The Johnny B. Goode Session

When Fred Rothwell a few weeks ago reported here on his new findings regarding the 'who-played-on-what' questions of Chuck Berry's discography, one of the most interesting changes to the Chuck Berry sessionography was made to the personnel which created Johnny B. Goode.
The session's recording contract encountered by Tim McFarlin during his studies of the Berry vs. Johnson suit of 2000-2002 lists Johnnie Johnson as piano player for the recording session dated January 6, 1958. According to what is listed in the discographies, this is the session in which Johnny B. Goode was recorded. Formerly, Fred and other experts had listed Lafayette Leake on piano.
Fred‘s sessionography change first got various comments posted here on the blog and then resulted in almost two months of (sometimes heated) discussions in email to which Berry experts from the U.S., from the Netherlands, from England, France, Norway, and Germany contributed.
In the end we had to agree that we do not agree on a common opinion. However, as this topic is of interest to most Berry collectors I will try to sum up the facts and the most important opinions.

Speaking of facts we found that we have astonishing few 'hard facts' to base any discussion or result on.

This starts with the date of the session which generated Johnny B. Goode. Depending on which source you consult the reported recording date for this song is February 28, 1958 (Michel Ruppli, The Chess Files) or December 29, 1957 (Mike Leadbitter/Neil Slaven, Blues Records). Berry's Autography has the date listed as February 28, 1958 as well. Who is correct? We have some hints:

We know that Chess Records assigned the matrix number 8633 to the final recording and mix of Johnny B. Goode. We also believe that Chess assigned matrix numbers in the sequence the master tapes were finished. The matrix numbers directly following Johnny B. Goode were assigned to different artists: The Pastels (8634/35), The Lewis Sisters (8636-40), Harvey & The Moonglows (8641-43), and so on. The next numbers assigned to Chuck Berry records are 8656/57 (A and B sides of EP 5121 Sweet Little 16), 8689/90 (side 1 and 2 of LP 1432 One Dozen Berrys). The next Berry recording Around And Around is sixty numbers after Johnny B. Goode and got the matrix number 8693. This points to at least a couple of weeks between the mastering of Johnny B. Goode and that of Around And Around. And if we assume that mastering in the 1950s was done either concurrently with the recording or soon thereafter, this also points to a couple of weeks between the recordings of the two. As a sidenote: From later sessions we know that songs recorded the same day got master numbers a dozen or so numbers off, probably because the mastering of the later songs was delayed.
Of more interest are the master numbers preceding Johnny B. Goode. They all are assigned to Chuck Berry recordings: Sweet Little Sixteen (8627), Rock At The Philharmonic (8628), Guitar Boogie (8629), Night Beat (8630), Time Was (8631), and Reelin' And Rockin' (8632). This means that all these songs including Johnny B. Goode have been mastered/recorded in one session or a set of consecutive sessions, in any case so close to each other that no other masters were made in between.
This is the reason why both Michel Ruppli and Leadbitter/Slaven had all seven songs listed as a single session. If this would be true, Ruppli's session date of February 1958 cannot be correct because Sweet Little Sixteen was already in the stores by January. Chuck Berry's list of recording sessions as published in his book places the six early songs (masters 8627 to 8632) in a session dated January 6, 1958 while he puts Johnny B. Goode (8633) along with Around And Around (8693) and five other songs (masters 8694 to 8696) to February 28.
When compiling his sessionographies, Fred Rothwell took the most probable route. He placed the recording of the consecutive masters 8627 to 8633 close together, i.e. put Johnny B. Goode close to Reelin' And Rockin'. However, because we know that the released version of Sweet Little Sixteen was take 14 and the released version of Reelin' And Rockin' was take 10, Fred had strong doubts that all these were recorded the same day. It would have been an awful long session. Therefore he used the December date from Blues Records for Sweet Little Sixteen and Chuck Berry's January date for Johnny B. Goode. These recording dates are so close together that consecutive master numbers are probable.

The session contract encountered by Tim McFarlin during his research of the legal papers filed for the 2002 lawsuit lists a date and personnel, but in contrast to later contracts it unfortunately does not give us a list of songs recorded. Thus if we believe the recording contract — and we should as the other contracts make perfect sense —, we know that a session took place on January 6th, 1958 (the date from Berry's book) and that personnel included Johnnie Johnson on piano.

