Harman’s Dr. Sean Olive explains the Harman Reference Curve and why immersive audio is the future

Dr. Sean Olive, Senior Fellow, Acoustic Research for Harman International, has written over 50 research papers on the perception and measurement of audio for which he was awarded the AES Fellowship Award in 1996, two Publication Awards, and the best peer-reviewed paper award at AES. He was also awarded the ALMA Titanium Driver Award for scientific contributions to the loudspeaker and headphone industry and in 2019, he received the Harman External Leadership Award.

In this Everything Audio interview, he explains how the Harman Reference Curve came to be and why it matters, as well as looking at industry trends and sharing his advice for the industry.

Before we delve into your role at Harman and the specifics of what you’re working on currently, could you provide some insight into your educational background?

Like many people in this industry, I started out as a musician. I studied music at the University of Toronto, mostly focused on classical, and I had an interest in jazz. Along the way, I became interested in audio. I lived in a residence with many engineers who built their own speakers and amplifiers and that somehow led me to a degree in sound recording. I studied at McGill University and graduated as a tonmeister. 

We went on a field trip during that course, and I met a guy named Dr. Floyd Toole, who is a scientist at the National Research Council in Ottawa, and he was doing these double blind listening tests where we went into his lab and listened to speakers and rated them on various scales. Then he measured them acoustically and I just fell in love with that whole idea of measuring people's perceptions and preferences. 

I ended up spending seven years in his lab, and in 1991, he got recruited by Harman International, and he moved to California. A couple of years later, he hired me, I joined them and we continued doing research into loudspeakers and listener perception in 1993.

Tell us about your current role at Harman in acoustic research and development; what does that entail and what do you oversee?

My title is Senior Fellow, and I work as part of a corporate research group called Harman X. Specifically, we're in a group called intelligent audio. We do research into product development that can benefit all of our divisions, which includes consumer audio, professional audio as well as automotive audio, so it's quite a broad agenda. 

Much of my research is focused on sound quality, so there's always a subjective aspect where we have to do listening tests and find out whether what we're designing actually sounds good and whether people like it.

You were the president of the Audio Engineering Society; what did you devote your time to while serving in that role?

For people that don't know, the Audio Engineering Society is the only international professional society dedicated to audio research technology and the allied arts like sound recording and music production. 

I've been involved with them as a student since 1985 and have worked in various roles. In 2013, I was elected as president – it's only a one year role – but you actually work five years in various committees; it's really focused on the needs of the organization and to serve the needs of the members. That involves conventions, various conferences on different topics of audio, and generally expanding and maintaining the membership levels.

Many in the pro audio sector will be familiar with the Harman Reference Curve (which you oversaw), which is also referred to as the Harman Target Curve, and when applied to headphones, the AKG Reference Response Acoustics Curve. For anyone unfamiliar with it, could you explain what the Harman Reference Curve is?

It's basically a frequency response of a headphone. If you measure it in a standard industry ear simulator or an artificial head, (which simulates the outer ear, as well as the ear canal on the eardrum), if you put a set of headphones on this device and put a frequency sweep through it, the output will be a frequency response. So we defined what the frequency response should be in order to sound good to most people, but getting to that point was a fairly long process. 

We started this research in 2012 and we also came up with a target curve for in-ear monitors. We tried to define what makes a headphone sound good. That was done by doing lots of experiments with hundreds of listeners all over the world – having them adjust this target, the bass and the treble, and then testing it against the current standards as well as many competitors and measuring people's preferences.

What did you discover through years of research and refinement?

We found that the majority of people generally prefer this curve, and if they didn't prefer it, they preferred it with some slight modifications to the bass and treble. In general, 64% of the people liked it as it is without any alterations, then there was a 16% segment who liked it with more bass – 4 to 6dB. Then there was a segment of around 22% of people who prefer it with less bass – 2 to 3 dB less bass and slightly less treble, or slightly more treble.

We think this could be related to hearing loss, particularly because as people age, they tend to lose hearing at high frequencies, and by reducing the bass and boosting the treble they may in fact be compensating for this hearing loss and making voices and instruments more intelligible. It's based on a preferred sound quality over headphones, which can be traced back to a frequency response.

Tuning speakers and studio monitors to be neutral is ideal, however AKG discovered that tuning headphones to be absolutely neutral presents a problem because they don’t sound right to listeners. Why is it that speakers sound different from headphones in a room?

There's many reasons. When we listen to headphones, we're not hearing the room, so that's something you have to account for. Secondly, when we listen to speakers in a room, we're hearing not only the direct reflected sound – there's really no effect on how this is affecting our physiology. In other words, we don't have something clamped onto our head. As soon as we put something on our head, it interacts with our pinna (or external ears).

Depending on how we angle it, it can change. So if there's a slight leak in the headphone, (because it doesn't quite fit properly), that can have a big effect on the bass. Those are all variables that can cause sound to vary a lot on a headphone. One of the goals at the beginning of this target curve is we took a pair of good loudspeakers that measured flat and they were accurate. 

