Thelonious Monk’s composition “Pannonica” is written in the key of ‘C’. This is a firm fact. There are several versions of this work by Monk, himself, that are always in ‘C’. That Monk is so insistent that “Pannonica” should be played in ‘C’ makes it all the more surprising that his unaccompanied celeste introduction on the version of “Pannonica” from the album Brilliant Corners – the first display of the piece – starts off in the key of ‘A’. As Monk, after this initial recording of “Pannonica” on October 9, 1956, never again plays an introduction on celeste allows a plausible explanation that the composer, using the bell-like sound of the celeste, chose to make that striking sound even more distinct by playing in ‘A’ to contrast with the ensemble playing in ‘C’. However, an incomplete alternate take of “Pannonica” from the same session contains an unaccompanied celeste introduction that is clearly played in ‘C’, as is the balance of this outtake and all subsequent performances of the tune by Monk. The presence of this outtake muddles any attempt to analyze the celeste introduction in the key of ‘A’. Still, the very common conjecture that Monk – for just this one time and off his main keyboard – chose to play the introduction to “Pannonica” in an unusual key before slipping back to the piano and playing, with his combo, the balance of the recording in the piece’s home key, ‘C’, seems supported by the evidence: the recording itself.
Enter Mark Micchelli. Mark is an excellent pianist and is currently the Director of {Western} Classical Music at WKCR. Broadly speaking, Mark is one of my students; but in this instance, I came to study at his feet - or ear, I should say. Micchelli is one of those very, very rare birds with genuine perfect pitch. I’ve only known a few, and first understood it when I took Wynton Kelly to hear music a few times: he wouldn’t go in if the piano was out of tune. I became more closely connected to the phenomena by befriending and working with Big Al Sears, Benny Carter, and Milt Jackson. Micchelli’s sense of pitch is in their league and, perhaps, then some.
Upon reviewing the simple analysis that the master take of “Pannonica” from Brilliant Corners has a celeste intro in ‘A’ and then moves to ‘C’ for the balance of the recording, Mark used his great ear to make a profound discovery: the pitch is constantly sliding – gradually ascending from ‘A’ to 'C’ - from the very start of the performance into its ninth bar. If having the celeste introduction to “Pannonica” on “Brilliant Corners” in ‘A’ before the balance of the piece was played in ‘C’ was surprising, then learning that the recording starts in ‘A’ and then slides upward in pitch to the key of ‘C’ in something close to nine bars is shocking.
The rapidity and near uniformity of this sliding of pitch is beyond the scope of the various instruments playing together. The recording slows down 12.65% to reach a uniform (and most likely correct) speed during its first 32 seconds (or nine bars). Whatever the explanation for the ‘A’ to ‘C’ shift in these first 32 seconds, the musicians could not have played it that way: it is impossible.
The balance of possible explanation thus falls into two concepts: either the recording was adjusted by technical means intentionally for a musical/commercial purpose; or, the technical means that changed the recording’s speed and pitch was a glitch, an inadvertent manipulation that caused the recording to play the way it does.
I see no musical or commercial purpose to this sliding of pitch. Furthermore, I attempted to “intentionally” duplicate the effect and found that, even using a well-tweaked analog vari-speed device, manipulating it to achieve uniform ‘C’ was a challenge. [Remember: it could not have been done digitally in 1956.] If the change of speed/pitch was done intentionally – and, again, a real reason to do this is not easily conjured – then who could confirm that it happened that way?
I contacted the producer of the session, Orrin Keepnews. On April 30, 2012, I spoke at length with Orrin. Given that he had just turned 89 and that the recording was 55½ years old, Keepnews was remarkably clear and entirely gracious in going over this technical point in detail.
Among Keepnews’ comments regarding the April 30, 2012 session was the notable claim that, compared to his other sessions with Monk, “the recording of ‘Pannonica’ was painless.” To the point of intentional tape speed manipulation, Orrin was very direct: “No! There wouldn’t have been anything like that.” To the possibility that some sort of technical problem had created the playback situation, Keepnews said he would turn to his handwritten notes made that day, October 9, 1956, at the session. Upon a preliminary examination of his notes, Orrin stated that there is no mention of it and, also, in retrospect, that he had no recollection of technical problems at the session. He made the additional point that the recording engineer might still be with us; but Keepnews added that he hadn’t spoken with him in upwards of a half-century.
Broadly, and not fully contemplating the playback specifics – a drop of speed of 12.65%, nearly uniform, in 32 seconds, as nine bars of music slid from ‘A’ to ‘C’ – Orrin seemed more accepting and bemused that Monk might’ve played the intro in ‘A’ and, not needing to explain that to anyone, didn’t. I finalized my preliminary assessment that the change was unintentional.
The point remained that whatever the inadvertent technical glitch that caused this long-unnoticed change of pitch and key in the opening bars of “Pannonica” on the original album “Brilliant Corners”, it is a most unusual one that challenges finite explanation. The concern of how it happened became a project for me and, although I have substantial background in audio, I turned to my peers and even betters: Tom “Curly” Ruff, “Midnight” Bob Shuster, and Mark Wilder.
Tom Ruff and Mark Wilder worked on this problem fairly early in my research. Their independent and joint conclusion was that the change in speed could not have happened inadvertently. A few months later (June 15, 2012), at a meeting here in New York City, I explained to them that it had to be unintentional. I also said that I had added the unbelievably gifted “Midnight Bob” to work on the mystery and find a plausible technical glitch that would explain how a reel-to-reel tape, presumably recording at the 1956 industry standard of 15 inches per second, could go from 16.8975 inches per second to 15 inches per second, descending in speed uniformly, over 32 seconds. [Dear Reader, please note: the performance is in ‘C’; the recording of the music is on a machine losing speed from 16.8975 inches per second to 15 inches per second; the playback speed is at 15 inches per second; so the music – in playback – is first played back slow, hence flat (in ‘A’) and gradually the recording speed slows down to the standard playback speed (15 inches per second) and the music plays back correctly, in ‘C, as it was performed.] That is what happened. But even Midnight Bob Shuster could not find or duplicate a plausible/possible explanation for the gear to induce this effect.
Eventually, I came up with one. In 1956, many reel-to-reel machines had 30 inches per second as a speed (“high”) as well as 15 inches per second (“low”). I conjecture that for some reason – possibly cueing the tape up to its next blank spot – the recording engineer threw the machine into “high” to get to the blank tape spot more quickly, but, at the command to start recording “Pannonica”, realized that he hadn’t changed the speed back to 15” per second. I think that the recording engineer flipped the recording machine from “high” to “low” speed as it was in motion, and that it took the machine at least 32 seconds to arrive at proper “low” speed, 15” per second.
Shuster allowed that this scenario does explain the resultant playback effect on “Pannonica”, but to this day has difficulty accepting it. Midnight Bob explains that no proficient machine would have needed 32 seconds to gain uniform speed as described in my scenario. But Midnight Bob Shuster, though he has tried mightily to find and duplicate one, has no other explanation for this effect. I spoke with him recently and he begrudgingly stated that my explanation is the only one known that could make the anomaly a reality.
Whatever the reason behind the situation, Monk’s first recording of “Pannonica” should only play in ‘C’. I’ve made a copy that only plays in ‘C’, but I look to the Mark Micchellis and Midnight Bob Shusters of the world to do it better and for the record.
