Archive for September 12, 2016

Back On Green Dolphin Street

Posted in Biographical, Jazz with tags , , , , , on September 12, 2016 by telescoper

I was listening to this wonderful track yesterday and couldn’t resist reposting a piece I wrote I wrote about it over 7 years  ago. If I were ever to be asked on one of those programmes where you have to pick examples of your favourite music, this would probably be the first I’d write on my list.

Years ago in 1980, when the great pianist Bill Evans passed away suddenly, Humphrey Lyttelton paid tribute to him on his radio programme “The Best of Jazz” by playing a number of tracks featuring him. I didn’t really know much about Bill Evans at the time – I was only 17 then – but one track that Humph chose has been imprinted on my mind ever since, and it’s one of those pieces of music that I listen to over and over again.

The track is On Green Dolphin Street, as recorded in 1958 by the great Miles Davis sextet of the time that featured himself on trumpet, John Coltrane on tenor sax, Julian “Cannonball” Adderley on alto sax, Jimmy Cobb on drums, Paul Chambers on bass and Bill Evans on piano. This is the same band that played on the classic album Kind of Blue, one of the most popular and also most innovative jazz records of all time, which was recorded a bit after the recording of On Green Dolphin Street.  I love Kind of Blue, of course, but I think this track is even better than the many great tracks on that album (All Blues, Flamenco Sketches, Blue in Green, etc). In fact, I’d venture the opinion – despite certainty of contradiction – that this is the greatest Jazz recording ever made.

On Green Dolphin Street was suggested to Miles Davis the band’s leader by the saxophonist Cannonball Adderley. It was the theme tune from a film from the late 1940s. It’s also the title of a more recent very fine novel by Sebastian Faulks.

I think the Miles Davis version demonstrates his genius not only as a musician himself but also as a bandleader. On Green Dolphin Street definitely bears the Miles Davis hallmark, but it also manages to accommodate the very different styles of the other musicians and allows them also to impose their personality on it. This is done by having each solo introduced with a passage with the rhythm section playing a different, less propulsive, 3/4 time behind it. This allows each musician to set out their stall before the superb rhythm section kicks into a more swinging straight-ahead beat  (although it still keeps the 3/4 feel alongside the 4-4, courtesy of brilliant drumming by Jimmy Cobb) and they head off into their own territory. As the soloists hand over from one to the other there are moments of beautiful contrast and dramatic tension, especially – and this is the reason why Humph picked this one in 1980 – when Bill Evans takes over for his solo from Cannonball Adderley. He starts with hesitant single-note phrases before moving into a richly voiced two handed solo fully of lush harmonies. It’s amazing to me to hear how the mood changes completely and immediately when he starts playing, and it always sends shivers down my spine.

Not that the other soloists play badly either. After Bill Evans’s short but exquisite prelude, Miles Davis takes over on muted trumpet, more lyrical and less introspective than in Kind of Blue but still with a moody,  melancholic edge. He’s followed by John Coltrane’s passionately virtuosic solo which floods out of him in an agonized stream which contrasts with Miles’ poised simplicity. By contrast, Cannonball Adderley is jaunty and upbeat, sauntering through his solo up to that wonderful moment where he hands over to the piano. Then Miles Davis takes over again to take them to the conclusion of the piece.

I’m not into League tables for music, but this is definitely fit to put up alongside the greatest of them all…

Advertisements

Bayes Factors via Savage-Dickey Supermodels [IMA]

Posted in Bad Statistics, The Universe and Stuff on September 12, 2016 by telescoper

How could I possibly resist reblogging an arXiver post about “Savage-Dickey Supermodels”?

arXiver

http://arxiv.org/abs/1609.02186

We outline a new method to compute the Bayes Factor for model selection which bypasses the Bayesian Evidence. Our method combines multiple models into a single, nested, Supermodel using one or more hyperparameters. Since the models are now nested the Bayes Factors between the models can be efficiently computed using the Savage-Dickey Density Ratio (SDDR). In this way model selection becomes a problem of parameter estimation. We consider two ways of constructing the supermodel in detail: one based on combined models, and a second based on combined likelihoods. We report on these two approaches for a Gaussian linear model for which the Bayesian evidence can be calculated analytically and a toy nonlinear problem. Unlike the combined model approach, where a standard Monte Carlo Markov Chain (MCMC) struggles, the combined-likelihood approach fares much better in providing a reliable estimate of the log-Bayes Factor. This scheme potentially opens the way to…

View original post 53 more words