BeatlesAnswers.org

Only a Northern Song

(Harrison)

Find on Amazon
Page status: In review

On this page

First lyric line — “If you're listening to this song…” (brief identification excerpt; full lyrics © Harrisongs Ltd / Sony Music Publishing.)

Story Section status: Outdated

“Only a Northern Song” is a song by the English rock band The Beatles, written by George and released on the 1969 soundtrack album Yellow Submarine. It was recorded during the sessions for Sgt. Pepper’s Lonely Hearts Club Band but was left off that album.

The title plays on George’s standing with Northern Songs Ltd, the publishing company built around the Lennon–McCartney catalogue, in which he held no comparable stake. Within the catalogue it sits with George’s other compositions, Within You Without You and Love You To.

The session work falls within the band’s Sgt. Pepper’s (1967) period. The basic track was taped on 13 February 1967 at EMI Studios, Studio Two, under the working title ‘Not Known’. George Martin produced; Geoff Emerick engineered. Bass guitar, a trumpet and a glockenspiel were added by overdub, and the song was mixed for mono in April 1967 before being set aside.

What’s distinctive

Two February nights inside the Sgt. Pepper sessions, logged under the working title ‘Not Known’, then nine weeks of silence — and when the song was picked up again in April the existing overdubs were wiped and replaced with bass guitar, a trumpet and a glockenspiel. It was mixed for mono and never for stereo, and it reached the public two years later on a soundtrack album.1,3

Recording

  • Two nights in February, then nine weeks of nothing. The basic track was taped in Studio Two on Monday 13 February 1967, 7.00pm–3.30am, as takes 1–9 — logged not under a provisional title but under the placeholder ‘Not Known’.1 The band came back the following night, 14 February, same room and the same 7.00pm–3.30am booking, reduced take 3 and superimposed onto it.1 Then the song stopped. Nothing further was done to it until 19 April, by which time the album it had been recorded for was finished. A song that goes quiet for nine weeks in the middle of the most intensively worked sessions of the band’s career is telling you something about where it stood.
  • What the setup sheet shows. Geoff Emerick’s setup sheet for the 13 February session survives, and it is the rare case of a per-song microphone plot rather than an era generalisation.2 The drums took what had become his standard arrangement — a D20 on the bass drum, D19c microphones on toms, hi-hat and overhead, and a KM56 below the snare; the guitar amplifiers were mic’d with U67s; and an AKG C12 stood at the Steinway.2 For the vocal he asked for a U48 and a KM56.2
  • The microphone that was given up for a box. The vocal KM56 did not survive the session plan: it was sacrificed to free an input for the RS144 premix box being used with the drums.2 That is the whole economics of a four-track control room in one decision. The desk has a fixed number of inputs, a drum kit under five microphones needs somewhere to be summed before it reaches a track, and the premix box that does the summing has to be fed from the same finite supply. Something on the list loses, and on this night it was the second vocal microphone.
  • Take 3 stayed the working generation. Nine takes were recorded on the first night, but it was take 3 that the session kept, reducing it on 14 February and superimposing onto it.1 Of the takes that followed the reduction, take 10 dissolved — understandably — into laughter, take 11 was good, and take 12 was the best.1 The song was carried across two tapes running in synchronisation, and the mono mix was made from both, mainly from take 12.3
  • April: the overdubs were wiped, not added to. When the song was picked up again on 20 April the existing material was wiped and bass guitar, a trumpet and a glockenspiel superimposed in its place.1 This is the decisive editorial act on the record and it is a subtraction before it is an addition — the version the band left in February no longer exists on tape. What survives is a second thought recorded over a first, which is why the finished track’s brass and tuned percussion sit against the arrangement rather than inside it.
  • Mixed for mono, and only for mono. Mono remix 4 was pulled from take 12 on 19 April, the day before the new overdubs; the whole numbering then restarted, and remixes 1–11 were made from takes 3 and 11 on 21 April.1 No stereo mix was attempted at all.3 The mono is therefore not the preferred mix of this song — for eighteen months it was the only one.