Is this a proof? No, because you can argue that we don't know of a contract for a December session (yet?), that there is no list of songs, that there may be other sessions between January and March 1958 (when Johnny B. Goode hit the stores). But placing the recording of Johnny B. Goode (and maybe the others) with the January session sounds reasonable given the information we have.

Other information we have is an audio protocol of what happened at the session which generated Johnny B. Goode. The recording tapes of this session have survived and have been released to the public. So they form some additional 'hard facts' we may base our discussion on.

From the tapes we know of three tries to record the song during this session. To judge the audio recordings one has to take into account that the released versions are not labeled correctly.

The correct sequence of the recordings has been discussed here in a blog post dated July 26, 2011:

Johnny B. Goode - take 1: was first released in 1986 on CHESS CH2-92521 "Rock 'n Roll Rarities" as the second part of a track named "Johnny B. Goode — previously unreleased version".

Johnny B. Goode - take 2: is a very brief take which starts correctly but is then interrupted. This second take has been released twice: Complete with the announcement "Johnny B. Goode Take Two" on Hip-O Select's "Johnny B. Goode - His complete 1950s recordings" and without this announcement but with a false start as the first part of the previously unreleased version on "Rock 'n Roll Rarities". Note that the sequence of the two takes is reversed on the 1986 release. Also note that the Hip-O set misses take 1 completely. It is only on the 1986 double album.

Johnny B. Goode - take 3: exists in two variants. The original recording of this take without any overdubs was first released on the Hip-O Select set in 2008.

Johnny B. Goode - take 3 including guitar overdub: This is the 1958 hit version. Like on many other recordings, Berry on take 3 played just the first part of the lead guitar intro but then continued playing the rhythm guitar. The remaining parts of the guitar intro as well as further guitar solos were recorded and overdubbed later, probably during the same session.

Again this provides us with some more facts, but how much of this can be considered as 'hard facts'? The takes are introduced as takes one, two and three. Take 3 is the basis of the final released record. So we can assume that there were no other takes. The final master (8633) however is a modified version of take 3. The most obvious modification is the addition of further lead guitar segments.

Another possible modification is a manipulation of the playback speed. It is known that Chess Records modified the playback speed of Sweet Little Sixteen to make Berry's voice sound younger. When a sound recording is played faster, the pitch becomes higher with the voices sounding lighter.

With Johnny B. Goode such a modification is not as obvious as with Sweet Little Sixteen. The running times of the undubbed take 3 and of the final master are almost identical. Whereas we have to keep in mind that we have access to the final master only in the form of 45 rpm records and digital copies like the one on the Hip-O Select box.

Just for academic purposes (and without any other use anyway) I have created myself an audio file in which one can hear the beginning of both variants of take 3 of Johnny B. Goode. The left stereo channel is the un-modified take 3, the right stereo channel contains the released master. One notices the overdubbed guitar which is now to be heard on the right channel only. And one can notice that there is a tiny difference in speed. It sounds as if the released master indeed has a slightly higher pitch and runs a little bit faster.
[I have asked Universal Music for permission to provide readers with a download link to this audio file. I have not received any permission nor any response at all, though (yet).]

What does this tell us about the recording session itself? Very little. We have no information when the guitar overdub was recorded and how. In the late 1950s there was no multi-track tape recording at Chess. Thus it is probable that the original take was played back into the studio where Berry then added the missing guitar lines. This is also where the speed difference may come from: The tape played back on a different machine and then re-recorded along with the solo guitar.

All we can tell for sure is that the mastering of the overdubbed take happened in temporal proximity to the mastering of Sweet Little Sixteen which in turn obviously happened before that song's release in January 1958.

In regard to the discussion on who was the pianist on Johnny B. Goode the differences between the three takes are relevant. Comparing the takes, it is obvious that Berry pretty much knew how he wanted the guitar to sound like. The main difference between the complete takes 1 and 3 is the piano playing. And how important this piano playing was becomes audible from the discussions taking place during take 2.

Take 2 starts just normal with the famous guitar intro. Then, when the piano comes in, someone shouts "hold it" and a dialog starts which I interpret as follows. To better understand the different sentences, I created another sound file in which I tried to level the loudness of those said through a microphone and those said without.
[Again I wanted to provide a download link to this sound-enhanced excerpt just for academic purposes but did not hear from Universal Music as the owner of this recording.]

