The Swingle Singers: The Bach Album. Swing CD5 (1990)

The Swingle Singers: The Bach Album. Swing CD5 (1990), no barcode






Blowing out
Originally published on, 2 June 2011, substantially re-written on 24-25 December 2018

Discographic minutiae: there was a mystery to The Swingle Singers’ “Bach Album”. It was published on the ensemble’s own label, Swing CD5, and is copyrighted 1990. Exactly the same program, augmented of 5 tracks, adding 13 minutes , was published by Virgin on “Bach Hits Back“. The Virgin disc is copyrighted 1994, but the booklet indicates that the recordings were made in May 1991.

It was tempting to consider that Virgin licensed the recording from the Swingle Singers in 1994 (they’ve clearly done so with another Swingle recording, “The Swingle Singers: 1812“, recorded in 1988/89 and first released on Swing CD 4, then reissued by Virgin in 1995). The respective timings didn’t really give a clue: they are generally very near without being identical, and there was no telling whether this is because they were different recordings or simply because the blanks after the music were parametered with different durations.

Still, the reissue hypothesis left open a few questions: why weren’t the five additional Virgin tracks published in the original Swing album? And how to account for the discrepancy of dates – the 1990 copyright of the Swing CD and 1991 recording year indicated by Virgin? To those questions, discographer Thomas Cunniffe on conjectured that May 1991 may have been when The Swingle Singers recorded the five additional tracks. But how to account for some more substantial differences of timing in some of the common tracks than just a couple of seconds, which seemed to contradict the idea of a straight reissue (Brandenburg Concerto No. 3: 2:57 on Swing vs. 2:45 on Virgin, In Dulci Jubilo 1:42 vs. 1:58, Et Resurrexit 2:42 vs. 3:17, Organ Fugue “alla gigue” 2:20 vs. 2:04, Es ist genug 1:37 vs. 1:55)? Was it then that the Swingle Singers re-recorded in May 1991 exactly the same Bach program with five additional track, shortly after doing it for their own label? To what purpose? And why the belated release on Virgin? Their Virgin Mozart album after all was released in 1991, so they had then an ongoing relation with the label. On the other hand, there was also a gap in time before the Virgin reissue of the 1812 album… Was there a re-recording done for Virgin but the publication, for some reason, was delayed?

An upload of a single track from what is said to be The Bach Album on YouTube (the Andante from Sonata No. 2 for unaccompanied Violin, track 9 on Virgin) sure sounded like the same recording as on Virgin. On the other hand, a consideration that would have cast doubt, had I had access to the information, is that in The Bach Album’s booklet, the credits mention a double-bassist and two drummers, while the Virgin CD says “The Swingle Singers produce all the sounds heard on this recording. No instruments are used.” Were we then to cast doubt on the Virgin claim? What I hear for instance, in track 2 of the Virgin CD, the Organ Fugue, or track 4, the Chorale “Sleepers Awake”, certainly sounds like voice imitating double bass; but the percussion (brushes on cymbals) sure sounds genuine. But then, that’s the whole point of a vocal imitation of percussion, brought to its apex in the Swingle Singer’s later album “Beauty and the Beatbox“.

When I first wrote this review, back in 2011, I concluded with “I’m not going to buy this short-timed album just to find out. Even if another recording, I doubt that the interpretations, at so short a distance, would be widely different (and the similar timings would seem to confirm that), and the Virgin CD does offer the five additional tracks: so this is the one to get”.

Well – 7 years later, as I was transferring my Amazon Swingle Singers reviews and discography over to this website, I found it cheap and did buy it. And, folks, here’s breaking news: they are… well… err… I’m still not sure. At least not sure for every track. But at any rate, they are not always different recordings from the Virgin CD. Just different mixings and remasterings.

Take the first track on both CDs, Ein’ feste Burg, sung both times purely a cappella. First thing that strikes the ear: the sonics on Swing are louder and cruder, those on Virgin softer and silkier. Other than that, many details of vocal color and pronunciation certainly sound the same; but then others suggest that these are not the same recordings; but they are so small and subtle, the way an or an is pronounced, that I’m not sure. So I’ve fiddled with both tracks of Ein’ feste Burg on Audacity and ran them simulaneously on four tracks. One is slightly shorter than the other; but ok, that could be purely a “fake” difference in the digital file’s running speed; so I ajusted them to run exactly the same duration; but they may then begin and end simultaneously: the inner unfolding isn’t the exactly the same.  Things seem even more obvious in “Es ist genug“, the end track on both CDs, also a cappella. There, it’s not just a difference of timing and the louder sonics of Swing: the balance of voices is also noticeably different, with men more audible on Swing. Different recordings then by a supremely professional ensemble, whose performances of a same piece will never vary much, right?

But wait. Go now to the Andante from the Violin Sonata BWV 1003 that’s on YouTube. Again, although the Swing track is cut at a significantly higher volume, they ARE very similar in tonal allure, breath control, many details; yet there are differences that would easily make you think that they were two different recordings, by this very professional ensemble: first, the Swing track audibly has real percussion, and the Virgin track audibly not; and again they don’t run exactly at the same timing, the Virgin track is a few seconds longer.

Still, despite those differences, those two tracks were SO tonally similar that I still had a nagging doubt. So I fiddled around again with Audacity, changed the speed of the Swing track to make it match exactly the duration of the Virgin track, played them simultaneously on four tracks.

And, folks: I wouldn’t go as far as to stake my hand on it, but really, I think they’re the same recording. Methinks the Swing CD only overdubbed the percussion to a vocal track recorded separately, and the Virgin CD did without the overdub (and slightly changed the tempo).

