Keith Howard

It's said that hi-fi, and especially loudspeaker performance, owes as much to art as to science, but audio journalism has all too often lost sight of the science completely. No-one can have done more to redress the balance than Hi-Fi News consultant technical editor Keith Howard.

"I'm fortunate to live in a six-bedroom detached Victorian house with high ceilings," says Keith. The listening room is the largest on the ground floor

For Keith Howard, it all started at school in the 1970s, where he wasn't the only one caught up in the quest for better sound. 'That was the zeitgeist. And lots of people were assembling systems from Sinclair kits or this, that and the other. Then I struggled on, as a student, with some pretty crummy systems. I bought a Thorens TD124 at some stage, I think I was using a Hadcock arm with it, can't remember the cartridge.

'I went through a period when I bought up old valve amplifiers. When I'd just started as a journalist, I was just living in a garret in Brixton, but I'd travel by train anywhere in striking distance in answer to adverts from people trying to sell old valve equipment. I'd pick up Quad IIs in particular, bring them back and try to get them working again.

'As part of that process I also bought four Quad ELSs, and put them together in a stack. I never got on with them. Disassembled the stack and sold them off!

'I came into journalism, in late 1977, by accident. I'd finished my first degree in biophysics at Leeds that summer. I thought I was going on to do a PhD at Birmingham University, but I'd applied too late, so I was forced to take a year out. And I was still working as a labourer for a stonemason's firm on Corpus Christi College in Oxford, shortly before I dropped a hundredweight of stone on my right big toe.'

Then Keith saw a recruitment ad in New Scientist magazine. 'It was just a simple few lines that said "Hi-Fi: is it your subject?" in capital letters. And I thought, well I've got to fill in the rest of this year and I don't want to be a labourer through the winter months. I was interviewed and obviously said the right things, because I became editorial assistant on Popular Hi-Fi. After a few months I was moved over to Hi-Fi Answers, as assistant to the late Paul Benson.

'But the position for me to do the PhD at Birmingham was now open, so I left in the autumn. I actually did an MSc in neurocommunications as the PhD didn't work out.

'I was still doing feature articles for Hi-Fi Answers, so I kept in contact with the magazine. It became clear that David Prakel was going to leave the job that he'd taken over from me. I was told that if I wanted it, I could have it.

'I decided that that was it for academe. Two years to the day after I'd left the first time, I went back to Hi-Fi Answers. My timing couldn't have been better. A few months later, in early 1981, Paul Benson left and I was promoted to editor.

'I was editor for almost nine years. I then became a freelancer, really because I liked being a journalist and didn't want to become a manager, and climb the corporate ladder.'

After Keith left, Hi-Fi Answers was relaunched as Audiophile, only to fold in 1994. But then he was offered the role of technical editor on a new magazine called What Home Entertainment.

'Indeed I did become technical editor, but the magazine only lasted seven issues. The only positive thing to come out of that was that I'd negotiated a three-months' notice period, so when they closed rather suddenly they needed to give me three months' money. And part of that was paid not as cash but in the form of the MLSSA measurement system they'd bought.

'So I had the MLSSA software and the MLSSA card. I bought myself a measurement microphone and I was in the business of being able to measure loudspeakers.

'Shortly after that, in 1995, I became audio consulting editor of Gramophone. This lasted about four years until Gramophone was bought by a large publisher. I didn't see that I had a future with it under those circumstances, so I left to write for Hi-Fi News. I felt that I'd been there and done that as far as equipment reviewing was concerned, and I was quite pleased to have the opportunity mainly to write features. But eventually I started to do all the loudspeaker measurements for HFN.'

At his workbench Keith has built a variety of circuits down the years to assist with feature articles for Hi-Fi News

'On Hi-Fi Answers, where we didn't measure anything, the argument had been: if measurements were really relevant, you'd be able to extrapolate from the measurements to the sound that the product makes. I came to change that point of view a bit, needless to say. But I'd never say that you can look at the measurements I take of a loudspeaker and know what it will sound like, because you don't. You can see what the tonal balance will be, and that's pretty much it.