I hear this dialog:

Voice 1 'Hold it, hold it, hold it!'
Voice 2 'What do you want, Jack?'
Voice 1 'You were making Roll Over Beethoven on piano that time, stay away from that!'
Voice 3 'Piano or guitar?'
Voice 2 'Keep playin' on the guitar during the solo.'
Voice 1 'Yeah, on the solo he was makin' Roll Over Beethoven. Stay away from that one!'
Voice 4 'Who is (he talking to)?' interrupted
Voice 1 'Johnny B. Goode, take three'

It is not clear which voice belongs to which person. In my opinion Voice 1, the main voice (on the microphone), is the voice of Jack Sheldon Wiener, engineer and from May 1957 to August 1958 co-owner of the recording studio at 2120 S. Michigan Av., Chicago. Jack is referenced and talked to in other segments from this session, e.g. in talks related to Sweet Little Sixteen and Reelin' And Rockin'. Voice 2, who instructs Berry to continue playing guitar during the piano solo, seems to be Leonard Chess. This fits to Berry's recollections of the session in his book where he writes: "Leonard Chess took an instant liking to this song and stayed in the studio coaching us the whole time we were cutting it." Voice 3 must be one of the musicians. Since he has no microphone I suspect this is the pianist asking. Voice 4 finally sounds like Chuck Berry to me.

Wiener obviously noticed that the piano solo on take 1 was too close to what they had released as Roll Over Beethoven. Berry himself did not care that much. He believed anyway that his songs differed in lyrics and solos alone. The rest was just standard. "Roll Over Beethoven, Johnny B. Goode, you name it, all of the songs could carry the same background or music that each other has." (Berry quoted by Tim McFarlin, see blog post of December 18, 2014)

I admit that both my interpretation of the studio dialog and my assignment of persons to voices is subject to discussions. The other Berry experts who listened to this studio talk had various different opinions. Some assigned Voice 1 to Leonard Chess, some even to Berry himself. Bob Lohr, who played piano behind Chuck Berry for the last decade, says:

The cat who stated "You were playing 'Roll Over Beethoven' ... stay away from that", and "he was playing 'Roll Over Beethoven' on piano" ... is clearly Chuck, not the engineer ... he's using the in-studio high quality vocal microphone and I'm 1000% sure it's Chuck ... after 18 years, I know his speaking voice like the back of my hand ... furthermore, the way Chuck pronounces "Beethoven" is pretty unique ... trust me, on my life I'm telling that was Chuck speaking, end of story!!! The engineer and LC [Leonard Chess] are speaking through the low quality studio talkback microphone.

I perfectly accept that Bob can identify Berry's voice as it sounds today. We should not forget that we are talking about a recording made when Berry was in his early thirties. Voices change and to me the instructing voice and the voice singing sound differently.

The main question the discussions about Johnny B. Goode circle around is the question "Who is the pianist instructed to stay away from playing Roll Over Beethoven". Some Berry experts point to Ellis "Lafayette" Leake, others favor Johnnie Johnson. Early discographies had listed Leake on piano, the ultimate discographical authority Fred Rothwell now lists Johnson as pianist — following the January 6 recording contract. It is unknown how the early discographies came to listing Leake. Fred writes in a recent article for "Now Dig This" magazine:

Session information about musicians has grown organically over the years and much of it has been based on anecdotal, word-of-mouth remembrances. In the '60s, blues fans would ask artists about old sessions and I'm sure guys like Willie Dixon, for instance, would try to placate them by giving info that was not always correct. Lafayette Leake was a big friend of Willie's and, I suspect, he got named as pianist for wont of someone else at times. Johnnie Johnson was not part of the Chess studio clique (he never recorded in his own name at Chess) and I think he may have been overlooked.

Those who favor Leake also say that both Berry and Johnson have denied many times over the years that Johnson was among the staff recording Johnny B. Goode. However, I was not able to find a single source for this claim. The only source to this effect is from Travis Fitzpatrick's biography of Johnson where he cites Johnnie saying "The only recordin' I didn't play on was 'Johnny B. Goode'. Chuck did that as a surprise for me."