Likewise with the Brandenburg Concerto No. 3. In all appearances, this would seem like two different recordings: not just because of the very present percussion on Swing, but also because of the substantial difference in timing, with Virgin more than 10” faster, making it sound, on comparison, like a  more dynamic rendition (although the punch of the percussion on Swing makes up for that).

Brandenburg Concerto tracks, Swing above, Virgin below

Yes but again, with tempo adjusted (8.2 percent faster on the Swing track) so that both tracks match? Percussion and amplification aside, just by ear: They’re. THE. SAME.

Et Resurrexit from the B-minor Mass, sung purely a cappella on The Bach Album, but VERY DIFFERENT from the Virgin version, taken at a slower tempo, making it sound comparatively powerful, but trudging. Obviously different performances…. Really? Adjust tempo so that both match: exactly the same performance, and here, even the amplification is very similar.

Original Virgin track above, Swing below
Both tracks adjusted to same tempo

And I didn’t need to change the tempo by much on the Organ Fugue BWV 542, track 2 (just accelerate the Virgin track by 0.162 percent) to get two identical tracks, except that Swing has a real double bass and percussion and Virgin had the bass vocalizing “dee-dums” and a faint “tchhh-tchhh” that you could easily mistake for real brushed cymbals – and the kick is that, running both tracks simultaneously on Audacity, I get both the real double bass and the vocal dee-dumming. I conjecture again that in the Swing recording the rhythmic section was recorded separately and then mixed in with the recording of the voices, and that for Virgin the vocal percussions were recorded over the former vocal recording.

On the other hand, there’s hardly a doubt on the beginning of In dulci jubilo, track 8a. Vocal color of the opening solo tenor is exactly the same and may raise doubts, but not only are durations, but also rhythms, slightly different – and you can’t fake two quarter-notes into a dotted quarter and eighth by changing tempo -, and the arrangement sung isn’t exactly the same, it’s a slightly shortened version on Swing: unquestionably different recordings. But when the others enter at 0:38, I’m not so sure anymore. Tempo adjusted, both tracks sound (percussion in Swing aside), if not like the same recording, at least like two takes recorded not a year apart, but in the same session.

I haven’t pursued those minute comparisons on all tracks, because, frankly, it is time-consuming and, while there can be a kind of excitement to the sleuth work (remember Antonioni’s “Blow Up” and Brian de Palma’s “Blow out”), it has little to do with musical pleasure. So what are the conclusions of this? Is this another “Joyce Hatto / William Barrington Coupe” affair? (for those not familiar with it, link will open new tab to Wikipedia entry) Well, evidently no, because here the Swingle Singers are not doctoring other performers’ recordingz to pass them off as their own, but their own, maybe to pass them off as new recordings. But more substantially, it would be too simplistic just to call the Virgin reissue “a fake”, because it possibly doctors the tracks already released on The Bach Album. Any record, LP, CD or else, is, in a certain way, a “fake”. Even if it is an unedited live performance that you are hearing on your all-surrounding SACD system, what you are hearing is not coming out of acoustic instruments playing in your living room, but from loudspeakers. It is an imitation, a simulacrum of a live performance, not the reality of a live performance. The word “Hi Fi” and all the similar words coined to pretend that the orchestra is in your living room, are somewhat misleading. High Fidelity, yes, if you understand that it means that the imitation tries to be faithful – but don’t lose sight that it is an imitation, not the real thing, and the flight simulator isn’t the real flight. A record may attempt to be just the documentation, as faithful as possible, of a performance – but it doesn’t have to:  it may be also a sonic object in itself. We fans of classical music may instinctively want to consider that a heavily spliced, edited and doctored recording is “a fraud” – and it is to the extent that, implicitly, it purports to be the documentation of the kind of virtuosity that the performer is capable of “live”. But we may take it as just a record, an artefact  – something in the same spirit as Conlon Nancarrow’s studies for player piano, something that can exist only through artificial and mechanical means, because while the human mind can conceive of it and hanker for it, the human hands (or voice) can’t achieve it. From very early on the recording of some of the arrangements of the Swingle Singers required a lot of overdubbing, because they were written for more than 8 parts; nobody found that a problem. We bought and listened to the albums, not to the documentation of a live concert. “The medium is the message”, as Marshall McLuhan would say. Or, as Gertrude Stein didn’t say, “a record is a record is a record”.

All this to say that it’s best to treat The Bach Album as a different recording from Bach Hits Back, even if the latter is made, in part or in toto, of manipulations of the tracks from the former. More significant than the similarities are the differences: the comparatively louder and possibly cruder sonics on Swing (but I hasten to add that it strikes the ear only when you jump from one recording to the other; also, what is lost in silkness, is gained in analytical precision of perception), the real percussion and double-bass on Swing, harking back to the ensemble’s Paris period (but tracks 1, 5, 6, 10, 11, 12, 15 are purely vocal) vs the vocal ones on Virgin, and the occasional differences of tempo, which transform the Brandenburg movement and the B-minor Mass’ Et Resurrexit into two very different “interpretations”.

Ultimately, the Virgin CD is the one to prefer: it’s been more widely circulated and is easier to find, and it does offer 5 more tracks and 13 more minutes of music. Also, because “The Swingle Singers produce all the sounds heard on this recording. No instruments are used”, it is a more faithful (did you say “high fidelity”?) representation of the style of the Swingle Singers since the late 1980s. The Bach Album is best reserved to Swingle collectors and completists.

For comments on the music itself, see my review of Bach Hits Back.




Comments are welcome