1971 | Klemperer’s EMI Recording of ‘Cosi Fan Tutte’ in the Kingsway Hall
A detailed look at recording a ‘studio opera’ in the early 70s
In 1971 I was a young sound assistant at London Weekend TV’s subsidiary sound studio ‘Intersound’ and had a growing love of classical music. I had discovered that both Decca and EMI were frequent users of the Methodist Kingsway Hall, just up from The Strand, and as often as I could I dropped into the hall to see if there were any recording sessions I could gatecrash. I realised that I’d have to be ‘self-taught’ with regard to classical music mixing instead of relying on anything passed on from my colleagues at LWT and I highly regarded the records with Decca’s Kenneth Wilkinson and EMI’s Christopher Parker on the credits that were often made in this hall.
On the 15th February, I came across an empty hall with everything set up for an opera recording and managed to easily walk around making notes and even do the same in the control room. This was the last few days of sessions with Klemperer conducting The New Philharmonia in Mozart’s ‘Cosi Fan Tutte’.
I carefully made an orchestra plan at the time showing the layout and microphones in use, and here is a version of that using a computer design programme:
Details of the heights and positioning of the mics can be found on the engineer’s desk layout a bit further down.
Walter Legge’s Philharmonia Orchestra becomes ‘New’
The Producer was Suvi Raj Grubb, an Indian who had come to the UK after working for All-India Radio and had become the assistant to Walter Legge in 1960. Legge virtually controlled the classical music recording side of EMI and he also ‘ran’ his own orchestra, The Philharmonia, giving them much recording work which was most often with Herbert Karajan, Otto Klemperer and more recently Carlo Maria Giulini. EMI wished however to curtail Legge and decided that future recording projects should be more carefully budgeted by their Artists Department. This decision was certainly one of the reasons that Legge announced in March 1964 that he was leaving EMI and also disbanding the Philharmonia Orchestra without any prior notice. The musicians rebelled at so instantly losing their jobs and they decided to rebrand themselves as the ‘New Philharmonia’ and carry on under their own management, and they were backed in this by both Giulini and Klemperer.
Unfortunately, Klemperer was in the middle of sessions for another Mozart opera, ‘The Magic Flute’ when Legge’s surprise announcement appeared and Klemperer took out some of his anger on Raj Grubb, as Legge’s assistant.
By the time these ‘Cosi’ sessions were starting in 1971, Klemperer, now aged 84, had patched up his relationship with Grubb. 
The EMI Control Room set up
Both Decca and EMI had virtually permanent control rooms at the Kingsway Hall, although I can’t remember if it was a shared one and both companies brought in the mixers as required. EMI based its Mobile Unit at the EMI Film Studios at Elstree, with the equipment being brought to the locations by truck and set up for each session.
EMI in this recording were using a new TG12345 console and like most classical sessions were directly recording in stereo only to a pair of Studer A62s tape machines. The tape in use was EMI’s 815 and both the A62 machines had the newly introduced Dolby A361 noise-reduction units. I noted down that the monitor speakers were Tannoy Majors, but I assume I meant Lockwood Majors, which were in common use in studios and always had Tannoy drivers. Decca however certainly were using ‘Tannoy’s’ in their sessions, opting for the triangular York model. In an adjacent room were a pair of Altec Lansing speakers, presumably for artists.
The recorded tape boxes were labelled with the Producer as ‘S.Grubb’ and the Engineers as ‘M.G’ and ‘R.L’, who were the balance engineer Michael Gray and his young tape assistant Richard Lush.
In the control room I took a look at the EMI TG12345 console and here is the engineer Michael Gray’s desk layout:
In the above sheet, I’ve used the designations for each mic that Michael Gray wrote on the scribble strip of the EMI TG12345 console. You can see that he didn’t completely follow the most convenient way of setting them out, as for instance, the ‘main’ orchestral mics, the 3 KM86s suspended over the front of the orchestra, aren’t all next to each other. This presumably would be because the TG12345 console had channels in pairs.
Those 3 ‘main’ KM86 mics are the only mics that I didn’t notate whether they were cardioid or omni. The KM86 was switchable on the mic body and I probably couldn’t tell at the height they were suspended. It would be great to know though; as if ‘omnis’, they would be positioned quite similar to the M50s that Decca used in their ‘Decca Tree’ arrangement. The left and right KM86s are however panned 20 degrees in from full, which is quite a lot and with omni mics would possibly be beginning to give some comb filtering cancellations in mono, so I think Michael Gray was probably using them in cardioid.
The other interesting question concerns the ‘ambience’ that would be added to the mix. The two KM56s up in the balcony looking down on the orchestra would be there for that purpose, particularly if cardioids were being used for those main KM86 mics. Desk channels 17 and 18 are labelled ‘ECHO’, and I can only assume that they are the returns from a stereo EMT140 Echo Plate hidden away somewhere.