We used a pair of Rebel consumer loudspeakers and we took a JBL M2, which is a professional monitor, and we put them both in this reference listening room. We measured at the person's eardrum – in this case, we're using an artificial mannequin and that became the basis of this reference curve. 

The goal from the beginning was to make a headphone sound like a pair of good loudspeakers in a reference listening room. We believe that because of this, we're as close as we can get to making a headphone sound like a pair of good loudspeakers in a critical listening room.

You mentioned JBL’s M2 Master Reference Monitor; how does this product tie into the reference curve?

The loudspeakers precede the headphones. It started back at the National Research Council in the mid ‘80s, where Floyd Toole published a set of seminal papers on loudspeakers and listener preferences. He did a lot of listening tests to show that a particular set of anechoic measurements would produce positive listener responses. 

That work continued on when Floyd and I both joined Harman in the early ‘90s. We continued to refine those measurements and did lots and lots of listening tests over the course of 15 years, and came up with a set of adequate measurements that can accurately predict with about 86% confidence how people will score it in a blind listening test.

We started out defining, ‘What is a good loudspeaker?’ and we used that to define what makes a good headphone. There's a connection there whether you monitor or make recordings through a pair of two speakers and listen to it through the AKG headphones – there should be a similarity between the two because we use the same criteria for all of our products, whether they're consumer, automotive or professional speakers. 

There's a continuity between the whole audio chain from recording, production, playback, a car, a home theater or a set of headphones.

How effective are the AKG K361 and K371 professional studio headphones at implementing the Harman reference curve, and why do professional users benefit from this?

Those were the first AKG professional headphones where the Harmon Target Curve was applied. They came out about five or six years ago. They were designed to this target so that when you switch back and forth between your monitors in the studio or over headphones, there's a continuity between how it sounds. 

The advantage for a mixer is that they can choose whatever is most appropriate depending on what kind of environment they're in. If they're doing a mobile recording, they can listen to their mixes over the headphones, or if they're in a controlled studio environment, they can monitor over their speakers.

There's a pretty strong chance that the consumer will hear what the artist or the mixer heard when mixing and mastering. It's all about being faithful to what the artist wanted, and giving them tools so that there's confidence that the consumer will hear what they're hearing.

What feedback have you had from engineers, producers and anyone working in a studio on these headphones?

Based on the reviews and speaking with people, it's been hugely positive – the K371 in particular. When they first came out, the HiFi community was quite impressed with them – you can find all kinds of awards, reviews and praise about how neutral they sound. 

The main difference between this new target and traditional professional headphones might be they have a little more bass. If you're monitoring with open-back headphones, for example, they tend to be very flat in the bass because it's very hard to produce bass with the open-back, whereas these are closed and have bass that's more representative of what a consumer might hear over their loudspeakers in their homes, for example.

What does the future of audio look like? In particular regarding immersive experiences, spatial audio, and the HRTF lab?

I think the future is already here. When we're talking about immersive audio, it's really being driven by new immersive audio formats like Dolby Atmos, DTS:X, MPEG-H (which is more common in Europe), as well as Sony 360 Reality Audio. 

These formats allow audio to be reproduced in 3D, so they're adding an extra layer of height channels so that you have not only front and back, but also the up and down experience. What we see is these formats are becoming very popular, because they're available now for music on streaming services like Tidal, Apple Music and Amazon, so consumers can get easy access to them through streaming – it's not disk space.

It's also object-based, it's not channel-based audio, so it's very scalable to whatever playback scenario you have, so you can play it back over two-channel headphones. It's normally rendered so that it knows when it's played back, what device you have, how many, and what its capability is. 

If you have a home theater system with 32 speakers, it will scale up to that – it gives you an added dimension of height, and it's much more immersive. It's also finding its way into gaming, virtual reality and augmented reality. It's where the future is and where it's going.

HRTF – we call it the immersive audio lab – or where we do research on various aspects of rendering audio for 3D as well as making products that can render this new audio. One of the features our lab has is something called a hyperion sphere. It's a very large structure that looks like a big sphere, which is a framework that holds a number of speakers. 

In the center of the sphere is a chair where we put people with microphones in their ears, and we basically rotate them through this 360 soundfield and we capture what's known as the head-related transfer function, or the HRTF.

Do you have any advice for the industry moving forward?

I think anyone moving forward will probably be working with immersive audio. That’s what will differentiate the really advanced products, and the less advanced products will include the ability to personalize the experience. 

So for 3D audio or immersive audio that has to do with incorporating their head-related transfer functions; it could be compensating for the fact that their headphones don't fit them quite well. Maybe there's some kind of way to adapt the headphone to account for some of these personal differences.

As far as hearables, which is a pretty new category, these are headphones that account for people's hearing loss. I think the future with headphones will probably have that incorporated into them, so they'll do more than just play music – they'll compensate for hearing loss. These are physiological anthropometric differences I'm talking about, so I would just say in the future, be aware of this.