Asked about this quote by Tim McFarlin and me, Travis said that when using Johnnie's quotes one should always keep in mind that Johnnie's interpretations and those of the reader might not necessarily match. The whole lawsuit between Berry and Johnson was based on the fact that to Johnson "writing music" was something completely different than what a copyright lawyer would understand thereunder.
Likewise, Johnson's saying that he did not play on Johnny B. Goode may not necessary mean "play piano", but more to the effect of "play a role" meaning "I did not contribute to Johnny B. Goode which I didn't know about before we went into the studio.".
In an 1999 interview with Ken Burke for the Rockabilly Hall of Fame (see http://www.rockabillyhall.com/DrIJJohnson.html) Johnson became more specific:

That's how we worked out all the tunes that's he's [Chuck Berry's] got practically, except "Johnnie B. Goode." I had nothing to do with that, that was sort of a tribute to me, I understand.

This is from Travis:

Since he is no longer around to clear it up, I can only guess what Johnnie meant with his original statement about not playing on "Johnny B. Goode." I probably should have questioned him more about it before he passed away. Remember, this is the man who said, "I didn’t write the music with Chuck, I was just in the room sometimes when he was writing" before describing the process he and Chuck used to write their music! It was years before we understood the reason why he said this. Johnnie believed writing music meant writing down lyrics or transcribing notes onto a lead sheet. Johnnie called what he and Chuck did "making up music" because it wasn’t written down. If you ever saw the movie Forrest Gump, that was very much how Johnnie viewed the world. As a consequence, more misunderstandings are coming to light. For example, Johnnie didn’t think he played on the early Mercury sessions because he thought the re-recording of all their Chess hits was due to a fire at Berry Park destroying the originals. He thought Chuck arranged to re-record the old songs on his own! That was Johnnie in a nutshell.

Whether or not Johnson or Leake played piano during the Johnny B. Goode session is still open to discussion. One would think that people who know Johnson's playing well can simply hear whether it's him playing. In the same 1999 interview Ken Burke asked Travis Fitzpatrick: "Even as low in the mix as some of Johnnie's piano work is, would you know his playing when you heard it?"
Travis replied:

Sure. I can always tell his playing. [...] I can listen to a lot of those songs and tell it's him. When I listen to some of those original Berry records I can say "That's Johnnie for sure!" I can tell that Lafayette Leake came in on some stuff, especially "Johnnie B. Goode." I can tell that's not Johnnie. Then, like he was saying, there's this whole thing where Leonard Chess would come in during his solo and run his hand up and down the keys, which Johnnie never does. So, that kind of made it more difficult, plus Lafayette Leake was a very good mimic. [...] But I'm 100% sure that was Lafayette Leake on "Johnnie B. Goode."

Another expert on Johnnie's piano playing is Bob Lohr. Bob is a pianist himself and has played with both Berry and Johnson. He likewise claims that he can identify Johnson's playing, too:

I'm extremely familiar with JJ's [Johnnie Johnson's] style. I have been called upon here in local studios over the years to 'clone' or mimic JJ's style on different projects as JJ's style is pretty much ingrained in my musical DNA. Based upon my familiarity with JJ's style, I would have to say that it was clearly LL [Lafayette Leake] instead of JJ on JBG [Johnny B. Goode] based upon style alone. The stylistic differences between LL and JJ makes me sure that LL was the man on the keys despite the union log of the date. They both played in a similar boogie/blues mode behind Chuck (and often on the same out-of-tune piano apparently!!!), but Leake ... with all due respect to JJ ... was a far more fluid and accomplished jazz player and generally threw in some nice fat jazz double-hand chording at the end of his solos ... something that JJ rarely if ever did. You'll hear Leake do this throughout Takes 2/3 and on the final take as well.

Interestingly, both experts did not know that take 1 of Johnny B. Goode existed which has a very different piano playing and was released only on the 1986 double album. When I asked them to re-check take 1, Bob Lohr found: "You are correct in that it sounds a lot like JJ's style, although I can still hear the stylistic difference."
Travis Fitzpatrick was even more astonished:

