MADE IN MAID­STONE

Since 1961 KEF has been engi­neer­ing cut­ting-edge speaker tech­nol­ogy, tech that looks ev­ery bit as in­cred­i­ble as it sounds

What Hi-Fi (UK) - - Feature -

"It’s kind of a mag­i­cal thing for me," says Jack 0clee-brown, . KEF's head of acous­tics. "You get a mag­net, a bit of pa­per and a bit of wire, and you can make a speaker."

Jack’s PHD in loud­speaker acous­tics sug­gests it might not be quite that sim­ple. KEF has been putting mag­nets, bits of pa­per and bits of wire to­gether for the past years at its head­quar­ters in .maid­stone, ,ent. It’s still in the same build­ings it used when EX-BBC elec­tri­cal engi­neer Ray­mond Cooke started the com­pany in 19 1, and it still fol­lows his phi­los­o­phy: make the speak­ers sound as nat­u­ral as the orig­i­nal per­for­mance.

Jack is in charge of mak­ing ab­so­lutely sure that hap­pens.

So he’s the liv­ing em­bod­i­ment of ,KEF's tone, then? No. "The ,KEF's sound comes from the tech­nol­ogy, in a sense," Jack says. Imag­ine the speaker in­dus­try or­gan­ised in a long line from tech­ni­cal to pure artistry – ,KEF'S is res­o­lutely at the engi­neer­ing end of the spec­trum.

That all started when the com­pany did. "It was founded by an engi­neer," says manag­ing direc­tor Steve Hal­sall. "He founded it be­cause

the com­pany he was work­ing for was us­ing late-19 0s pa­per cone tech­nol­ogy. That was fine for mono speak­ers – but when stereo came around, you sud­denly wanted two speak­ers to mea­sure ex­actly the same. Ray­mond re­alised the pa­per cones of the day Kust weren’t con­trol­lable enough to be able to do that."

A new set of stan­dards

So he ex­per­i­mented. Ray­mond brought new ma­te­ri­als to the party: more sta­ble ma­te­ri­als, such as plas­tic, that would give a bet­ter stereo im­age. The com­pany he was with at the time wasn’t hav­ing any of it, so he went and set up ,E' to pur­sue those av­enues by him­self. Ôdur­ing the 19 0s, ,E' was fa­mous for thor­oughly re­searched, sci­en­tific de­signs – de­signs that gen­uinely tried to make the sound bet­ter,õ says Steve.

And that’s why, on ,E'’s site, there’s an ar­ray of pur­pose-built test­ing fa­cil­i­ties, in­clud­ing an ane­choic cham­ber (an in­su­la­tion and acous­tic wedge-filled five-me­tre cube that’s so quiet you can hear your own heart beat­ing) and an im­pulse-re­sponse room (a mas­sive, high­ceilinged concrete cathe­dral of a space). KEF was the first to use a com­puter to mea­sure a loud­speaker,õ Jack says. "This room was built specif­i­cally for that – to mea­sure the re­sponse of the speaker to a lit­tle click.õ The room is so big be­cause you need to get the speaker far enough from the walls and floor to mea­sure it ac­cu­rately. There are hoists and pneu­matic lifts and clamps and mi­cro­phones sus­pended from other pieces of equip­ment – all for a click. But that click is ab­so­lutely cru­cial for see­ing how a pro­to­type is be­hav­ing.

Ôbe­fore com­put­ers, you’d use a plot­ter: you’d put the speaker in the ane­choic cham­ber, put dif­fer­ent fre­quen­cies through it, and you’d plot. A pen would draw the mea­sure­ment on a piece of pa­per. So if you wanted to store it, you’d put it in a fil­ing cab­i­net. If you wanted to send some­one the fre­quency re­sponse, you’d fold it up and put it in an en­ve­lope. If you wanted to com­pare two speak­ers or do pro­duc­tion mea­sure­ments with sev­eral, you’d have to use a light ta­ble. A big light ta­ble and a guy with a lot of pa­tience,õ says Jack.

The process was so la­bo­ri­ous that pro­duc­tion con­trol us­ing acous­tic mea­sure­ments was rare. Com­put­ers, and ,E'’s first 'ourier Trans­form anal­yser, changed that in 1974. Ôit’s a stan­dard type

of com­puter anal­y­sis," says Jack. "Your phone can do it nowa­days, but at the time you had to buy this big huge box – it was about c2 0,000 even back then."