'So loudspeaker measurements can only tell you so much. But that doesn't mean you shouldn't be making them. One very good reason is that you need to be sure that you're not being hoodwinked by a product, which you perceive you like but actually has got a big problem.

'Not only that, it's really only fair to the readers and to the product to make sure that it's functioning properly. It is anything but unknown for products to arrive that aren't working as they should.'

The measurement computer, which lives in the listening room, runs ARTA and LIMP software, plus other code Keith has written himself. The GRAS 40BE measurement mic has a 4Hz–100kHz bandwidth. At the time of our visit Grimm Audio's active LS1 was awaiting measurement. Its frequency response turned out to be the flattest Keith had ever recorded

'Journalists are often accused of wanting to hear differences where they don't exist, wanting to exaggerate differences between products – so that they can carry on flogging their expertise, whether real or imagined, to the public.

'I really don't think that happens. And as far as I'm concerned, the differences between products that I write about are real and anybody can hear them if they want to make an effort, and if they prize the sort of differences we're talking about. I don't claim to be golden-eared, but I'm experienced at listening and I know what I want from music, and that equips me to review products.

'I think those are the key issues. You've got to have a clear idea of what you're expecting hi-fi to sound like. That means that people who read you have some chance of assessing what it is you're listening for. And deciding, yes, they're on the same wavelength as you and therefore will take some notice of what you write.'

So how has Keith's own system progressed through the years of writing and reviewing?

'Like many journalists,' he says, 'I can be lazy about owning equipment, because quite a lot has been coming through the door. There must have been occasions in my life when I couldn't assemble a complete system of my own.

'Pretty quickly, after becoming a journalist the second time around, I bought myself a Linn LP12, because that's what you did in those days! I had a Mission arm on it.

'I had an eye-opening experience the first time I heard a Koetsu cartridge. I had a Koetsu Black on loan for review in Hi-Fi Answers. It went back to the distributor on a Friday. On the following Monday, even though I couldn't really afford it, I phoned Absolute Sounds to place an order, because a weekend without the Koetsu had been too much to bear. But those sort of experiences have been relatively few and far between. And now I've left the source component world behind, inasmuch as I source audio from computer all the time.

'I don't remember exactly what happened to the Linn. I also had a Townshend Rock at one stage, which I had to sell in the course of a bout of freelance penury (I think I had to pay the taxman). I do still have a turntable, a Roksan Xerxes X, but I haven't used it for years. And frankly, I feel no inclination to use it now.

'It's now feasible along with downloads to have hi'res material that's ripped from DVD'v and DVD-A. So the only reason left to have a player as a separate component is to play SACD, because that's the one thing you can't rip. Frustratingly!

'Why would I want to spin an optical disc and read data from it in real time when I can transfer that data to hard disk or, even better, read it from a USB flash drive?'

'This is an example of how the old tweakiness, so prevalent in the '80s carries over to computer audio. I find, and I know that lots of other people have had similar experiences, that I get the best sound quality if I play, firstly, uncompresssed files. In other words, if I buy a FLAC file, I don't play the FLAC file from the computer. I decode the FLAC file and play the WAV file. I find that sounds better!

'Now, that seems bizarre, because we get back to the old business of "data is data", the data in the losslessly-compressed FLAC file is the same as the data in the WAV file. But it seems to be, getting back to familiar power supply concerns, that if you are doing more computing than you need to do, in other words if you are unpacking the FLAC file in real time while you're playing it, you're making the processor and the power supply work harder, and the sound suffers.

'Whenever I'm reviewing I use a USB memory stick. In fact whenever I want the best sound quality (which is effectively all the time!) I'll copy the track over to a USB stick because it sounds better that way. But the bottom line for me is that if you take care over these details, computer audio can deliver extremely good sound.'

