Wolfgang Amadeus MOZART (1756-1791)
Bruno Walter conducts Mozart: the Pre-war Recordings
Die Zauberflöte overture, K588 (1790) [6:45]
Le nozze di Figaro overture, K492 (1786) [4:08]
La clemenza di Tito overture, K621 (1791) [4:19]
La finta giardiniera overture, K196 (1775) [2:14].
Three German dances, K605 (1791) [5:35]
Piano Concerto No. 20 in D minor, K466 (1785) [27:37]
Symphony No. 38 in D major, K504, Prague (1786) [22:32]
Symphony No. 39 in E flat major, K543 (1788) [25:37]
Symphony No. 40 in G minor, K550 (1788) [23:10]
Symphony No. 41 in C major, K551, Jupiter (1788) [27:21]
Eine Kleine Nachtmusik, K525 (1787) [15:03]
Requiem, K626 (1791) [59:23]
Elisabeth Schumann (soprano), Kerstin Thorborg (mezzo), Anton Dermota (tenor), Alexander Kipnis (bass)
Vienna State Opera Chorus
Berlin State Orchestra, BBC Symphony Orchestra, British Symphony Orchestra, Mozart Festival Orchestra, Vienna Philharmonic Orchestra/Bruno Walter (piano:
PRISTINE AUDIO PASC564 [3 CDs: 223:44]
There are two ways you can assess a set like this: the music-centred, and the artist-centred. The music-centred compares Bruno Walter’s early recordings with those of contemporary conductors. The artist-centred compares Walter’s early recordings with those in which he returned to the same works. You may personally prefer one of these approaches, so I’ll do a bit of both to try to please everyone. First, as noted on Pristine Audio’s inlay card, not all Walter’s pre-war Mozart is here. Four items already issued on Pristine are omitted: the 1924 recordings of the Cosi fan tutte and Idomeneo overtures (on PASC 142) and 1928 recording of Le nozze di Figaro overture and 1931 recording of Eine kleine Nachtmusik (PASC 142).
I’ll start this set from CD1 track 1, so you can feel its varied mix in presentation: track titles and recording locations I list after this review but give dates within the review. The 1928 Die Zauberflöte overture with the Mozart Festival Orchestra, while spirited and featuring merry woodwind exchanges and lively Mannheim crescendos, is a bit too scrambled swashbuckling in its articulation of fast chords. I compared it with Walter’s 1961 recording with the Columbia Symphony Orchestra (Sony MYK 37774 now licensed as a Presto CD) and find it has more poise and keener sense of the portentous drama, if less lightness in the faster violins’ passages. On the other hand, the 1932 Le nozze di Figaro overture with the British Symphony Orchestra compares more favourably with Walter’s 1961 disc. A good test of any performance is that the first tutti should be really arresting and the 1932 performance is pretty good at this, where the 1961 is a touch stiffer, just as its introduction is more reserved than the frisky 1932. This latter is a terrific account: pacy, good sforzandos and clarity of line. Timing at 3:59 to 1961’s 4:39, it is no wonder that 1961’s sforzandos are comparatively leaden. I wish in 1932 Walter hadn’t made the exposition codetta theme (CD1, tr. 2, 1:26) a bit over-leisurely, but he does the same in 1961.
Next, we come to the first of many appearances of the gold standard in orchestral playing then, that of the Vienna Philharmonic in a 1938 account of La clemenza di Tito overture. This has spirit, dramatic edge and pace, first in the strings and then the woodwind, plus a particularly exciting close and all in splendidly sonorous sound. La finta giardiniera overture, also from 1938, is brief but by turns hearty and nifty and here the Vienna horns have more chance to shine. Then comes the lighter Mozart in three German dances recorded in 1937, of which the third, the Sleigh ride, has the most character. This has spruce loud sections, while its quiet sections feature the charm and delicacy of sleigh bells and a beaming posthorn solo, the latter returning in the coda and in 1937 quieter, as if the procession is now in the distance. There’s a 1954 Walter recording of these dances (Sony G010004059731J, download only available in the UK) with the Columbia Symphony Orchestra which is just a touch stiffer except for a warm posthorn. This doesn’t get quieter at the end, which is the only aspect of this later account that I prefer to the 1937.