Spending the equiv­a­lent of c2. mil­lion in to­day’s money shows ,E'’s to­tal com­mit­ment to (et­ting It Right.

Why? Ôbe­cause that’s what keeps you go­ing,õ Steve says. Ô:ou’re al­ways try­ing to get that 0.1 closer to what re­al­ity sounds like. 0nce you get it work­ing and you hear things on record­ings you never no­ticed be­fore, the hairs on the back of your neck stand up.õ (et­ting It Right means emo­tion con­vey­ing that per­for­mance to the lis­tener.

(et­ting It Right also means pro­to­types, test­ing, anal­y­sis, tweak­ing, sketch­ing… all of which can now be done vir­tu­ally. Even as re­cently as the turn of the mil­len­nium it was tricky to model, say, a tweeter on a com­puter you Kust didn’t have the horse­power avail­able. ÔI’M run­ning a model now that will take about ten hours – but you’d have had no chance of do­ing that in 2000,Õ says Jack.

Test­ing sim­ple mod­els on com­put­ers means putting a tur­bocharger on de­vel­op­ment times. :ou can find out if you have the tweeter dome the ex­act right shape the cab­i­net the right si[e the ports the cross­over the glue that holds the thing to­gether. All ex­actly right. And then you can build a pro­to­type.

But it isn’t Kust a case of plug­ging in a few mea­sure­ments and see­ing what hap­pens. ÔIF you lump ev­ery­thing into a com­puter model, it’s Kust as com­pli­cated as look­ing at the real thing and try­ing to fig­ure out what’s go­ing on,õ says Jack.

The team re­alised that they could iso­late a com­po­nent vir­tu­ally to see what it’s do­ing: whether the is­sue is an acous­tic one, or a me­chan­i­cal thing maybe it’s even the air be­hind the dome. ÔSO you tell the com­puter the dome is per­fectly rigid – some­thing im­pos­si­ble in real life – and see if that gets rid of the prob­lem,õ adds Jack.

How to get it right

)e brings up a model on his com­puter – it’s a pie-slice of a midrange driver, com­plete with sus­pen­sion, bas­ket, mo­tor and cone ma­te­rial. Jack shows how chang­ing the si[e of the driver, the cab­i­net and the port all af­fect the re­sponse graph. Ôit’s a very quick sketch­pad. :ou can quickly work out the ba­sic parts and pa­ram­e­ters for the thing you’re de­sign­ing. Then you add more and more de­tail.õ

The sys­tem uses pic­to­rial schemat­ics to as­sem­ble the speaker’s in­nards, and then spits out a graph – which can be ma­nip­u­lated how­ever the engi­neer wants. Want to make the cone im­pos­si­bly rigid? Click. Re­sult. Want a ten-foot cab­i­net with a six-inch driver Kust to prove to fea­ture-writ­ers that it’ll never work? Click. Re­sult. Want to place a vir­tual mi­cro­phone in a vir­tual room to test the vir­tual speaker’s off-axis per­for­mance? (Ô.icro­phones are free when they’re vir­tual, so I’ve put loads of them in Õ) Click. Re­sult.

It’s those ex­per­i­ments that let Jack and his team it­er­ate quickly and ef­fi­ciently, drilling down to iron out prob­lems be­fore they ever make it to pro­to­type stage or, god for­bid, the fac­tory floor.

:ou can set it go­ing overnight, then get back to your desk in the morn­ing and see if you’ve cracked it. This has trans­formed the way ,E' works. Ôprob­a­bly the key prod­uct in terms of re­ally show­ing what we could do was the Uni-2 and the orig­i­nal .uon,õ Jack adds.

Say hello to Uni-Q

The Uni-2: the point-source driver. The Kewel in ,E'’s acous­tic crown. It was in­vented in 1988 to solve the tim­ing prob­lems you get when sound from a tweeter placed a few inches away from the midrange driver reaches your ear at a dif­fer­ent time from the midrange sound. It’s a mi­nus­cule dif­fer­ence, but it affects the sound neg­a­tively – es­pe­cially when you’re sit­ting off to the side of the speaker.