All three of the mics above are ‘side-address’, with sideways mounted capsules.
The EMI Mobiles’ TG12345 Consoles
EMI’s first recording console using transistors was the TG12345, designed and built by the EMI Research Labs team led by Mike Batchelor. The expert on these TG12345 consoles nowadays is Brian Gibson who worked for many years as a maintenance engineer at Abbey Road and currently maintains most of the 17 or so surviving consoles. Here’s a photo by Brian Gibson of one of the ex-mobile MKIIs after it was modified by combining a pair of 8-track desks to make a 16-track version:
Classical recording crews have always favoured ‘de-rig control rooms’ at each venue and the MKII consoles for the EMI Mobile Unit were constructed in three sections that could be wheeled into whichever location was being used by EMI presumably like the one in Brian’s photo above.
MIKE HEDGES – Engineer/Producer and a TG12345 owner:
“The Mk.1 was first installed in Studio 2 at Abbey Road and amongst other things recorded the Beatles ‘Abbey Road’ album. When the TG Mk.2 became available it was moved next door into Room 4. Many of the Mk.I Mic cassettes were changed to the newer variation Mk. IIs with more comprehensive EQ. Some of the removed cassettes were used in Abbey’s mobile console and some put into storage and thought lost until recently. A few years later several of the remaining cassettes had their electronics removed as valuable spares for the Mk. II’s (mobiles) and Mk. IV’s (Studio 1 and 2).” 
Above is an ex-mobile desk now preserved at Abbey Road. It has 10 main meters here, for the 8 Group Outputs plus the Main Stereo.
TG12345’s were obviously fitted with varying numbers of those big ‘cassettes’ depending on the requirements and the console I saw at the Kingsway Hall for ‘Cosi’ was fitted for 20 mic channels and 4 Groups. Below is a detail of a TG12345 console’s pan-pots on one of the dual input ‘Microphone Cassettes’. This photo is from a MkIII console and you can see the positions that I’ve referred to in my desk layout document shown earlier:
The use of Painton Faders perhaps wasn’t surprising at the time Bob Batchelor’s guys were designing this desk as although Neve had been using ‘flat faders’ since their earliest germanium transistor desks, they’d had to import them from Germany until they got Penny & Giles to start making some.
However, I’ve found a photo that shows that the Paintons were probably not old 600-ohm ‘stud’ types anymore:
A photo of three Painton Quadrant Faders, presumably all from a TG12345 MkIII and described as being ‘under test’ by the Californian dealer. The sheet states ‘1200 ohm’ which would be correct for the type of early transistor circuitry being used. The one on the left is still a ‘studded’ type, but the other two have continuous tracks. The same dealer also showed a photo of part of the Mic Amp circuit, with both NPN BC109 (silicon) and PNP BCY71 (germanium) in use in the output section.
The Kingsway Hall ‘Cosi’ recording
Both EMI and Decca used the Kingsway Hall frequently for their classical orchestral and opera recordings because of its superb acoustics….and despite the rumble from the London Tube line underneath. EMI had done many previous operas here and so undoubtedly followed a fairly consistent pattern in layout and miking. The vocalists were always set up behind the orchestra on the stage, which was about 4ft high, and it was divided into 5ft squares with white tape, as shown in the earlier drawing. This enabled the singers to consistently keep to previously decided positions for stage movements, but it’s often remarkable how quickly an opera singer can go from one side of the stereo image to the other in recordings!
The mics were fed into the control room behind the stage and all those Neumann mics would require separate mic cables to their power supply units, as would also at that time the single phantom-powered KM84 on the Tymps. The really large boom stands were I think big Neumann models, and a TB speaker and phone were on the stage for communicating with the singers, plus the red cue light and phone were beside Klemperer’s music stand, which incidentally was padded, as can also be seen in this photo.
Asked in 1970 why he so much prefers having his second violins on his right, Klemperer replied:
“They play a big role in symphonic music. They are very seldom in unison with the firsts and therefore they must be independent. On the other hand, I think it is important for the cellos and basses to be together, rather than about as far separated as can be“. Another unmistakable feature of a Klemperer performance is the clean articulation of the strings. “It is most important for a conductor to indicate the bowing and dynamic marks he wants in the parts. I do this in my score and then the Philharmonia librarian transfers these to all the parts. To conduct from old parts is really a punishment.“
These days, he finds the strain of recording very tiring, yet at 85 his enthusiasm to put on disc many more interpretations is only held in check by his daughter’s understandable concern for his continuing good health. ” 
The photo above shows Producer Grubb at the console conducting a playback for the singers and it’s always worrying to see drinks on the front of the console faders, and with those scores alongside, I hope Mike Gray had marked his fader positions.
