Jeremy Polmear talks to Malcolm McMillan
on the challenges and rewards of bringing 78rpm recordings back to life
JP: Malcolm, before we talk about the transfer of these Goossens recordings to CD, can I ask you about the recordings themselves, which come from your huge collection of 78rpm discs. How did it all start?
MM: It was when I joined the Royal Ballet Sinfonia almost thirty years ago; playing the ballet in six evenings and two afternoons per week, the question came up as to how I was going to spend my spare daytime hours. Some people go swimming, play football, snooker or tennis, but I was also setting up home at the time, and would go round the junk shops looking for period furniture. As I came across lots of old 78s and a friend had already started collecting, I got hooked. There's also the bonus of getting tax relief on my collection!
JP: But you are a clarinettist - what are you doing with oboe recordings?
MM: I like lots of music. I started collecting Jazz records too until I realised that most of them had already been transferred to LP and, eventually, CD. And I love the old string and piano players - if I come across a good one I'll mention it to friends. But I decided to stick to woodwind and brass, which includes the oboe, and especially Lee Goossens.
JP: Have you got any valuable recordings?
MM: Some of the ones I didn't collect - string players or rare early Jazz - would probably be more valuable, but I have got some very early recordings that span back to about 1896. One example is the 1911 recordings of The London Woodwind Quintet, who were principal players of the original London Symphony Orchestra, which must be pretty valuable.
JP: And how do you play these records? Have you got an old machine?
MM: I used to have a wind-up gramophone, but with about a quarter pound weight pressing down on the record, playing them would wear out both the record and the steel needles, which had to be replaced each time. Then I tried thorn needles, which are real thorns, rose or pyracantha that you can sharpen as they wear. But now I have a special electronic turntable with five diamond styluses of different sizes.
JP: Why five sizes? Are the grooves not all the same?
MM: No they aren't, but that's not the only reason. Sometimes a smaller stylus, which goes further into the groove, will not pick up the surface damage that comes from storing records with no wrapping. On the other hand, sometimes granite from the record itself gets embedded in the very bottom of the grooves and can't be cleaned out, and in that case a larger, truncated stylus would be more successful. Although 78s are mono, of course, I record in stereo, and in some cases one side of the groove will be quieter than the other. In general, though, it's better to mix them together because some of the background noise tends to cancel itself out. It's a case of trial and error.
JP: And what's this granite?
MM: Some companies like Columbia added powdered granite to the shellac mixture to make it more durable, which it did. But it is an abrasive, and contributes to that characteristic background whoosh you hear on 78s. That's why vinyl LPs were so much quieter.
JP: So you pick the best stylus. Are there any other things to look out for when playing a 78rpm disc?
MM: Yes, it might not play at 78 rpm! The actual speeds these records were made at vary from 60 rpm to sometimes as high as 100 rpm. This might be noted on the record, but basically I use my ears to bring the result out at A440. This is not quite a purist approach, as there were at least three pitches in use in England in the early part of the 20th Century, but it's not easy to work out for sure what the original pitch was, and it would create uneasy moments moving from track to track on a CD like this Goossens compilation.
But there are more problems with speed, in that the pitch at the beginning of a track might not be the same as the pitch at the end. Then you've got the whole problem of equalisation, in that old recordings - especially the very old, pre-electric recordings - would pick up and emphasise some frequencies more than others. Plus the 'modern' amplifier has got its own equalisation curve, which has to be allowed for in order to produce a neutral sound.
JP: Sounds like a job for a computer.
MM: Yes, the amplifier plugs into my computer, so I transfer the data on to my hard disk, where I can work on it.
JP: Are there special programs for this kind of thing?
MM: Yes, many. The one I like to use is called Cool Edit Pro, which can handle equalisation and speed changes.
JP: And what about clicks and pops, and general gritty background noise?
MM: Cool Edit has a plug-in accessory called Click Fix, which recognises the characteristic pattern of a click and removes it automatically, and then you take out by hand the ones it misses. As for the general background noise - like mains hum, turntable rumble etc, Cool Edit samples the pattern on a silent part of the record, and then tries to take that pattern away from the other signals.
