Harman Kardon's quest to standardise sound

Two men are on a mission, a mission to create a series of speaker systems that recreate the essence of the original recording that was laid down in the studio rather than a speaker that makes it sound better.

It's an idea that has had Floyd Toole and Sean Olive working tirelessly for the last 19 years as they try to create a standard for sound, something that has never been done before.

"You have to know what good sound is before you can create speakers that produce good sound", Floyd Toole tells us.

Dr. Floyd Toole, now an acoustic consultant for the company started the programme which his protégé Sean Olive has now taken over. Both academics, rather than your regular company men, the two have worked together since 1993.

Their department in a rather boring American office campus in Northridge, California tests the latest Harman Kardon products not just until they sound good, but according to them, until they scientifically sound better than the competition.

The "scientifically sound better" is a very controversial statement. There is no standard for sound (yet) and many believe that there can't possibly be one either.

What makes you think something sounds better than something else? According to Toole, for most, it's about something called psychoacoustics the study of subjective human perception of sounds.

However, Toole and Olive have a different idea, a different approach. They believe that sound can not only be measured, but that you can pinpoint good sound as well as recreate it over and over again to appeal to everyone rather than just a select few.

Why are they bothering? Well apart from the sheer science of it all, the economic reasons are simple. If you can recreate a more natural sound that people like, more people will hopefully buy your product.

It is a very different approach from most manufacturers, claim the pair.

"Most of our competition either opt for a "golden ears" that approves the speakers for market or simply add bass or other elements to the speakers to make the recordings you play sound either louder or more enhanced", claims Olive.

What the two gentleman from Harman Kardon believe, is that the speaker should just be a resonator of the sound rather than yet another editor in the process of delivering the music.

"Sound isn't a personal preference we just think it is", comments Sean Olive. So to combat that the two gentlemen have set about creating a series of tests to democratise the process of measuring the "best" sound to ultimately help in creating the most natural sound instead, whether that's a good or bad thing.

The speakers are hidden behind a curtain and then selected randomly by a computer

To do that, the pair has devised a double-blind test that, with thousands of helpers from all walks of life, lets them find that "good" sound through a scientific process.

In simple terms, the double-blind test merely pits Harman Kardon products against its competitors with neither the tester or the tested knowing which system they are listening to. Three speaker systems are tested at a time with all three hidden behind a curtain. A computer randomly chooses the system to play without anyone knowing.

The team use a range of tracks to test people's perception of the speaker systems including Tracy Chapman's “Fast Car”, from the album Fast Car; Jennifer Warnes' “Bird on a Wire”, from Famous Blue RainCoat, and James Taylor's “That’s Why I’m Here” From the album with the same name, plus many more.

While they might not be tracks you are expecting, Olive tells us that they are selected based on experimental evidence from years of testing loudspeakers and listener training exercises.

The important physical characteristics is that the music is well recorded and relatively neutral.

"We use a lot of music recorded and produced by Grammy Award winning George Massenburg because of their high sound quality". Olive comments.

Can he hear what sounds best?

Music that has a broadband in frequency (contains the highest and lowest sounds we can hear from 20 Hz to 20 kHz) and spectrally dense (like pink noise) is also very helpful Olive tells us. In English that means music that offers a wide range of vocals, bass, and instrumental notes.

Music that is continuous in spectral content which allows listeners to compare different loudspeakers under the same conditions or context is also good.

"We further improve this by making short (20-40 s loops) that automatically repeat is a must too", says Olive who admits that he is now very good at Karaoke when it comes to certain tracks.

Harmon says this test is carried out over and over again until a winner is chosen. If that isn't a Harman Kardon product then the product is sent back to the drawing board and worked on until it does.

"In the past, we've had inadequate data and subjectively we allowed ourselves to be biased", comments Olive after letting us sit through one of the gruelling tests where we had to determine which sounded the best out of a Harman Kardon MS 100, A B&W Zeppelin and Bose SoundDock 10.

Were weren't told which was which of course until the end, nor where we told what systems we were listening to until after the test was over. However, after listening to three systems three times each, you could start to hear patterns to the performance of the speakers, with the Harman Kardon MS 100 creating that natural sound Olive and Toole are striving for. Whether that is for better or for worse is up to you and your listening habits, but it's the one, as Olive puts it, "most true" to the recording.