The Uni-2 driver places the tweeter at the ex­act cen­tre of the midrange driver and re­moves that ‘sweet spot’. The sound comes from a sin­gle point in space, there’s no mud­dy­ing of the acous­tic waters, and you get a more nat­u­ral per­for­mance.

In the­ory. ÔIT had a load of is­sues,õ says Steve. ÔIN the very be­gin­ning it wasn’t a good hi-fi drive unit, but be­cause the idea was great, it was de­vel­oped and de­vel­oped. It’s prob­a­bly in its 13th or 14th gen­er­a­tion

You're al­ways try­ing to get 0.1% closer to what re­al­ity sounds like

now – and all the is­sues have been solved. It’s de­liv­er­ing on the con­cept’s prom­ise.õ

And more re­cently, that de­vel­op­ment was done vir­tu­ally. Ôbe­fore the .uon we had a Ref­er­ence range with a Uni-2 driver and a hy­per­tweeter,õ says Jack.

Back­ing up a Uni-2 with an­other tweeter? Surely that’s al­most an ad­mis­sion of de­feat? Ôthat’s ex­actly how I feel. We learned that you need to put ev­ery­thing in the com­puter model or it won’t work prop­erly. The next Uni-2 could do ev­ery­thing the hy­per­tweeter/ Uni-2 ar­range­ment could do, but in one driver. And it was point-source. That was a turn­ing point for the whole tech­nol­ogy – of say­ing, ‘this can work bet­ter than dis­crete driv­ers if you get ev­ery­thing right’.õ

8a%inç science eau iful

Uni-2 has taken the point-source con­cept to an ex­treme with the Blade se­ries (which has a sin­gle unit on its rounded front, flanked by force­can­celling driv­ers on its ta­pered sides), but Jack stresses that the key is to make prod­ucts that peo­ple enkoy lis­ten­ing to: ÔWE see it as a tech­ni­cal device that’s meant to re­pro­duce a sound go­ing inõ.

But it has to look as good as it sounds, he adds, which is why they col­lab­o­rate with peo­ple like su­per­star in­dus­trial designer Ross -ove­grove. ÔWE worked with him on the .uon and the .uo. :ou can give him a re­ally com­pli­cated brief and he’ll come back with six or seven ideas – com­pletely dif­fer­ent – and they’ll all work. We could use any of them.õ

Ev­ery­thing is en­gi­neered and de­signed at ,E'’s site in ,ent. Ev­ery­thing is tested and de­vel­oped there. Jack and his team are re­spon­si­ble for any­thing that affects the speak­ers’ sound. And when it comes to the high-end speak­ers that the com­pany builds in its small fac­tory, what they say, goes: Ô'or some­thing like a Ref­er­ence or a Blade speaker we can have re­ally strict re­quire­ments for where driv­ers are, shapes, vol­umes and con­struc­tion.õ

Jack shows a pic­ture of an .D' Blade Two pro­to­type they made in the work­shop to prove that it sounded good. ÔIT looked pretty ugly. The brief says which sur­faces need to be smooth, but it doesn’t say what the shape is go­ing to be. This one didn’t have the sub­tle cur­va­ture of the fin­ished model it didn’t fade in to­wards the bot­tom. :ou wouldn’t ask Ne to de­sign it, ba­si­cally Õ De­sign engi­neer Phil (idley is the chap who has to trans­late what Jack and the de­sign­ers pro­duce into some­thing that can ac­tu­ally be man­u­fac­tured. Ôand that goes for ev­ery lit­tle part,õ he says, ro­tat­ing an in­sanely com­pli­cated CAD model of a Blade 2 on his screen. ÔWE can as­sem­ble all these parts to­gether, make sure that things meet to­gether and don’t in­ter­fere.õ

It’s aerospace tech, essentially, adapted for the acous­tics world – and it makes that pro­to­typ­ing process far eas­ier. Ôit’s fairly dif­fi­cult to con­vey a shape like the Blade on pa­per,õ Phil says, adding, Ôthe CAD is very tweak­able, so we can get around the screen and change it un­til we get things right you can even see how the light hits it and so on.õ

They send the CAD off for rapid pro­to­typ­ing (3D-print­ing on steroids) and then dou­ble-check the fit, form and func­tion. Not quite right? Tweak the CAD and re-send.