We stay in 1937 and with the Vienna Phil, but what a contrast in tone with the orchestral introduction to Mozart’s Piano Concerto 20 in D minor. The first violins, though soft, are rich and baleful and once all the violins take up the rising semiquavers and then demisemiquavers the tone becomes more waspish. The second theme (tr. 8, 0:58) attempts normality, but the pace rightly doesn’t allow this, nor the tutti protest, during which Walter emphasises its forlorn melody and especially the pause (1:58) as expression itself seems to break down in crisis. Then comes a codetta to this second theme (2:07) in which Walter is a little romantically indulgent. The piano now enters with the third theme (2:18), a lament of beauty in which Walter doesn’t indulge, giving it a chaste objectivity, but before long in running semiquavers he gets involved in the orchestral alarm. However, when the second theme returns, his semiquavers are brusque, perhaps to clear the path for his solo introduction of the fourth theme (3:44), whose tripping quality is soon taken up in enjoyment by the orchestra. For the first time there’s a period of contentment. This even affects the return of the third theme until the orchestra resumes its introduction’s brooding and motifs. We get a succession of piano pleas and orchestra rejections which could be creepier, for instance the sudden fortes at 5:28 and 6:11. Similarly the change in the soloist from a piano which looks like it might be submission to a defiant forte (7:08), here is barely mezzo piano. This is the moment of the orchestra’s recapitulation of the first theme and when Walter the pianist later joins it, I feel he ought to be more combative. The recap of the third theme (8:08) he takes a touch slower, which makes it less confident, while that of the fourth theme, now in the key of D minor (8:32) has become forlorn. Walter plays the cadenza by Carl Reinecke (1824-1910). This begins mulling over the third theme, elaborating and playing around with it to end rosily, so Reinecke can then waltz into theme 4, weave intricate patterns around that and then some more haunted recollections of theme 3 and a descending chromatic scale to finish off. You remember the frilliness, not the drama, though, as generally, Walter underplays the latter anyway.
I compared this with piano-conductor Edwin Fischer with the London Philharmonic in 1933 (APR 7303). The difference here is that Fischer is first a pianist, whereas Walter is first a conductor. Accordingly, it’s not surprising that Fischer’s third theme is more ruminative and inward, more assured than Walter’s and his recap of it more affectionate. Fischer’s fourth theme is also more considered and rounded, while at the same time hopeful in its lighter articulation. Walter as conductor makes that pause in the first presentation of the second theme more telling, but with both Walter and Fischer I feel there could be more bite in the louder orchestral passages. However, it’s Fischer who is better at the sudden fortes I comment on above. Fischer plays his own cadenza. You notice its different, more modern harmony as much as the showcasing of theme 1, then lots of passage work before alighting on the joy of theme 4, attractively cast in the baritone register for the repeat of its opening two phrases. After this the cadenza becomes more dramatically Beethovenian.
The second movement, in B flat major, entitled Romance is also a rondo and, while there’s no tempo marking, this means it shouldn’t be slow because, while the opening piano solo is a touch reflective, the orchestral repeat shouldn’t be laboured. Of course, Walter the conductor appreciates this, so the tempo he sets as pianist is good, allowing a certain poise but also smoothly flowing and his presentation is limpid. The movement offers escapism to an idyllic past, but also with the poignancy that the troubled present is ever lurking. For instance, the discord made by the rondo theme’s penultimate note (tr. 9, 0:21), often in Mozart an appoggiatura but here unusually sustained as a minim: compare this with the one occasion when it’s absent (1:29). The first episode (1:53) is a growingly searching, aria like expansive musing. The second episode (4:09) is a panic attack in G minor, its animation achieved by the sudden use of continuous semiquavers in the piano. For proper emphasis, both opening strains need to be repeated as marked but, unfortunately, in Walter’s recording, repeats aren’t made. Another irritation for me is that the forte chords which punctuate the episode aren’t as crisp and biting as they should be, the opening one being the shock of panic. All the same, when the rondo theme returns, Walter’s gentle simplicity is a welcome relief.
Fischer’s opening rondo theme solo manages to convey both an absorbed musing and blithe simplicity. With the LPO orchestral repeat there’s an airier, less de-luxe crafted feel than with Walter’s Vienna Phil. In the first episode Fischer emphasises the gracefully melodic and sunnier aspects: I prefer Walter’s clearer attention to the more clouded elements. Fischer is preferable in the second episode as he does repeat the first and second strains and has more trenchant forte chords. His final rondo theme is balmily relaxed.