JP: So it can tell the difference between music and noise?
MM: No it can't, and that's where you have to be careful. Obtrusive noise and music tend to share the same frequencies - especially on the oboe, and it's not because I'm a clarinettist that I say that! So if you go all the way in taking out the background noise you can get the oboe sounding as if bits are missing, and I remember you commenting when we were putting this CD together that sometimes we went too far in taking things out. It's a matter of finding the right balance.
JP: I can see that each record has to be treated individually. What were the hardest tracks to do?
MM: The earliest ones, for example the Colin and the Bax. The Colin was an acoustic recording and the first few grooves were pretty rough. The Bax was an early electric recording, but discs from The National Gramophonic Society tend to be noisy, even the good-looking ones, and I had two copies of this particular recording. There was a lot of granite!
JP: And the easiest?
MM: The oboe and piano tracks with Gerald Moore were fairly straightforward, and of the later recordings the Scarlatti was particularly well recorded. I'm glad I persuaded you to include it in the CD...
JP: Yes, I remember I didn't even bother to listen to it at first, because it looked like a thoroughly inauthentic arrangement. But I remember you saying that it was full of committed music-making, and you were right. We are talking now a few months after the release of the CD, and I can tell you that I am getting a lot of positive feedback. Why do you think people get value from listening to these old recordings?
MM: Well, nostalgia maybe has something to do with it, and perhaps curiosity. There is a danger that today's musicians and music lovers think that the old players were OK in their time, but things have moved on. Yes, technically, they have, but this CD shows that people like Lee Goossens (and Reg Kell on the clarinet) were masters of their instruments, and masters of musicianship - they were like Heifetz and Horowitz, who happened to play oboe and clarinet. They are part of our heritage, and deserve to be celebrated, and enjoyed.
from his extensive 78rpm collection.You can email him at email@example.com
Addition, 2012: Malcolm has subsequently provided and transcribed material for The French Accent, using essentially the same techniques. I asked him whether this set posed any special challenges:
MM: The Poulenc Trio was the hardest, I tried two copies of it. Columbia records were using a lot of granite with the shellac at that time, so there was a lot of surface noise. The balance between the instruments was also not ideal - not enough oboe - and it also changes between movements. Apart from that it was mainly a question of slightly adjusting the tonal qualities of each recording so that the jump from one to another wasn't too extreme.
JP: Looking at the older recordings - the Poulenc and Thuille from the late 1920s, and the Auric, Ferroud and Roesgen-Champion from the mid 1930s; did you see any improvement in recording quality?
MM: Certainly. The discs themselves were produced by the same methods, but there was progress in microphone technology, with the introduction of the ribbon microphone. There were also improvements in the electonics used to process the signal before it went on to the wax master.
After being Principal Clarinet in the Kent Youth Orchestra and an exhibitioner at the Royal College of Music in London, Malcolm McMillan went on to study with Jack Brymer whilst freelancing and teaching in and around Canterbury. It was at this time that he was asked to perform the clarinet concertos by Mozart and Busoni with the Orchestra of Kent Opera and to organise performances of Walton's Facade in Walmer Castle.
He then joined the Royal Opera House in their Royal Ballet Sinfonia, where his work has taken him to most parts of the UK and Europe plus visits to the USA, South Africa and Hong Kong. He has also worked with the Northern Sinfonia, giving him the chance of playing challenging twentieth century music as well as the more standard classical chamber orchestra repertoire.
Some of the highlights of his career were playing for the final appearance of Rudolph Nureyev with Margot Fonteyn at the Edinburgh International Festival, performing Romeo & Juliet in the open air theatre Herod Atticus on sultry mid-summer Athens evenings, and being broadcast playing the jaws-harp in an orchestral piece by Charles Ives!
Malcolm's Current CD transfer projects include Reginald Kell's lighter offerings (complete with sleeve-notes by the late Steve Trier) and Eli Hudson - England's virtuoso flute/piccolo player, who probably made more recordings in the first half of the twentieth century than any other woodwind player.