ÔIN the old days, that would’ve gone straight off to tool­ing and could have meant spending thou­sands of pounds on mod­ding the tool­ing if it wasn’t right. Now we trap that kind of thing. In lots of ways we can cut down the de­sign cy­cle and im­prove the qual­ity of the prod­uct by hav­ing more de­sign it­er­a­tions be­fore go­ing for the fi­nal part,õ says Phil.

It’s es­sen­tial for all sides of the depart­ment to work in har­mony. ÔI can’t do what Phil does, and Phil can’t do what I do,õ says Jack.

T $+ (É $ É +$ É ÉÈÉ È + É + $ È $ ÇÊ

No com­pro­mise I Ys Ço o e Mbbx

0ver in the fac­tory, qual­ity man­ager Dan 'owler is in charge of en­sur­ing all that work is car­ried out per­fectly, ev­ery time. )is team of nine puts to­gether all of ,E'’s high-end prod­ucts: the Ref­er­ence, Blade and .uon speak­ers. Ôthe crew is bril­liant. They’re so com­mit­ted and pas­sion­ate. Each speaker is built in­di­vid­u­ally – (raham tests them, and then a cer­tifi­cate is printed out with the builder’s name on it. ‘Built by .ark Ab­bott’.õ

It’s those per­sonal touches that re­mind you that, while ,E' might be a global brand with some speak­ers cost­ing as much as a house, it’s Ôthere’s a lot of ex­per­tise here. :ou have to have a mul­ti­dis­ci­plinary team: the pro­duc­tion, the sourc­ing, the tool­ing…õ

Phil agrees: ÔI think you end up with a more co­he­sive prod­uct, be­cause it isn’t de­signed by com­mit­tee. If you’ve got one per­son look­ing af­ter that as­pect of the de­sign, I think you tend to end up with a bet­ter prod­uct.õ still a proud, tight-knit team do­ing the graft. .el .atthews, an as­sem­bly tech­ni­cian, might be hand-build­ing Blade crossovers one day, and as­sem­bling driv­ers the next (raham )umphries, the as­sis­tant pro­duc­tion man­ager, painstak­ingly tests up to 0 speak­ers a day against a ref­er­ence line­s­tandard model in a ded­i­cated acous­tic booth Jon 0’Toole, an­other as­sem­bly tech­ni­cian, lav­ishes care on end­less Blade 2 speak­ers as they pass through his workstation.

They’re all pol­ished, checked, buffed, checked again, care­fully packed in situ and then shipped off across the world. Ôthey re­ally care about what they’re build­ing,õ Dan says. Ôthey get frus­trated when they get poor-qual­ity cab­i­nets in, you know, that kind of thing. :ou can see their pas­sion com­ing out when they get an­gry – I un­der­stand it, be­cause they’re com­mit­ted. We don’t mess about try­ing to buff out mis­takes we send it straight back. Ev­ery­thing’s got to be 100 per cent.õ

,E' takes this un­com­pro­mis­ing ap­proach to ev­ery­thing it does: its re­search, its de­sign, its engi­neer­ing, its man­u­fac­ture. It even hand-picks acous­tics grad­u­ates from U, uni­ver­si­ties – grad­u­ates like Jack. Ray­mond Cooke started ,E' in re­sponse to peo­ple set­tling for the imperfect. )is legacy – plain to see in ev­ery depart­ment – has re­sulted in some of the most spec­tac­u­lar speak­ers ever made.

Ôthe com­pany is as good as the peo­ple that work here, and I think what’s in­ter­est­ing is the num­ber of peo­ple who’ve worked here ten years plus,õ says Jack. They’re all af­ter a bit of that magic, it seems. .agic that’s still made in .aid­stone.

E"ery­thing 7EF does is based on hard science. And itys all de"eloped and en­gi­neered in-house in 7ent

freat de­sign goes hand-in-hand with acous­tic engi­neer­ing

Mea­sur­ing im­pulsere­sponse in 7EFYS ded­i­cated lab­o­ra­tory

Newspapers in English

Newspapers from UK

© PressReader. All rights reserved.