Tuesday, August 06, 2013

Why multi-channel music is a commercial conspiracy

Our ears tell us that multiple microphone set-ups compromise sound quality in classical recordings, and that, conversely, simple microphone arrays enhance sound quality. So, extrapolating that finding to the extreme, the best sound quality will come from a single microphone, which means, of course, mono sound. In a recent post David Derrick celebrates the sound of Thibaud, Casals and Cortot playing Beethoven’s Archduke Trio captured in mono by EMI in 1928, while the 1955 Decca recording of Blochs magnificent and neglected string quartets by the Grillers seen above provides further proof of the sheer musicality of mono sound.

What is striking listening to those Decca discs of the Bloch quartets is the solidity of the sound. It is ironic that the word 'stereo' comes from a Greek root meaning 'solid', yet today stereo and other multi-channel formats all too often trade solidity of sound for technical complexity. Stereo and multi-channel sound - multi-channel defined as more than two channels - crafted by a first-rate engineer and producer brings tangible benefits; these include the illusion of being present at the live performance by creating a plausible sound stage and untangling complex musical strands. But the opportunity cost of the mediocre sound so often heard in contemporary classical recordings and broadcasts is ignored. It is not just the solidity and presence of the sound that is compromised: the importance of bass and the associated 'slam', and of perceived loudness in attracting new and younger audiences has been discussed here many times, and phase errors and other nasties created by multiple microphone set-ups and by multiple speaker reproduction are the enemy of bass, slam and loudness.

Recently I quoted Sir Adrian Boult as saying "It isn't at all necessary to sit in front; in fact I prefer the back..." At the back in a concert hall, where young listeners sit, there is no discernible sound stage, so you do not need more than one channel to create the illusion of the orchestra being in front of you. Yes, there is hall ambience, but that can still be reproduced by a mono system. As can that much sought after low frequency slam which is non-directional - that is why the new generation of multi-channel home cinema systems have only one low frequency sub-woofer.

Nobody talks about mono today, but it wasn't always like that. In my days at the BBC, radio mixing desks had a mono switch which allowed the stereo balance to be checked for compatibility with mono receivers. Some of the earliest - and best - stereo microphone set-ups such as the legendary ORTF array were specifically designed for mono compatibility. Many very early stereo recordings sound better than their contemporary counterparts not only because they used just a few microphones, but also because they were mixed to sound good in mono as well. But as stereo systems replaced mono record players in the 1960s, mono compatibility was forgotten. Then in the 1970s quadraphonic multi-channel sound arrived, and the rot really set in.

History shows that multi-channel sound is a commercial conspiracy. Above is the inner liner for an LP encoded in the abortive multi-channel SQ quadraphonic format developed by CBS and EMI to compete with JVC's misleadingly titled CD-4 multi-channel LP format and Sansui's QS format in the 1970s. Below is the sleeve for the notorious 1974 CBS SQ LP of Bernstein conducting The Rite of Spring that placed the listener in the middle of the orchestra. Predictably the BBC jumped on the quadraphonic bandwagon in 1974 with an experimental broadcast using their Quadrimono system. This, quite unbelievably, required listeners to have two stereo receivers in addition to a four channel system. Topically a Proms performance of Stravinsky' Oedipus Rex was one of the recordings used in the experimental BBC broadcast; the results were judged as "absolutely dreadful...scrappy, scratchy violins and feeble percussion... image very unstable, with positive localisation only at the corners – a moving image jumped alarmingly from one speaker to the next with no progression".

Despite the commercial backing of global electronics manufacturers and leading record labels multi-channel sound was a dismal failure in the 1970s. Confused consumers resisted investing in the required decoder and an additional amplifier and speakers. The competing systems were the brainchildren of a misguided alliance of backroom boffins and smooth salesman, while the vital people in the middle, the record producers and engineers were deeply cynical - while I was at EMI quadraphonic sessions were monitored and balanced in stereo only, with the rear channel information captured on tape for a later quad remix by studio staff, not the producer. Then there was the small problem that matrix four channel sound did not really work: the separation between front and rear channels was as low as 3dB meaning that the results sounded like double stereo. There is more on EMI's involvement with SQ Quadraphonic sound in More maestros, myths and madness.

The popular perception of mono sound is moulded by the technical compromises - restricted frequency range etc - of vintage recordings. But hearing a high quality mono recording such as the Decca Bloch quartets played on a high end audio system is a revelation. (Note the recording must come from a mono master; combining the two stereo channels of a CD player to give a mono source simply introduces phase cancellation at an earlier stage). There is a dawning understanding that technical complexity is a mixed blessing, which means 'retro' technologies such as vinyl are making a comeback. I am not seriously suggesting a return to mono. But it is worth reflecting on how much the development of multichannel music has been driven by an altruistic quest for better sound, and how much by commercial interests in the form of audio manufacturers trying to sell yet more complex and costly equipment and record companies trying to sell yet another 'improved' music format.

This has been very much a personal view of stereo and multi-channel music. While my personal listening ranges from mono to the latest stereo recordings I have had less exposure to contemporary multi-channel formats such as SACD 5.1, although my main replay system does use the excellent Arcam two channel SACD CD 37 player. Contemporary multi-channel systems with their discreet audio channels avoid the phasing and separation problems that bedevilled the matrix systems of the 1970s. But in an email discussion about the benefits, or otherwise, of SACD - yet another 'improved' music format - Andy Moore, one of the senior engineers at Arcam who developed the CD 37 made this telling comment: "Again unless properly control SACD may sound less than overwhelming, and I learnt many years ago that the technical prowess of SACD and DVDA does not make up for poor mastering or production/recording techniques".