The Mac mini  (left, middle shelf)is the music source but runs Windows XP and J River Media Center v15 as the player software. Beneath it is the RME Fireface 800 multichannel interface, which Keith uses for speaker development. He writes software to pre-process stereo into multichannel WAV files, one channel per speaker per driver

Returning to conventional electronics, we asked Keith about his choice of amplifiers.

'I stumbled across the Exposure XVIII Monos by accident. I knew they had a good reputation, of course, but it wasn't until an ex-Gramophone colleague was looking to sell a pair that I decided to give them a go – and I was glad I did. The chap in question later sold me his second pair so that I can bi-amp two-way speakers. They are quintessentially musical – by which I mean they're informative and fun to listen to, and those two qualities are sine qua non for me.'

And his loudspeakers?

'I don't own the Thiel CS1.6s, but they have acted as my stable reference point for speakers for many years. They've been on long'term loan ever since I reviewed them. As I said when I wrote them up, they have a marvellous midrange lucidity – which explains why I've been happy to spend so much time with them.

'If you make a change to the system, even if it's small, the Thiels are transparent enough to let you know whether it's an improvement or not. Many more expensive speakers that come my way during my measurement duties for HFN fail that crucial test. If hi-fi equipment isn't informative about how a piece of music is played and how it has been recorded then it isn't hi-fi for me.

'I'm not interested in sugar coating, I do not believe reproduced music should always sound "nice-. How dull that would be.'

Today Keith has a sonically excellent listening room, and he has much to impart on the 'final frontier- of hi-fi, the room itself and its interaction with the loudspeakers.

'I think there are certain basic rules, such as keep clutter away from your speakers. The American approach to hi-fi is to have a stack of equipment, what's often referred to as the altar, between the speakers. OK, it's practical in many ways, but having anything large and reflective between the loudspeakers is a bad thing.

'And if you're going to apply any absorbent, or any attempt at diffusion, then I would say, don't stick the absorbent or the diffusor at the point on the side wall where the first reflection comes from. I came to that conclusion very many years ago. I bought some 50mm acoustic foam and put it in the places where the first-order reflections would come from on the side walls, and on the back wall. I proudly called in a colleague. He listened for half a minute and said "I think it sounded better before!'"

It seems to be as if, somehow, the reflection gives you a second bite at the cherry. In crude terms you're hearing the sound twice. The important thing is that this second bite at the cherry should be spectrally messed about with as little as possible. That's why you don't want to put absorbent in there because conventional loudspeakers become more directional with rising frequency anyway so the off-axis sound which is bouncing off the side walls is already spectrally different from the direct sound. All practical absorbents absorb more in the treble than they do at lower frequencies, so you're making that situation even worse.'

Measurement equipment sits cheek by jowl with kit used for reviewing. Upstairs in his office Keith has an extensive collection of audio reference books and papers – and a few car books too!

Keith would certainly assert that hi-fi equipment continues to get better.

'Yes. A lot has improved. I should say in this context that I've never been somebody who wears rose'tinted glasses when it comes to audio's past. I don't hanker after "the old days", and I don't think products in the past had a certain magic that products lack now. I think that generally speaking, products have got a great deal better and one of the reasons they've got better is because computer-aided design techniques have become available.

'But what makes this whole subject so interesting, is that this hasn't got us to a situation where everything sounds the same. We still have an awful lot to learn.'

Keith's listening room is a working environment so has to be capable of being quickly adapted into a loudspeaker measuring space – which is achieved simply by pushing back the sofa, as here. In the photo at the top of the article he is in the process of doing a near-field bass measurement on one of the speakers in a recent group test – others await their turn against the wall. The room is acoustically treated, but in a light'touch way that controls the reverberant sound and suppresses flutter echoes without interfering with first-order reflections. It's a good sounding room.