[Dixielandjazz] Jelly Roll Morton, was NO Wanderers

[Ken Mathieson]

Anton wrote:
There is also a memorable difference in George Mitchell's second, brief solo. In the first take he blows an astonishing flurry of notes; in the second take he articulates. Can any of our DJML trumpeters tell me what is the technical term for what Mitchell does in that first take?

Hi Anton et al,

I'm not a trumpeter, but I'd venture the technical term for Mitchell's flurry in the modulation of Cannon Ball Blues (Take 2) might be panic or even goof. Take 1 was taken at a leisurely 110 bpm, whereas Take 2 was at 140 bpm, so Jelly's chart would be a whole lot harder to articulate in the latter. When Mitchell gets to the last bar of the modulation in Take 2, he fluffs the line as played in Take 1 (probably written out) and tries to cover it with a flurry of notes. Take 1 was actually the issued take, but ran to 3:36, which was close to the upper limit for a 78rpm record. I suspect that Take 2 (at 2:59) was meant to be the preferred take for a couple of reasons: it was about the ideal length for a 78 and, in JRM's solo (behind the horns playing organ chords), the piano is more prominent in the mix than in Take 1. However Mitchell's articulation problem in the modulation and Jelly's slight rhythmic unease in his solo behind the horns (he rushes in bars 5 & 6) might have convinced Victor to release Take 1. 

The arrangement is identical in the 2 takes, so any perceived difference in feel comes from the 2 tempos being so different. I haven't transcribed the various horn parts from the 2 takes to check for differences, but all the theme statements are identical apart from minor ornamentations and inflections, so it's pretty clear that Jelly hadn't left much to chance. I've seen the quote about solos and breaks being improvised, but everything else was as Jelly had written or instructed (I think it was Simeon who wrote this). However, comparison of alternative takes shows that a lot of the solos are very similar from take to take. There may have been more freedom in ensembles, but it would be a massive task to transcribe the individual parts for comparison purposes. I wonder if anyone has attempted this? 

Cheers,

Ken Mathieson