[mhennessy]

I've always felt sorry for the LS5/9. It had a difficult birth. The design was done after Dudley Hardwood retired from R&D, and Designs Department also lost key members of staff. As a result, expertise levels were less than usual, and confidence even lower. I have copies of memos from this period that acknowledge this - unfortunately I'm not allowed to share them - but they show that the new LS5/9 was not welcomed, to put it lightly.

As was the case with the LS5/8, the LS5/9 that we all know bears little resemblance to the prototype that R&D produced. R&D did the right thing; they produced speakers that had substantially flat frequency responses. But after going through Designs Department, we have what we see today...

In the case of the LS5/9, ageing is a real problem. Not as bad as 15 ohm LS3/5As in dB terms, but unluckily for the poor 5/9, the peak that develops is in the 3kHz region, an octave above the ~1.5kHz peak that 15 ohm LS3/5As are (in)famous for. You might get away with a peak at 1.5kHz, but no-one wants a peak at 3kHz! The peak is around 3-4dB in the samples I've measured - but it's the worst place for it. The 1.5dB "Bextrene peak" with LS3/5s can easily be 6dB or more, and it's a higher Q too.

This peak in the LS5/9 is caused by the PVC surround, which gets stiffer with age as it attempts to return to its original shape. You can prove this with nothing more sophisticated than a hair drier!

To get a sensible bass resonant frequency, the stiff PVC surround mandated an unusually "loose" spider, meaning the unit is prone to drift. Rubbing voice coils are all too common. Rogers helped by enlarging the gap (and upping the magnet to compensate), but that just postpones the inevitable. Periodic rotation is recommended.

Another problem with the ageing PVC surround is the effect on the bass tuning. It's actually not too bad compared to other non-BBC speakers, even today, because the LS5/9 lacks the bass boost that the LS5/8 (and LS3/5A to a lesser extent) have, it doesn't have the usual "warm" BBC sound that everyone is used to. If you only work at the BBC and are only experienced with BBC speakers, then it's no surprise that you wouldn't like the 5/9. But those folk who are used to more typical commercial designs don't really comment on the bass IME.

This is one of the things that makes them fussy about placement. If you put them on a solid mass-loaded domestic hi-fi stand of 40cm or so, near to the rear wall, they get back some of the bass warmth that's lacking when they are on tall, wobbly stands in the middle of nowhere. And if you have a decent rug or carpet, plus plenty of soft furnishings, the 3kHz excess becomes much less objectionable. Under those conditions, I'd pick the 5/9 over the 3/5A or the 5/8 - and I know that because I've tried it, as the photos on my website show!

The 5/9 - like the 5/8 - has a HF lift, thanks to the Audax tweeter. It doesn't show on the 5/8 design report, but is there on the 5/9 report, suggesting that the tweeter changed in those early years. Looking at some of my measurements, it's about 6dB. There were 2 versions - the one the BBC picked had a lift at 11kHz, but the other one had a lighter diaphragm and it had a peak at 15kHz. That's the version that's available today, and this is one of the things that Graham Audio had to contend with when designing their version.

The production LS5/8 lacks the slot in front of the woofer that Harwood's prototype had. This makes the off-axis response much worse - you get away with that in a heavily treated control room, but not in more typical domestic environments. So out of the professional environment, the LS5/8 is actually quite hard to make work (quite apart from the bass excess!). Graham have found the same with their version, so have stopped taking it to shows. The LS5/5 is infinitely better in that regard, and their version is probably the ultimate expression of the BBC monitor concept from the Harwood era. Definitely take a listen if you get the chance.

I've only ever seen a single LS5/5A - it was in our telecine area, which was ripped out in 1997 to make way for 5 bays worth of file servers and other IT tatt. I'm told that they weren't well enough matched for stereo, but I can't deny or confirm that. They didn't play loud enough for many in the 1970s, hence the LS5/8. Since hearing the Graham version, I have played around a bit with putting slots in front of drive units, and have been impressed by the results - it's a shame aesthetic concerns stop it being done more often.

We were lucky - we had a separate (small) sound gallery. That changed in 2003 in the new building, sadly. It's very common today, and makes the job extremely difficult. For regional TV, a pair of LS5/9s were absolutely fine. In the mono days, before my time, there was a single LS5/8, but I dread to think how it must have sounded in such a small space. That was the only one on that site, but we had loads here at WN. Mostly gone now - can't say I miss them, as they all struggled to image properly, and often hummed. Changing C5 and C10 on the Quad boards helped a lot. We now have a representative mix of the sorts you'd find across the corporation, and have a lot of fun comparing them and wondering how on earth some of them made it past even the most basic selection exercises! Speaker selection exercises are always controversial because some people have very strong feelings (which they don't always manage to hide on their results sheets, meaning their results have to be discarded, wasting everyone's time) - as a result, people just tend to look around to see what everyone else bought. It's a bit of a mess, as you'd expect. Still, I guess that's the least of the BBC's problems right now...