Equipment

StudioEMI Studios, Abbey Road — Studio Two1
DrumsD20 on bass drum; D19c on toms, hi-hat and overhead; KM56 below the snare2
Drum premixRS144 premix box2
PianoAKG C12 at the Steinway2
Guitar ampsNeumann U672
VocalNeumann U48 (a KM56 was requested and given up to the drum premix)2

Recording Timeline

  • Monday 13 February 1967 — Studio Two, 7.00pm–3.30am: ‘Not Known’, takes 1–9. P: George Martin. E: Geoff Emerick. 2E: Richard Lush.1 The same booking mono-mixed ‘A Day In The Life’, remixes 2–5 from take 7.1 Emerick’s setup sheet for this date records the microphone layout: a D20 on the bass drum, D19c microphones on toms, hi-hat and overhead, a KM56 under the snare feeding an RS144 premix box, U67s on the guitar amplifiers and an AKG C12 at the Steinway.2 Sharing a night with the mono mixing of the album’s closing track tells you exactly where this song sat in the queue.
  • Tuesday 14 February 1967 — Studio Two, 7.00pm–3.30am: tape reduction of take 3, superimposition onto take 3, mono mixing.1 Take 10 collapsed into laughter, take 11 was good and take 12 was judged the best of them.1 A reduction on the second night is early — it means the four tracks were already full after one session, and every later idea would have to be recorded over something rather than beside it.
  • Wednesday 19 April 1967 — Studio Two (control room only), 7.00pm–12.30am: mono remix 4 from take 12. P: George Martin. E: Geoff Emerick. 2E: Richard Lush.1 The bulk of the night went to ‘Good Morning Good Morning’, remixes 10–23 from take 11.1 This mix was made the day before the overdubs were replaced, so it captures a version of the record that stopped existing twenty-four hours later.
  • Thursday 20 April 1967 — Studio Two, 7.00pm–2.15am: superimposition onto take 3 and onto take 11.1 The existing material was wiped and bass guitar, a trumpet and a glockenspiel superimposed in its place.1 Seven hours for one overdub pass on a song already set aside is a real commitment of studio time, and it is the night the record acquired the instrumentation it is now known for.
  • Friday 21 April 1967 — Studio Two, 7.00pm–1.30am: mono remixes 1–11 from takes 3 and 11. P: George Martin. E: Geoff Emerick. 2E: Richard Lush.1 The same session recorded ‘Edit for LP End’, take 1.1 The remix numbering starting again at 1 two days after remix 4 is the paperwork registering that the source tape had changed underneath it.
  • Wednesday 1 November 1967 — Abbey Road, Room 53, 10.00am–1.00pm: tape copying, from mono remix 6.1 A copy rather than a mix, made in a small transfer room alongside copying work on ‘I Am The Walrus’.1 Copies like this are how a shelved recording gets moved toward a use nobody had in mind when it was cut.
  • Tuesday 29 October 1968 — Studio Three (control room only), 10.00am–1.00pm: stereo remix 1, from mono remix 6. E: Geoff Emerick. 2E: Graham Kirkby.1 The session carried the Yellow Submarine stereo mixing as a batch — ‘Hey Bulldog’, ‘All Together Now’ and ‘All You Need Is Love’ were mixed the same morning.1 Making a stereo remix *from a mono remix* is the tell: this is mock stereo, not a return to the multitrack.3 No producer is credited on the session sheet.
  • Friday 17 January 1969 — released on Yellow Submarine, Apple [Parlophone] PMC 7070 (mono) / PCS 7070 (stereo).1 Twenty-three months after the basic track, on an album whose second side is a George Martin orchestral score rather than a Beatles recording.

