Art Tatum's greatest solo rendition?

[cuberdrift]

I'm an Art Tatum fan and I want to know who here thinks is the "greatest" (or one of the greatest) of his solo renditions.

I think that his 1953 version of Tea for Two may be at least one of the greatest here.

Lulu's Back in Town, on the other hand, is perhaps the most "virtuosic" recording of his I know.

What do you think?

EDIT: Oh, I just noticed dFrankJazz's post that it was his birthday last October 13. Pretty cool coincidence!

[chopinlover01]

My favorites are his ballads, strangely enough. The material lends itself well to him, though they aren't really ballads the way he plays them. Check out Sophisticated Lady, Body and Soul, etc.

[j_tour]

The most impressive to me showcase for his technique is his 1938 take on "Tiger Rag."  The LH 432 or 543 patterns are pretty tricky to me in general, at that tempo, whether or not he improvised the diminished chord patterns played in scales or worked it out ahead of time is a neat trick to have come up with.

You can compare the equally nice Jelly Roll Morton US Library of Congress recording of the tune from about the same time (sort of more how the tune is supposed to go, but at a pretty slower tempo) to see how far Tatum sort of tore the tune apart and just built up a remarkable little composition of his own.

[cuberdrift]

My favorites are his ballads, strangely enough. The material lends itself well to him, though they aren't really ballads the way he plays them. Check out Sophisticated Lady, Body and Soul, etc.

Weren't ballads his main repertoire? In any case, you should also check out his non-mainstream stuff; his Blues, Begin the Beguine, afro-cuban Taboo, Pole Boogie. I think he shines there in a certain way that he doesn't in his more common recordings.

The most impressive to me showcase for his technique is his 1938 take on "Tiger Rag."  The LH 432 or 543 patterns are pretty tricky to me in general, at that tempo, whether or not he improvised the diminished chord patterns played in scales or worked it out ahead of time is a neat trick to have come up with.

You can compare the equally nice Jelly Roll Morton US Library of Congress recording of the tune from about the same time (sort of more how the tune is supposed to go, but at a pretty slower tempo) to see how far Tatum sort of tore the tune apart and just built up a remarkable little composition of his own.

1938? I know that there are only four recordings of Tiger Rag; 1932 (slower, but more improvisatory), 1933 (the most famous version), 1935 (really fast), and 1940 (cleanest version).

And what LH patterns are you referring to?

Speaking of the Tiger Rag, Tatum there showcases his LH stride and he's probably the most virtuosic stride player of all time. Unfortunately we don't see him do this super-stride, but to give you an idea of what it's like Liberace actually does some stride in his own version of Tiger Rag and it is every bit as fast as Tatum (don't underestimate this guy lol).

It's seen in 1:15 of the below video.

[j_tour]

1938? I know that there are only four recordings of Tiger Rag; 1932 (slower, but more improvisatory), 1933 (the most famous version), 1935 (really fast), and 1940 (cleanest version).

And what LH patterns are you referring to?

Oh, I don't remember.  I knew I had the date wrong, but this is the recording I was thinking of.  Somebody made a transcription available online, but I've only glanced at it and don't have a link.  

His little LH ostinato thing comes in at around 0m55s and kind of continues as a little device from there.  

This time listening to the Tatum, I'm even more impressed than before -- I forgot he actually does more than play lip service to the "tune" (I'm still not sure "Tiger Rag" was ever really much more than a loose idea with some melodic fragments and a loose structure).  Pretty impressive.

I think Fats Waller could have done the straight cliché stride LH (octaves/tenths + chord in middle register) as good, but Tatum really just exploded that convention without losing coherence.

I suppose a thanks is in order for the Liberace flamboyant version -- I'm still laughing about that, whatever it was I just saw.  Well, he certainly put on a little show.  

Not that it's relevant, but for me if I want to fool around in that style, I stick closer to the way Jelly Roll played it -- I'm sure lots of people can play all kinds of Tatum copying from the records, especially with all the pre-written transcriptions available nowadays, but it's kind of deflating to try to improvise like him, even though it's pretty reasonable to use some of his RH runs, if you're really making a staple of that style.  Not to mention studying his substitutions, still relevant playing straight bebop.