I must revise my opinion (an ultimately my book) concerning Johnnie Johnson’s playing on "Johnny B. Goode." Until Dietmar pointed it out, I did not realize that take one was misplaced as take three on Rock 'n Roll Rarities. Consequently, I never really listened to it. Well believe me ... I have now listened to it. I listened to that first take of "Johnny B. Goode" for hours last night. My immediate reaction was "Holy COW! The AFM contract was right! That is Johnnie Johnson!" Just to be sure, I jumped into my Johnnie recordings both issued and unissued and found examples of every lick. His phrasing and the way he resolves his licks is Johnnie’s fingerprint. It is him. The flashiest lick has been right under everyone’s nose. Watch the rehearsals for "Carol" in Hail Hail Rock and Roll or better yet, Johnnie’s backing behind the sax on "Almost Grown" in Hail Hail Rock and Roll. Those songs are in C and G respectively, but you can see that the lick is in his repertoire and in fact, he uses it quite a bit. Just not on most of the Chuck Berry recordings — which is why it hasn’t become recognized as a standard JJ lick.

Travis' remark on the song keys is significant because Johnny B. Goode is written in B flat (Bb). Bob Lohr again:

It's harder to play Chuck's style (or blues in general) in the flat keys ... E flat (Eb) or B flat (Bb). Johnnie was not too good at playing in those keys and would never use those keys when playing in his own band. LL on the other hand, was technically a much more accomplished classically-trained/jazz player who could play almost as well in Bb or Eb as JJ could play in G. JJ could certainly play in B flat or E flat, but nowhere near as well and as fluid as LL plays here in B flat. If in fact JJ played on JBG, he played in a completely different style which we have never heard before or since from JJ ... and better in B flat than he ever played before or since.

Bob's statements about Johnson's facility in playing jazz licks or playing in B flat at all had Travis to disagree:

Regarding Johnnie’s proficiency in B-flat and as a jazz musician: Johnnie Johnson was playing jazz professionally in Detroit at age seventeen. As a nineteen-year-old Marine, he was handpicked by Bobby Troup to play piano in the Barracudas, a USO service band featuring members of Glenn Miller’s, Tommy Dorsey’s, and Count Basie’s orchestras. After WWII he returned to Detroit where big band jazz and swing had gone out of style and he had to learn bebop like Dizzy Gillespie and Bud Powell to get jobs. Johnnie could play "How High the Moon" in every key; I saw it with my own eyes when he was giving me lessons. His trio in Chicago, with Milton Rector on bass, was a jazz combo. The Johnnie Johnson (or Sir John’s) Trio was a jazz combo when they started at the Cosmopolitan Club — remember, Chuck replaced a saxophone player. If you asked Johnnie to list his piano heroes, they were all jazz greats, no blues and certainly no rock pianists. His idol was Oscar Peterson. Johnnie knew jazz very, very well. Defining the limits of Johnnie’s jazz ability based on what he played with Chuck Berry and on his blues albums is like determining someone’s knowledge of chess by how he plays checkers. Johnnie’s favorite keys were G, C, B-flat, F, D and E-flat. He did not like the way A sounded (thought it was too bright), he played E grudgingly, and he absolutely hated C sharp/D-flat and B. So yes, if you just turned him loose on a blues or rock song, he would do what he enjoyed and what came easiest to him. I have a recording of Johnnie playing "Johnny B. Goode" at the Rock and Roll Hall of Fame. It is in the original B-flat, and he nails it. We even had the rest of the band pull out so he could show what it sounded like on piano. I can think of at least one great live blues off the top of my head that he does in B-flat as well. And he used to do this gospel style breakdown playing "Maybellene" live with Chuck that was really cool. The point being, Johnnie was very good in B-flat.

At this point we have moved far away from the 'hard facts' issue we started this article with. We're now down to rely on expert opinions — the same way a judge would have to rely on expert testimonials. Like a judge you will have to come to your own conclusions.

One thing both of our experts have not taken into account enough is — at least in my personal opinion — that during the change from take one to take three Jack Wiener (or whoever) clearly instructs the pianist to play in a completely different style. This can be a good reason why Johnnie Johnson does not sound like Johnnie Johnson on the final recording.

All in all I finally agree with Fred Rothwell's sessionography change. Whether you do depends on how you judge the various facts and opinions by yourself. This lengthy article should give you everything to come down to your own decision.

[Many thanks to all the Berry experts who provided useful information during the creation of this article, most notably Fred Rothwell, Morten Reff, Josep Rulló, Bob Lohr, Tim McFarlin, Travis Fitzpatrick, Michel Ruppli, and Arne Wolfswinkel.]