Studio Notes

  • An input is a currency, and the vocal microphone paid — the KM56 requested for the vocal was given up so that its input could feed the RS144 premix box used with the drums.2 A four-track machine is not the constraint people assume it is; the constraint sits upstream, at the desk, where every microphone needs a channel before anything reaches tape. Five drum microphones cannot each occupy a channel and then be summed to one track without something doing the summing, and the premix box that does it is itself an input. So the arithmetic runs: keep two vocal microphones and lose the drum premix, or keep the drum sound and record the voice on one. The sheet shows which way it went. It is worth noticing that this is a decision taken before a note was played, on paper, by an engineer working out what the night could physically hold.
  • Emerick’s standard drum setup, and what it was for — a D20 on the bass drum, D19c microphones on toms, hi-hat and overhead, and a KM56 below the snare.2 The pattern is close-miking with one dynamic per drum and a condenser doing the detail work, and it is the opposite of the two-overheads-and-a-room approach the studio had used a few years earlier. Placing the snare microphone *below* the head is the characteristic move: it puts the capsule where the snare wires are, so the microphone hears the rattle rather than the stick, and it also puts the drum shell between the capsule and the rest of the kit, which reduces the hi-hat and cymbal spill that would otherwise dominate a close snare channel. A kit captured this way does not sound like a kit in a room; it sounds like a set of individual instruments that can be balanced against one another at the desk.
  • An AKG C12 at the Steinway — the setup sheet puts a single C12 at the piano.2 The C12 is a large-diaphragm valve condenser with a remotely switchable pattern, and one of it on a grand is a deliberately simple choice: a single capsule cannot capture the instrument’s width, so what it gives instead is a coherent, centred piano with no phase relationship to manage. On a record that is going to be mixed in mono, that is not a compromise at all — a stereo pair on a piano folded to mono can lose more than a single microphone ever had. Choosing one microphone here is choosing the format the record was actually going to be heard in.
  • Two tapes, running together — the song was carried on two tapes locked in synchronisation, and the mono mix was made from both, mainly from take 12.3 Running two four-track machines together is how you get more than four tracks in a building that has no eight-track machine, and it is a fragile way to work: the two tapes have to start together and stay together, and any drift shows up as flamming and phase movement between whatever is on machine one and whatever is on machine two. The reason the mix is described as coming mainly from one of the two tapes rather than evenly from both is the practical consequence — the engineer leans on the tape carrying the material that matters and uses the other for what it uniquely holds.
  • Wiping is an editorial act, and it is irreversible — on 20 April the existing material was wiped and bass guitar, a trumpet and a glockenspiel put in its place.1 Superimposing onto a full four-track tape means recording over, not alongside; the previous pass is gone at the moment the new one is committed. Two months had passed since the song was last touched, so the decision was made cold, against a version nobody had heard in a while, and it was made without a safety copy of what was being replaced. Everything a listener now identifies as this record — the trumpet in particular — exists because somebody chose to erase what was there in February.
  • Why the remix numbering restarts — mono remix 4 was made on 19 April, and then remixes 1–11 were made on 21 April.1 Numbering that goes 4, then 1, is not a clerical error: remix numbers count mixes made from a given source, and the source changed when the 20 April overdubs replaced the material remix 4 had been drawn from. The old series could not be continued because it no longer described anything on tape. The paperwork is quietly recording the same fact the tape is — that there are two versions of this song, and only the second one survives.
  • Mono was not a preference here, it was the only mix — no stereo mix was attempted in 1967 at all.3 Eleven mono remixes in one April night and no stereo pass is a stronger statement than the usual mono-first bias of the period, because most Sgt. Pepper-era songs did eventually get a stereo mix within days or weeks. This one did not, and the mono was what went into the original film print of Yellow Submarine.3 The consequence is that for the first eighteen months of its existence this recording had exactly one released form, and every later stereo version is a derivation rather than an alternative reading of the same tape.
  • What mock stereo does, and why it is audible — the 1968 stereo was made from mono remix 6 rather than from the multitrack.1,3 A stereo image built from a finished mono mix cannot separate anything, because there is nothing left to separate; what it can do is split the signal, delay or filter one side, and let the ear infer width from the difference. The result has a characteristic hollowness in the middle and a phase relationship that collapses awkwardly if the two channels are summed back together — which is exactly what happens on a mono radio or a single speaker. Genuine stereo mixes of this song were not made until 1995 and 1999.3
  • A placeholder is not a working title — the 13 February session sheet carries the song as ‘Not Known’.1 Working titles are ordinary in this catalogue and usually descriptive of something in the song. ‘Not Known’ is different in kind: it is the studio recording the absence of a title rather than a provisional one, which means the piece arrived at the session without a name attached to it in the paperwork. Set beside the nine-week gap that followed and the wholesale replacement of its overdubs in April, it is the first of three separate pieces of documentary evidence that this recording was not, at the time, being treated as finished work.