Where the second movement is a simple rondo, the finale is an irregular sonata rondo, with a second theme as well as rondo theme, a first episode of two themes and a second episode development. This creates a movement which is a complex whirl of experiences, not least because they happen very fast, the marking being Allegro assai. Walter the pianist takes this on with admirable dexterity and the Vienna Phil match him, but for me the sheer adrenalin takes away something of the ferocity and grit the restored home key of D minor should also have, for instance in the furious orchestral extension of the piano’s opening with the rondo theme. The second theme (tr. 10, 0:50) is a cousin of the third theme of the first movement, pleading for more tolerance, but it also has a dancing quality that’s subversive, coupled with an uncanny ability to keep popping up. Here Walter gives it for me a peremptory quality, as if here’s an underlying sadness that must be brushed aside in a hectic search for pleasure. The first episode first theme (1:13), in F minor, Walter does very well, with a brief illumination of pain and then the high flute (1:25) breaks through its barrier to sunlight, cue the first episode second theme (1:50), a jolly woodwind one in F major. Thus, the concerto’s endgame is foreshadowed. The second episode (2:23), the movement’s development, elaborates the rondo theme, recaps the second theme, rondo theme again further developed and second theme again, culminating in a dialogue between oboe, flute and piano, done with delightful friskiness by Walter and the VPO. Now comes the recap of the first episode first theme in the paler light of D minor (3:36), as is also its formerly jolly second theme (3:59), but Walter’s consistent pace and piano’s continuous quaver runs, maintains the stimulating quality of a winter sprint. Another cadenza must resolve this, Walter with Reinecke’s again, little more than a sequence of recalls of the rondo theme’s first phrase to create an airier, bubbly atmosphere before the last appearance of the piano solo rondo theme, whereupon the tutti coda produces a skipping celebration of the first episode second theme in D major. By now long since I’ve warmed to Walter’s pace.
As it happens, Fischer adopts much the same tempo. It doesn’t look like it, with his timing of 7:11 to Walter’s 6:08, but Fischer’s own cadenza takes 1:14 whereas Walter’s one by Reinecke takes 0:24, leaving the respective timing of pure Mozart: Fischer 5:57 and Walter 5:44. Fischer’s piano opening is cooler, more inward-looking, yet his following orchestral tutti has more bite than Walter’s, though this is somewhat diminished by the smaller body of the LPO than the VPO strings. Fischer’s articulation, crisper than Walter’s, gives his playing more edge. Fischer’s piano solo second theme is at first more fractious than Walter’s, in its second solo appearance initially more reflective, but then thrown off. Fischer’s first episode first theme is cooler than Walter’s, though it gets a concentrated gaze. Fischer’s first episode second theme maintains tension with a more wilful merriment which has nothing of Walter’s unbridled joy. Fischer’s cadenza brings more weight and bombast than hitherto with a sardonic recall of the first orchestral tutti, but then a poignant one of the third theme of the first movement before using a fast version of the finale’s ‘merry theme’ in the lead-back.
Now come the late symphonies. Fluency is the characteristic which strikes me most about Walter’s Vienna Philharmonic Prague Symphony from 1936. Walter’s introduction is contrasted more in melodic than dramatic terms, so you note the sonorous tutti and sweet violins. Clarity of the density of the orchestration is a notable feature and says much for the quality of Mark Obert-Thorn’s restoration. Weight isn’t something you note much after the introduction. Rather in the Allegro main body of the first movement it’s the shine and deftness of Walter’s tingling, tensile strings, even when soft much of the time, their nimbleness and clean articulation. The whole movement rests on the ingenuity with which motifs flow from instrument to instrument within the frame of a continuous spurring accompaniment. To give a micro-example: the first violins descend (tr. 11, 3:51), the violas provide a lively accompaniment (3:52), the second violins answer (3:54). The second theme (4:20) is appropriately comelier without any need for unmarked contrast of tempo because Walter’s violins’ shaping is so sure and the lightness when they finally reach high register lovely. A particular feature of this theme is the luxuriant spicing obtained by Walter when Mozart sets its second strain (4:31) in the minor, as if to say this is a theme for all seasons. As usual in this period and throughout these symphonies there’s no exposition repeat. In the development Walter brings more verve to the argument’s progression and clarity to the independent features different instruments show at the same time. Loud playing is incisive without being oppressive.
I compared this with the London Philharmonic Orchestra conducted by Sir Thomas Beecham in 1940 (Warner Classics 9099462, download only available at the time of review). His timing for the movement is 10:26 against Walter’s 9:22 and a good deal of the distinction comes from his slower introduction, 3:31 against Walter’s 2:54. His more spacious Adagio has more character: a dramatic confrontation between masculine tutti and wheedling first violins, neither yielding. His sforzandos have more of a shiver than Walter’s, the latter’s from 2:14. Beecham’s tutti are more incisive but this halts the progression a shade. This is an account more in heroic mode which has remained very much the fashion for interpretation, but I prefer Walter’s lightness and sheen which is surprisingly closer to historically informed performance. Beecham turns on the sinuous charm for a sexier second theme yet makes heavier weather of a more rugged development. Admittedly his peroration is more exciting.
