M7/I with Haitink/BRSO - sheer perfection for me

[barryguerrero]

I'm generally not a big fan of the first movement to M7. I feel it's often times played too fast. Clocking in at 22:50, I feel Haitink's 'live' BRSO first movement is outstanding - he gets it just right for me. I just wish his third movement Scherzo weren't stretched out beyond 11 minutes. And while the ending of the finale is outstanding on the Haitink, I wish he had kept the tempo up a bit more in the back half of the finale. Still, I may try to find a used CD of it   .    .   .   .   .  .  I feel the first movement is pretty good on the recent Sado/Tonkunstler one as well. F.Y.I.

[Roland Flessner]

Overall this is really a fine performance, well characterized and executed but I agree that the scherzo seems too cautious and the finale would benefit from more momentum. I'll have to listen again to Jansons/BRSO because during my most recent M7 obsession I thought that recording was at the top level.

Two quibbles about Haitink: At rehearsal number 254 in the finale, the bass drum plays two forte eighth notes, a pattern that repeats two bars later. In this BRSO recording the first two strokes are missing. (My score is the Eulenberg which may be less authoritative than the UE or newer edition.) Then, in the penultimate bar, the trumpets and low brass are supposed to start fortissimo then diminuendo to piano. This is one of Mahler's jokes, because we've endured several false endings already and it sounds like he's going to trail off and do it again, then ends with the "stinger" in the last bar. I can't think of any other conductor who ignores this direction except for Haitink. In his old RCO recording he starts the brass quietly then does a crescendo, the opposite of Mahler's instructions. In later performances, such as one live here in Chicago and on this BRSO edition, he plays a kind of sforzando, loud at first then somewhat quieter, but not a diminuendo. He seemed to have a bee in his bonnet about those last two bars and the musical logic escapes me.

[erikwilson7]

Roland, I've also noticed how Haitink handles the last couple of bars in his old RCO recording. I think the only other times I've heard it done similarly are both of Bernstein's versions (Sony and DG). The difference is that Haitink's is a real sforzando, whereas what Bernstein's trumpeters do in both of his recordings are more of a 'bowtie' diminuendo-crescendo into the last note. As we both know, neither methods are what Mahler actually wrote.

It's less successfully pronounced in the earlier DG recording; it almost sounds like the trumpets are a car whizzing by ― a sort of doppler effect. In the later DG recording, we get to hear much more of the bowtie dynamics I mentioned.

[Roland Flessner]

Thanks, Erik, good observations. I still have the Lenny Sony issue but the DG recording went away in a shelf cleaning. I never really cared for either one.

M7/V has to be really tough for a conductor and orchestra to pull off with such frequent changes of tempo, meter, and mood. After Haitink/BRSO last night I spun up Inbal/Frankfurt in V and it's outstanding, establishing and maintaining a narrative line start to finish despite all the kaleidoscopic changes along the way. The recording is foggy in the bass, the upper woodwinds would benefit from more presence, and chimes are used instead of real bells, but those aside, it really is a compelling performance.

[barryguerrero]

Great observations, guys. Thank you.

[Roland Flessner]

Barry, I have to thank you for dragging me in to yet another M7 obsession. Just kidding; for me this is the meaning of life and I was overdue.

I've been surveying available recordings and will offer a few observations. I started out by concentrating on the finale since it's always made the least sense to me. Now after all these years, it seems structurally solid, in a good performance. In some cases I've listened to most or all of the work, sometimes in order, sometimes not.

The performances that annoy me introduce tempo changes that Mahler didn't indicate. In the score, the tempo modifications are usually quite subtle. In a work that is inherently episodic, keeping steady tempos is essential. One might say that not many roses line the path traced by this symphony and we have no reason to stop and smell them.

When I go through one of these surveys, I rarely have a radical change of view but this time around I've noted some surprises. Some may admittedly owe to  memory that has not improved with age.

First, I'll note the excellence of the Kubelik/BRSO studio recording. His first movement is the only one I'm aware of that clocks in at under 20 minutes. It doesn't feel rushed but conveys a welcome sense of urgency. While the SQ is dry and dated, it is also unusually transparent and the woodwinds speak with a color and presence that embarrasses many newer recordings. The seating arrangement helps too, with antiphonally divided violin sections and low strings on the left. The remaining movements are in the same vein. Nachtmusik I is often played too slow (it is marked Allegro moderato), but moves along smartly here. And since it's a march, steady tempos really help it to work. The Scherzo is just terrific.

One surprise was how much I like Haitink/RCO, aside from that notorious crescendo in the penultimate bar. Like Kubelik he keeps the tempos moving along and applies a light touch with changes. The recording while dated is also detailed and transparent. Even the Scherzo is excellent, perhaps surprising because Haitink tended to mute the demonic elements in Mahler's music.

I'm not sure I had ever listened to Kondrashin/Leningrad all the way through. It's another standout, provided you can acclimate to the congested Soviet-era sonics.

I tried two movements of MTT/LSO (RCA) but went no farther because the SQ is nothing special, tempos are stretched around, and it doesn't feel contrapuntal.

I haven't listened all the way through the Boulez/Cleveland but the Finale is quite good. Perhaps the Emil Berliner remaster helps the SQ; on the original issue, it lacks presence and bite, as though we're listening from the lobby.

The Ozawa is not one of his better performances. He was usually among the most straightforward of conductors but frequent ritards repeatedly kill momentum. It doesn't help that the trumpets, and to a lesser extent the horns, tend to play a phrase where each note is accented as a smooth legato.

Levine/CSO is a very strong contender, with one caveat. He maintains momentum and the recording is better than many of the others in the set. The soft strokes on the tamtam in Nachtmusik I are wonderful, in the deep bass as they should be. My concern is with a too-slow Nachtmusik II clocking in at over 14 minutes. He doesn't mess with the pace but for me this movement works better at a more flowing tempo even if I'm not sure I'm on board with Boulez's breezy 10:38.

Jansons/BRSO is one of the strongest performances though the bells in the Finale are underwhelming.

I need to spend more time with Nott/Bamberg but his Finale is outstanding and in excellent SQ, although in that penultimate bar he plays a late-Haitinkesque sforzando rather than a diminuendo. Like Kubelik he seats the orchestra in the common 19th Century arrangement.

I'll be spending more time with these and a few other recordings in the coming days and will offer comments if I come up with anything that might be interesting.

[barryguerrero]

Thank you for the interesting observations.