Releases

  • 1969 UK mono — Yellow Submarine (17 January 1969, Apple [Parlophone] PMC 7070, side one) — the mix the song was made for and the only one that existed for eighteen months. Everything is centred and dense, with the trumpet and glockenspiel sitting hard against the organ rather than being placed away from it. This is also the version heard in the original print of the film.
  • 1969 UK stereo — Yellow Submarine (17 January 1969, Apple [Parlophone] PCS 7070, side one) — width without separation. The same balance is spread across two channels, thin through the middle, the instruments moving as a single mass rather than occupying places of their own; summed back to mono on a single speaker it thins further. This is a mock stereo derived from the mono, not a mix made from the multitrack.
  • 1969 US — Yellow Submarine (Apple SW-153) — the American issue of the same album, carrying the same pairing of a mono master and a derived stereo. Nothing distinguishes the American pressing musically; it is here for completeness.
  • 1987 CD — Yellow Submarine (EMI CDP 7 46445 2) — the first digital issue, carrying the derived stereo forward. Cleaner and brighter than the LP, and unchanged in balance; the hollow centre of the mock stereo is if anything easier to hear on a clean transfer.
  • 1996 — Anthology 2 (Apple CDP 8 34448 2) — carries a stereo mix made in 1995, the first time the song was addressed in stereo from tape rather than manufactured from the mono. Worth hearing next to the 1969 stereo simply to establish what the earlier version was not doing.
  • 1999 — Yellow Submarine Songtrack (Apple CDP 5 21481 2) — the version most listeners now know, and the first on which the instruments occupy defined positions. The bass is a separate object rather than part of the middle, and the trumpet and glockenspiel read as overdubs placed on a track rather than as a smear across it. Made in full stereo from the multitrack.
  • 2009 — Mono Masters (EMI 99463 2) — the 1967 mono master remastered and back in print as part of the 2009 programme. This is the closest thing available to what was signed off in April 1967, and the natural reference against which every stereo version of this song should be judged.

Sources

  1. The Complete Beatles Recording Sessions — Mark Lewisohn (Harmony Books, 1988), pp. 97, 103, 109, 130, 163, 164, 200.
  2. Recording the Beatles — Kevin Ryan & Brian Kehew (Curvebender Publishing, 2006), p. 452 (‘13 February 1967 Session for “Only A Northern Song”’).
  3. Joseph Brennan, Usenet Guide to Beatles Recording Variations (rev. edn.), Only a Northern Song entry.

Frequently asked

Why was Only a Northern Song left off Sgt. Pepper’s Lonely Hearts Club Band?

It was taped during the Sgt. Pepper sessions, in February 1967, but was not taken up for the album. It was set aside and reached release two years later on the Yellow Submarine soundtrack.1

What instruments were overdubbed on Only a Northern Song?

On 20 April 1967 the existing material was wiped and bass guitar, a trumpet and a glockenspiel superimposed in its place.1

What was the working title?

‘Not Known’. The 13 February 1967 session sheet carries the song under that placeholder rather than under a provisional title.1

What microphones were used?

Geoff Emerick’s setup sheet for the session records a D20 on the bass drum, D19c microphones on toms, hi-hat and overhead, a KM56 below the snare, Neumann U67s on the guitar amplifiers, an AKG C12 at the Steinway and a U48 for the vocal.2

Is there a real stereo mix of Only a Northern Song?

Not from 1967 — no stereo mix was attempted then, and the stereo on the 1969 LP was derived from a mono remix in October 1968. Genuine stereo mixes were made in 1995 and 1999.1,3

How many takes were there?

Takes 1–9 were recorded on 13 February 1967; take 3 was reduced the following night, and of the takes that followed, take 11 was good and take 12 was the best.1