The Andante slow movement from Walter brings the opportunity to hear Mozart’s development of themes at a more measured, graceful tempo. It’s like watching two dancers in accord: one proposes and the other responds. The first proposal by the first violins, repeated by the tutti, suggests this is going to be a relaxed affair, but straightaway all strings respond with a motif of five quavers adventuring ever higher through five repetitions before resting on their laurels then gliding down. Then the first violins’ proposal, the second theme (tr. 12, 1:04) is more substantial, tersely stern and melting in successive phrases, to which the response is lolling in sequences of running semiquavers (1:31). The first violins now propose a third theme (2:02) and we’re into an exposition codetta which incorporates both rising five-note and six-note motifs, falling five-note ones, varied semiquaver patterns, tension and resolution. All this with an appreciable blend of distinctive wind contributions glowingly realized by Walter. The development features the semiquaver patterns continuing and the opening theme in an autumnal contentment. But then the motif of five quavers finds a dark corner, for the first time turning up loud (4:16) and ominously from Walter, and the opening theme turns into mourning in D minor. But thereafter for me there’s something of a blip, as the loud five quavers (4:37 and 4:54 onwards) lack conviction from Walter. All the continuing quaver patterning then seems rather square in the transitional passage which needs to be suaver, because this is the sleight of hand that gets Mozart back to the recapitulation in G major.
Beecham, timing the Andante at 10:06 to Walter’s 8:44, is perhaps closer to Adagietto, but this is effective in creating more space and flexibility of line and clarity of orchestral detail, though some may feel he hangs about a little too much. Beecham thus brings more charm than Walter. His contrasts in the second theme are more dramatic and here I prefer Walter’s quieter approach. But Beecham’s third theme has the quality of longing which I find more attractive, as I do his quieter treatment of its later tension and relaxation. His D minor theme is sad where Walter is mournful, but he avoids Walter’s blip by keeping all the loud five quavers firm and purposeful and then becoming smoother for the return to the softer adventuring approach as at the outset.
The Prague is the only late Mozart symphony without a Minuet and Trio, but its finale is rather like a Scherzo, anticipating by 16 years Beethoven using that term in his Symphony 2. Walter treats it like one, beginning light and fairy-like in the strings, then a boisterous tutti of demons. The jolly second theme (tr. 13, 0:46) could be heard as a quietly bubbling Trio and its lovely contrast of scoring for strings and woodwind, whether separately or together, would grace any Trio. The first ‘Scherzo’ theme then returns and, in the codetta, (1:23) the first violins’ triplets in quavers propel the exposition to a conclusion. The development (1:45) starts with a sustained loud call to arms which then alternates with the movement’s opening theme before that goes through some assertive transformations to confirm this is also a finale with a sense of purpose and emphatic closure. And where you think there’s a recap of the first theme (2:29) and it’s the only one, Mozart immediately plunges you back into the development’s fusillades. So, the work and Walter end very much looking towards Beethoven.
There’s little to choose between Walter’s and Beecham’s finale in terms of tempo and quality. Beecham’s is daintier and more amiable, especially in the second theme which he makes more swinging by emphasising the first beat of the second bar in both its phrases, but I like Walter’s more fairy-like opening and bolder tuttis. Beecham’s development’s fusillades are more stimulating where Walter’s are momentous blasts.
We’ve reached CD2 for Symphony 39, Mozart’s last with a first movement introduction. This has two points of interest. First, the contrast between the sturdy tutti and the gossamer violins as they gently fall in semiquavers and rise more expansively; second, the increase of tension in the later tutti dotted rhythms offset by shadowy rising semiquavers in violas, cellos and double basses. Done well, the effect is very spruce, but Walter with the BBC Symphony Orchestra in 1934 isn’t at his best here. His violins are fine but the tuttis too heavy and the later tension is underplayed. The point of the introduction is to allow the Allegro main body to begin in total relaxation: you couldn’t get a more lyrically indulgent opening theme. Only Mozart could have written it and Walter lets it glow, but it’s also a preface to the robust tutti you’d expect and can enjoy with Walter. In refreshing contrast, the simplicity of the second theme (CD2, tr.1, 4:01) couldn’t be bettered: a short rising and falling phrase, then a high, pirouetting phrase on violins and violas as clarinets and bassoons repeat the first phrase. This is in turn swept aside by an invigorating exposition codetta, a ghostly echo of whose final phrase begins a development which is grim but short-lived. In the recap Walter’s violins’ descents are scintillant, the second theme attractively dreamy and a good head of steam to the coda.