Despite complexity being the enemy of fidelity mono is unlikely to make a comeback. But there is still a lot we can learn from it: today's technically contrived classical recordings and broadcasts lack the immediacy and impact of their mono and early stereo counterparts such as the legendary 1959 Mercury recording of Antal Dorati conducting The Firebird. And adding that elusive immediacy and impact back into new recordings may generate much needed sales and win a new classical audience. As an example of the power of mono, Glenn Gould's mono 1955 Goldberg Variations sold more than forty thousand copies in its first four years in the catalogue and more than one hundred thousand during Gould's lifetime - sales that today's technology empowered record companies would die for. There is a more detailed discussion of sound reproduction in Music to listener of listener to music?. The superb Griller recording was produced for Decca by the late Peter Andry who later moved to EMI; my memories of him are in No flowers please for Herbert von Karjan.

* I may believe that multi-channel music is a commercial conspiracy and my main listening room contains just two Bowers & Wilkins Nautilus 803 speakers. But our den retains a two channel four speaker system that I built in the 1970s using the now forgotten Hafler 'sum and difference' matrix to extract out-of-phase ambient information from stereo recordings. And very good it sounds too on the increasingly rare broadcasts and recordings that are not close-miked.

** The tribute website to multi-channel audio pioneer Michael Gerzon is one of the best sources on this subject.

Also on Facebook and Twitter. Any copyrighted material on these pages is included as "fair use", for the purpose of study, review or critical analysis only, and will be removed at the request of copyright owner(s).


Lyle Sanford, RMT said...

Thanks so much for this post, and all the other audio ones you've done that it builds on. If anyone else is talking in such detail about the nature of audio quality of recordings and how that affects the listening experience, I haven't found it. It seems profoundly strange that something so crucial to the art and business of classical music only gets talked about on a single blog.

I'll never have your "golden ears", so the suggestion to go for the "slam" instead of trying to recreate the full complexity of the sound is a very handy take-away when it comes to mics and mixing.

Pliable said...

Lyle, many thanks for that. Yes, isn't it strange how little attention is paid to the quality of sound in classical music? Particularly with so much talk of winning over rock audiences, a group that can surely relate to to the sheer visceral quality of, for instance, a Mahler symphony. But all too often today that visceral quality is diluted in the recording process.

Play a rock fan EMI's 1969 recording of Sir John Barbirolli conducting Mahler's Fifth Symphony and you have a classical fan. Play them many recent recordings of the same work and you have a puzzled look.

Pliable said...

Looking at the Overgrown Path traffic log it is good to see readership for this post running ahead of the not inconsiderable readership for those depressing Sinfini Music posts.

It appears the sound does still matter - http://www.overgrownpath.com/2008/11/does-sound-matter-anymore.html

Michael Strickland said...

When I bought my first record player in the 1960s as a 13-year-old, mono was just being phased out in record stores which meant they were usually being sold at fire sale prices which made them affordable for my boy budget. My little stereo played both mono and stereo recordings, and I soon realized that there was little appreciable improvement in sound between the two formats. For instance, Bruno Walter's mono set of the Brahms symphonies is completely superior in every respect to a stereo set he conducted a decade later. Nice to be reminded that I wasn't imagining it.

Philip Amos said...

When I gave this marvellous post its first read, I was so very tired that I realized at the end of the second paragraph that my mind, agreeing with such great enthusiasm with your words about the Griller recording, had in fact at some point confused both composer and performers with another recording altogether. What had crept into my mind was the Vegh Quartet's complete Beethoven Quartets recorded for the Haydn Society in 1952. My first hearing of those in the Music and Arts reissue a couple of years ago struck me very forcibly re sound quality.

I think in the post it was your use of the word 'solidity' that led me to transpose the Griller and Vegh recordings. It's been a while, but even now I can recollect that 'solidity' would have perfectly fit one of my first reactions to the sound. That quality was of particular importance to the Vegh, who were much intent upon a blended sound. At the same time, the cello is as firm a bass line as one could wish for, and in many parts of the cycle there is a toughness, a degree of aggression, entirely appropriate in Beethoven (and perhaps not surprising given who the first violin is!). Call that 'slam', and you have solidity, bass and slam, and I must think the reason, or one reason, why the 1952 recording is so much better than their remake exactly twenty years later.

Keith said...

"At the back in a concert hall, where young listeners sit, there is no discernible sound stage, so you do not need more than one channel to create the illusion of the orchestra being in front of you."

I recommend the very top 'shelf' on audience left at Birmingham Symphony Hall myself for the really large forces. You can opt to look at the orchestra in full flow, or to chill out by reclining. Not so good for those with vertigo!

Oddly enough, I have just moved the sitting room around and have the mini-stereo bookshelf speakers about 3ft apart on a cupboard about 4ft high at the moment while decorating. Toe them out 45 degrees and it sounds fine. (Gould's Art of Fugue on the organ has plenty of 'slap'). A local jazz musician is hanging on to his radiogram specifically for playing his BlueNotes.

Here is an experiment I tried 40 years ago: single sided amplifier using a 6L6 in pentode connection with RC coupled 6J5 salvaged from old radios. Dad wasn't letting me play with full HT so only 200v. But variable negative feedback via a potentiometer. High feedback: 'creamy' but boring. Low feedback, 'alive' but obvious distortion. Magic.