video in this series, be sure to check our YouTube channel which covers the history of JBL Professional.
Check out the video above to see more from the JBL Professional facilities and be on the look out for the third part in our Hearing the Truth series titled Design. If you would like to pick up a pair of JBL LSR 305 Studio Monitors, be sure to visit us over at UniqueSquared.com.
TranscriptMy name’s Sean Olive. I’m the Director of Acoustic Research for Harman International. I’m also the President elect of the Audio Engineering Society. Harman takes a very scientific approach towards how we measure sound quality in terms of listeners and also how we objectively measure sound in terms of acoustical measurement. So our basic philosophy is if you want to measure good sound and develop good sound loudspeakers, you have to have very accurate subjective measurements and very accurate objective measurements. So over the last few years we have developed some of the most sophisticated listening rooms, listening test methodologies, and also some of the world’s best anechoic facilities as well as measurements. We actually developed our own measurement system which we call HATS which stands for the Harman Audio Test System. My name is Charles Sprinkle. I am the Senior Acoustics Engineer here at JBL Professional and I am going to show you a little bit about how we voice our loudspeakers and the metrics that we use. We use a system that measures loud speakers not in just one location but in 72 different locations in 10 degree increments across the horizontal and vertical plane. What that does is it gives us a better indication of how the loud speaker is performing in space not just in one location in the room. It allows us to predict what the perception of timbre will be in the room, what the timbre of the first reflections are going to be and I am going to show you a couple of the metrics that we use. The black curve on the top is the direct sound. This is what is picked up directly on axis of the loud speaker. The green curve is what we call the listening window and that is a combination of measurements plus or minus 15 degrees horizontal plus or minus 10 degrees in the vertical plane and that tells us about somebody who is sitting nearly in the center of the louds peaker, what they can expect to hear over that space. The red curve is the first reflections or early reflections and that’s what the listener can expect to hear from the walls and the floor and ceiling integrated and how that relates in timbre. Then the blue curve is the total sound power which is an integration of all of those 72 measurements together. The blue curve on the bottom is directivity index or basically how the sound power relates to the listening window. The red curve is the first reflections directivity index and how the first reflections relate to the listening window. What makes our loud speakers and how we voice our loud speakers so unique is we’re paying attention to the first reflections, the total sound power to make sure that that sound is substantially the same in timbre with the on axis frequency response. What that does is it gives you a greater sense of imaging of soundstage and depth. It makes the loud speaker sound more natural. You’ll notice that these green curves, that the actual general shape of these curves are very similar and you’ll also notice that these directivity index curves are also very smooth. This is something that is fairly unique to JBL loud speakers. You can have a loud speaker that has very flat frequency response, you are only going to get that frequency response flat in one space. Anywhere else in the room it’s not going to be right and you can’t make it right. If you don’t have good directivity response, you cannot get that loud speaker to sound the same in different spots across the room. The time we spend making sure our speakers have good directivity index gives our loud speakers the ability to sound good not just in one position but across the entire room. A very broad sweet spot gives the ability, as an example for a studio monitor for a studio engineer, to be not just at their primary mix position but anywhere across the board and get the same story and to be able to have confidence in the mix that they are creating. So what we’ve done is we’ve developed our own training process. It’s called How to Listen. It’s a piece of software we’ve developed and each listener undergoes several weeks of training where they learn how to identify and describe different distortions that we simulate and add to loudspeakers. It gives automatic feedback on the performance, it gets more and more difficult as they progress, and it gives us a way to quantify how discriminating and how reliable a listener is so we only select the very best listeners for our product evaluations. So one of the keys to developing really good listening tests is you have to control what we call nuisance variables. These are psychological factors, physiological factors, for example the listeners hearing, and other nuisance variables include the listening room, the acoustics, the speaker position. What we’ve done is we’ve looked at every possible nuisance variable that can bias or make a test unreliable and we’ve taken great strides to try to remove them or control them. One of the first things we did was develop a listening lab which we call the multi channel listening lab. You’ll see this in a little while. We have pneumatic speaker shufflers that position each speaker in the exact same position in the room and that means that every speaker is basically exciting the room and resonances, the reflections, they are all basically the same from each loud speaker. The second major variable that we’ve controlled is the sighted biases. So we have an acoustically transparent screen which comes down so the listener has no knowledge of what speaker they are listening to. They have no information about the price, the brand. All they have information is the sound so they are strictly focusing on how good the loud speaker sounds. What we do is we go out and we purchase our competitor’s products and we do A, B, C, D comparisons of our prototype and we compare them against our competitors and the goal is to have a speaker that sounds better than our best competitor. What people like, what they prefer is a very accurate and very neutral loud speaker and that’s exactly the goal that we had in developing the LSR 3 Series. Hi I’m Peter Chaikin, I am the senior manager of recording and broadcast marketing for JBL Professional and that’s the segment responsible for studio monitors. Speakers that are used in post production, recording studios, and broadcast facilities. This is the power test room. Before a speaker becomes a finished design, a market ready design, it has to go through a series of power tests and the power tests prove that the speaker can stand up to the kind of abuse that its specced for. Assuming it passes that test without the amp blowing up, without the speaker blowing up, without the logo falling off of the grill or anything like that, then from a power test perspective the speaker is ready for the market. When you buy a JBL Professional studio monitor you know that that design has been through the 100 hour power test. We don’t know of anybody else that has this kind of stringent test. What this display shows you is the health of each speaker. How long it’s gone through the test, what the signal is that we are putting into it. This measures failure or lets you know when the failure occurred and of course they do a deep analysis, the engineers do a deep analysis of why it failed and then their task is to fix that problem so that the speaker passes the 100 hour power test. So I am testing our stadium line of speakers called PD Line which is precision directivity line. I am testing the SPL, putting it on this lift, and measuring the on axis response and also running entire polar spins on it to characterize the entire acoustic performance of the speaker. That’s measured by putting the speaker on this lift, lifting it up, having a microphone thats six meters away and essentially putting pink noise through it and simulating the SPL, the transfer function, and numerous other acoustic properties that we could obtain through post-processing. The end result is that we get a complete characterization of our speakers so we not only get the on axis response but we get the response at every angle, we get 3D polar patterns, we get the sound power, we get other things such as directivity. So we are able to fully characterize exactly how this speaker could perform in different environments and let us know if we need to tweak anything. So it’s like the best way to characterize and to actually design a speaker. My name is Paul Bauman. I’m associate director of tour sound at JBL Professional. Been with the company seven years and I’m basically responsible for things like product design, conception, and the technical support side of things. Where are we now? This is what we call the large arena. It’s not a place where we would have a hockey game or a basketball game but we do a lot of product listening tests here. You can see there is a cinema perf screen. We can listen to some cinema product. We use these four chain motors here when we are doing Vertec VTX training. We do some hands on practical work with the systems and go through all the different suspension techniques, listening tests on the different preset options for the products and just for our own use when we are doing listening evaluation on new products, well set up listening tests in here and sign off products. That’s a very powerful tool for developing speakers. Being able to optimize their sound quality so we’re not wasting money on things that don’t matter. We’re just focusing on getting these measurements to be such a way that we know with high confidence that they’re going to sound really good.