On EMI sessions the Producer usually sat behind the Engineer who occupied the ‘stereo seat’ at the desk, so the music stand in front of Yvonne Minton here is probably where Grubb would normally be, whereas Decca had the Producer alongside the Engineer. Decca also used two Engineers on an opera, one for the orchestra and one for the vocals.
Opera recording has now become just too expensive to undertake in studio conditions and in the ‘Recording Schedule’ below that Suvi Raj Grubb produced shows that there were a total of 16 ‘Cosi’ sessions, and illustrates how an opera had to be broken up to accommodate the artists:
How did it sound?
The schedule shows that, as expected the recordings started with the Overture to establish the orchestral balance and it was to be repeated on the second session as well.
Here’s how it came out. This though is from a CD issued in 2012, which it states was ‘digitally remastered in 1991’ at Abbey Road. That was during the period when most record companies were madly digitising their back catalogues to reissue on CD, and there wasn’t always that much care in doing that, with little remaking of the tape edits for instance:
AUDIO EXCERPT ONE:
Klemperer is frequently criticised for the slow tempos he adopted as he got older but he takes the Overture at a fair lick. The sound balance seems good to me but it’s a bit ‘thin and bright’ and definitely lacking in the bottom end.
To hear some of the vocal balance against the orchestra, next is the well-known trio from Scene 2, ‘Soave sia vento’. It is No.10 ‘Terzettino’ in the score, and the schedule shows us it was recorded on February 9th. The singers are Yvonne Minton (Dorabella), Hans Sotin (Don Alfonso) and Margaret Price (Fiordiligi).
AUDIO EXCERPT TWO:
Some of the CD is slightly ‘edgy’ sounding, that’s with some peak distortion creeping in. EMI didn’t use ‘peak reading meters’ like Decca but VUs but with Dolby ‘A’ on those tapes, Michael Gray should have been able to avoid tape distortion surely. It’s more noticeable on this next solo vocal excerpt and also on the louder strings, which sound surprisingly bright considering there appeared to be no EQ at the time of recording.
Finally, the Aria No14 in the score ‘Come scoglio immoto resta’ with Margaret Price (Fiordiligi):
AUDIO EXCERPT THREE:
How was it received?
Klemperer’s Mozart operas were often considered to be ‘too slow’ overall but undoubtedly had some really good sections that some have praised:
“But the ladies, the ladies! What a glorious trio of voices, unencumbered by breathless tempi;
Klemperer’s leisurely but loving phrasing gives them the space to soar like eagles over Mozart’s
melodic landscape. The story goes that in reply to Lucia Popp’s complaint that his tempi were too
slow, the libidinous Klemp replied, ‘When I make love, it has to be very slow!”. This certainly applies
to “Come scoglio” but when Margaret Price sings it like a goddess you want it to last as long as
Despite its attendant frustrations and weaknesses, this set remains in my shelves purely so that I can
hear the best trio of female singers ever to sing Così – albeit rather slowly…”
The reviewer in the New York Times was however fairly critical of Klemperer’s tempi and of his singers but managed to slightly favour the ladies’ singing:
“Margaret Price, a young talent definitely on the way up, creates a Fiordiligi far colder than she need be. Her two big arias are strongly sung but they chill rather than inspire, Yvonne Minton’s Dorabella seems far more human and vocally she does not let down the sister act.
Actually, the serious, meditative, sometimes dreamlike side of “Cosi fan tutte” is responsible for this performance’s most pleasurable musical moments. The “Soave sia it vento” trio goes beautifully and delicately, as does Fiordiligi’s surrender duet with Ferrando and, perhaps best of all, the “E nel tuo, nel mio bicchiero” toast just before the Act 2 denouement with Miss Price’s powerful yet never heavy soprano riding the crest of the ensemble.
No, I’m afraid that, unlike “Figaro” it doesn’t make it. Yet, for all its strangeness, it is a painstakingly prepared, meticulously executed performance. And, alas, it is stillborn.“ 
CREDITS and REFERENCES
 More details of the relationship he had with both Walter Legge and Otto Klemperer and the change that brought about The New Philharmonia can be found in Suvi Raj Grubb’s fascinating biography “Music Makers On Record” published by Hamish Hamilton in 1986.
Grubb’s immediate boss, Peter Andry also wrote about how he was involved in Klemperer’s ‘Magic Flute’ opera record and the drama of Philharmonia change at the time of Legge’s departure in his biography “Inside The Recording Studio”. My copy came from Scarecrow Press 2008.
 Mike Hedges’s quote was from the Gearspace Forum discussion about an ex-Abbey Road TG12345 MkIV console when it was sold for a record-breaking figure in 2017.
 Klemperer’s quotes from an interview with Alan Blyth in The Gramophone of May 1970.
 From a review of Klemperer’s ‘Figaro’ and ‘Cosi’ in The New York Times of July 1st 1973 by Harvey E. Philips.