What is clear, however, is that the test creates quantitative rather than qualitative data, knocking down those misconceptions of good sound is perceived rather than scientific.

"If you make it blind, people can't be influenced by the looks", Toole adds.

Of course some people just aren't good listeners, not very helpful when it comes to rating a system, so Toole and Olive have created a listening test to examine prospective listeners before they are given real testing to do.

Listeners are tested to train their ears to hear better quality sound

The software, which has been developed in house at Harman Kardon tests listeners over a number of pieces of music to see how they rate certain types of music.

It's not a case of whether they like or dislike James Taylor Quartet, but whether they can tell the difference between whether the music is flat or with more bass.

With around 20 levels to the test, the higher the level the harder it is to determine the difference. Steve who created the test is at level 14, although there are listeners in the programme that are around 15 or 16, he tells us.

The good news, is that Harman Kardon are planning on launching the software for Mac and PC users so they can train their ears to be better. Better ears, says Dr Floyd Toole means the urge to listen to speakers that produce better sound, and that commercial reasoning is there again.

"Nobody asks to buy the same window when they see a great view out of it, they expect the window to allow you to see the view as untouched as possible", says Olive, when describing the goal for the company's speaker system, before going on to tell us that "the ultimate goal is that if you don't like the sound it isn't the loudspeaker but the recording".

It's one of the age old problems of the recording industry. They record it to sound one way, and then you play it in a room that is completely different from virtually everyone else's. How do you make it sound like the artist intended it to?

"The room is the final audio component, regardless of what you do with the box [sound system] it is the room that affects everything. Bass accounts for 30 per cent of perceived quality, however bass is the sound that is most effected by the room in which the music is played".

Through science, Toole and Olive are hoping to work out what audio sounds like in certain rooms and then tackle that within the settings. Settings that are about recreating the original sound rather than just adding an equalised Pop or Jazz to the music.  

That's the future plan: to have scientifically based subwoofers that deliver the same sound in different rooms because it knows what it is supposed to sound like in the first place, and it knows how to create that sound in any environment.

It's what Harman Kardon says it is already able to do in the car, because it knows the dimensions of the environment and how that reacts to the sound created in it.

Will there be a time where your sound system in your car automatically adjusts depending on how many people are in it. Olive says it's all possible, and sooner than you think.

Testers never see what they are testing until afterwards

But back to standardising sound. With Harman Kardon pushing for that to happen, but not many other manufacturers currently keen to support it, it looks like we have some time to go before every speaker will sound the same.

It's an interesting dilemma and one that if successful could mean all speakers sound the same, it's just the design that differentiates them.

But more interestingly it throws the task of creating good sounding music back to the recording artist, producer and the equipment they use.

Olive and Toole both talked at length about another artistic area that has similar questions: oil paintings. Should galleries show old oil paintings dimly lit by candles to represent the artist's gallery at conception, or in the bright light environments that they normally opt for?

With music it's a given that you strive to recreate the event. Live music sound engineers try their hardest to recreate the music as best they can, and Olive and Toole believe they can help deliver that experience.

Some, of course, believe that sound is perceived, that you can't simply measure it, and that even different cultures have different sound tastes with the Americans preferring more bass than Europeans and Japanese. Olive believes that to be not true.

"Very little research has been done in cross-cultural preferences in the sound quality of reproduced sound. If cross-cultural preferences exist, the music and audio industries have largely ignored catering to them, instead distributing products that are optimized for a single universal audience".

As for Toole, he still hopes that his life work will see a standardisation, when? "Hopefully in my lifetime", says Toole, but with the researcher already retired (he is 71) Toole hopes that's sometime in the next 20 years.

Floyd Toole and Sean Olive are proving that sound can be standardised, that we can for better or worse recreate good sound over and over again. Whether that is a good thing is yet to be heard, however one thing is clear, the science of sound isn't just about the notes that are created but how we listen to them as well.

What do you think? Should sound be standardised, or isn't it as simple as that? Let us know your thoughts in the comments below.



>