[cuberdrift]

Oh, I don't remember.  I knew I had the date wrong, but this is the recording I was thinking of.  Somebody made a transcription available online, but I've only glanced at it and don't have a link.

Here's a video showing the sheet of the 1933 version. Some of the runs in this sheet music, however, are a bit inaccurately notated, particularly in the parts where Tatum does his "double-note cluster run" thing (it's mistaken for his triplet runs). The 1940 version you're referring to is probably the fastest and cleanest.

To be more specific, Tatum often plays a certain "triplet run" figuration, say, for example, over C7: Bb-C-G-E-G-C - Bb-C-G-E-G-C, etc. as sextuplets with possibly 3-5-1-2-3-1 fingering. In Tiger Rag, however, there are parts where he does a variation of this run, adding a flat second to every other note, so it becomes B+Bb - G - F+E - C, repeat. This second kind of run is sometimes mistaken for the sextuplet kind of run, perhaps due to extreme speed. Remember, a sextuplet figure in 180 beats-per-minute is equivalent to a sixteenth-note figure in 270 beats-per-minute - that's why this kind of Tatum run is really dazzling, it allows you to play really fast runs without having to move the hand that much.

His little LH ostinato thing comes in at around 0m55s and kind of continues as a little device from there.

Isn't that just a trill? How is that hard?

This time listening to the Tatum, I'm even more impressed than before -- I forgot he actually does more than play lip service to the "tune" (I'm still not sure "Tiger Rag" was ever really much more than a loose idea with some melodic fragments and a loose structure).  Pretty impressive.

Indeed. They say it was his showpiece for his entry into the "stride world" in some sort of contest with Fats Waller, James P. and Wille "The Lion" participating.

I think Fats Waller could have done the straight cliché stride LH (octaves/tenths + chord in middle register) as good, but Tatum really just exploded that convention without losing coherence.

I'm not in doubt or anything, though I'd just like to know if you know any recording where Waller plays those tenths at at least 160 bpm? My experience is limited, but I don't know of any stride player who could play tenths that fast apart from Tatum, aside, of course, from Tatum imitators. I feel Waller possibly could, though.

I like to think of Tatum as a quasi-Impressionist Fats Waller with a Scarlatti temperament. I simply can't associate his virtuosity with the likes of Liszt, who is always the hallmark for "piano virtuosity". The two's styles are very different, Liszt banged his way with his romanto-concert pianist figurations, Tatum I think was way more linear and can be likened to a Baroque improviser with more modern harmony. The latter relies more on runs and ornaments as a way to develop a melody rather than enhance its "power" (can't really find the right way to describe this), as in Liszt. For Lisztian jazz, if there is any such thing, I would agree more to Oscar Peterson, whose virtuosity seems to me much more intentional and inherently "bravura" in nature - he's said to have studied under Liszt's student's student, in fact.

I suppose a thanks is in order for the Liberace flamboyant version -- I'm still laughing about that, whatever it was I just saw.  Well, he certainly put on a little show.

And his left hand gives us a peek at how Tatum would have played in his super-stride performances. I really do like Tatum's super-stride. Sometimes he plays for like an entire two minutes doing tenths at 160 beats-per-minute. I haven't heard many players play that way WITH all the elaboration and development Tatum does melodically, and with the control and swing feel that he achieved.

Not that it's relevant, but for me if I want to fool around in that style, I stick closer to the way Jelly Roll played it -- I'm sure lots of people can play all kinds of Tatum copying from the records, especially with all the pre-written transcriptions available nowadays, but it's kind of deflating to try to improvise like him, even though it's pretty reasonable to use some of his RH runs, if you're really making a staple of that style.  Not to mention studying his substitutions, still relevant playing straight bebop.

"Deflating"? Could you elaborate more on this?

Glad to find another one who shares this interest.

Regards,
cuberdrift

[j_tour]

Here's a video showing the sheet of the 1933 version. Some of the runs in this sheet music, however, are a bit inaccurately notated, particularly in the parts where Tatum does his "double-note cluster run" thing (it's mistaken for his triplet runs). The 1940 version you're referring to is probably the fastest and cleanest.

Thanks for the information.  Also, for pointing out the transcription's problems -- I haven't looked closely at it, but it drives home one of my central theses.  Namely, never, ever trust any transcription you haven't done yourself, or at least checked carefully.  IMHO, if it's worth studying, it's worth doing it the old fashioned way, although it can take a while.  

With audio processing tools now in digital, it's a lot easier than using a reel-to-reel recording slowed to half speed (and dropped an octave), the way I had to learn Ray Charles tunes and stuff as a kid.  It's pretty easy to be so accurate that the older transcription books of jazz, you can see that in those days "close enough" was OK, but nowadays the fastest runs you can basically just write down, including the "wrong" or unusual notes, and worry about the best ways to notate rhythms as accurately as possible, or spend more time on voicings (I don't bother writing voicings out anymore -- I figure if I'm not getting paid for publication, I don't care that much).

Isn't that just a trill? How is that hard?

Well, I guess it's trill-like.  I find that pattern is hard on the independence of the LH 432 (543) finger at that speed.  It's like a big-boy version of the B section of the Bourée of the A major English Suite, LH, on steroids, or as it appears in a tiny part of the G major prélude of Chopin LH pattern that (I'm guessing) most people have to practice a little bit (where the LH moves from the third down to the tonic, stepwise, with roots on G, A, D, C).  Maybe not.  It's not my strongest LH motion, but certainly not impossible -- just something I spend some time improving with the usual scales and made-up patterns.

I'm not in doubt or anything, though I'd just like to know if you know any recording where Waller plays those tenths at at least 160 bpm? My experience is limited, but I don't know of any stride player who could play tenths that fast apart from Tatum, aside, of course, from Tatum imitators. I feel Waller possibly could, though.

I don't really remember any specific recording, it's just I remember having been amazed at how cleanly Fats was able to play various stride in LH at full tempo.  It probably didn't hurt that his hands were likely big enough to hit the wider tenths like Db-F "straight on," instead of sort of having to think about it and stretch like me.

"Deflating"? Could you elaborate more on this?

Sure.  Well, just speaking for myself, when I'm just screwing around with a given style, I like to feel I'm playing stuff I understand completely, not to mention that's not stretching my technique.

No doubt everyone should always be trying to enrich their palettes, harmonically and technically, but I only play these days a few improvised stride tunes if I've had a few or am just kind of bored, although I used to take it more seriously as a young teenager, although mostly just playing transcriptions then.

Even in proto-stride, ragtime, I still take an hour every month to try to keep a few pre-written tunes in memory (I'm happy just to keep a few, like "Solace" and "Kitten on the Keys," in memory, although they're easy to forget, unless you dust them off every now and then), just in case you need a little pre-written tune to fill some time and please the people, maybe at a party or something.

Glad to find another one who shares this interest.

Absolutely!  You're way ahead of me in knowledge of recordings and some of the details, but I think doing jazz without at least some of the basic ingredients is kind of taking a shortcut that, IMHO, probably shouldn't be missed.  Similar to how, I think it was Joe Henderson, I think Herbie's said something similar, that however more "avant" you get in jazz, bebop and playing changes was a great school.  So, for pianists, bebop's still core, but I'd be shocked if McCoy or Chick or Fred Hersch or Bill or Don Pullen or whoever couldn't do a little stride if they wanted that sound.  Certainly Dr. John and James Booker and the New Orleans people made it a part of their rock, R&B, blues styles, when it suited them.

Not trying to disrespect others, because time available and tastes can vary, but there's some value in studying at least a little of most things, even if only to reject it or put it off to another day/year/decade.

[huaidongxi]

Mingus had a definite affinity for pianists who could play from any period of music and come up with a quasi paraphrase to fit the Mingus eclecticism (he was close to the equivalent of a post modern Ellington, who bridged stride and swing) or contribute to the modernist post bebop as needed.  Jacki Byard, Don Pullen both had immense stride piano chops.