Again, I compared this with the LPO/Beecham, 1940 vintage. His faster introduction, taking 2:43 against Walter’s 3:00 I find more effective in the edge, incisiveness and momentum of its tuttis without detriment to the beguiling sleekness of its violins’ descents. Beecham brings a quieter radiance to the first theme, though you may not care for his violins’ portamento. His tuttis are more vivacious than Walter’s while his second theme, finely phrased, is suitably deft.
The slow movement, marked Andante con moto, isn’t particularly slow and Walter begins it as a soft, genial strings’ dance, its opening theme warmly flowing with everything in courteous balance. The repeat of its first strain is made, but not its second, so we’re more quickly plunged into its loud second theme in F minor (tr. 2, 1:54) when the first violins become tetchily demonstrative while still maintaining precise pointing. The clarinets and bassoons try to mollify them. They refuse, yet Walter slows down a touch, awkwardly deliberate at 2:38 as if they are reluctant, leaving space for more woodwind entreaties to succeed and the return of the first theme now tenderly adorned with counterpoint across the community. However, that second theme returns more desolately in B minor (5:17), as if the first violins are suddenly possessed, yet somehow with Walter you don’t feel their heart is in it. The same dialogue recurs and the same irresolution in the first violins. With Walter the others’ amelioration is now slower, more painstaking, longingly hanging on, with roseate clarinets in particular, to the possibility that all may now progress well. The ambivalence of the movement’s progress is acknowledged.
Again, Beecham is faster, this movement taking 7:47 to Walter’s 8:45 and I prefer this approach. His first theme is less warm than Walter’s but stylishly phrased. His second theme is more dramatic and alarming, not so much in terms of dynamic as caustic attitude. Beecham keeps up the tempo in the first violins’ rejection of the early pleas and places the focus of the change of mind on the violins’ musing of afterthought (2:51 in Walter), making it an active decision. Beecham’s second theme return is similarly starker, yet his woodwind close more halcyon, though the shadows in the string bass remain clear.
The Minuet from Walter is sturdy yet also sprightly as its opening tutti swings along confidently, then its strings’ response has an ingratiating charm. Walter’s Trio is all mellifluous innocence, a gentle clarinet solo opening and tender violins’ response. Beecham’s Minuet is sturdier with heavier accents and thereby not quite as dashing, in both senses, as Walter’s, though Beecham brings out more the inner detail. Beecham’s Trio is sweetly melodic but misses Walter’s insouciant smoothness.
The finale is marked Allegro and I’d suggest Walter takes it at Allegro molto, yet it begins light and fluffy and Walter succeeds in making it all like that; even the loud tuttis are a touch understated. I like it because I appreciate the quality of both the strings and woodwind articulation, which manages to be both precise and joyous. Mozart gets away with using only one theme, maintaining the variety and momentum with endless re-scorings of the opening six semiquavers plus one quaver descending motif. He also has two ways of varying the theme. In the first, where you’re expecting a second theme (tr. 4, 0:33) the violins dialogue with the woodwind and then go into their own dancing flight of fancy, beautifully gauzy from Walter, while the woodwind hold the core motif. In the second ’variation’, which happens in the development, the clarinets and bassoons introduce a chant (1:54) to accompany the motif, soft and suitably monastic here.
Beecham’s finale is also nifty, taking 3:42 in comparison with Walter’s 3:34, but he sounds a touch more measured in a sheer determination to be athletic which, though finely crafted, is less cheerful than Walter. Beecham’s dynamic contrasts are stronger and tuttis heavier, full of bonhomie. His second ‘variation’ is an opulent pearl, glowing unabashed.
Symphony 40 was Walter’s first Mozart symphony recording, made in 1929 with the Berlin State Orchestra, today the Staatskapelle Berlin. But don’t think of it as a youthful endeavour: Walter was 53. To me this has something of the manner of an acoustic recording about it: a small body of strings, though tight ensemble, emphasising the plaintive tone of the soft opening. This is an advantage, because the symphony itself is sparely scored: no trumpets, no drums and no clarinets until Mozart gave them the plums of the oboe parts in his revised version played here and in most recordings. This movement is a lament, fairly uncommon in symphonies and after Mozart not at Molto Allegro, but his pace makes the grief all-encompassing, from which there’s rarely escape. The minuscule manner of the first theme in Walter’s account has something of the quality of an improvised folksong. The first loud tutti is just firm punctuation, but at the second (tr. 5, 0:36), the second part of the opening theme, Walter makes an upliftingly fuller sound for its extended melody, of hope and fighting spirit. The second theme (0:55), in B flat major and divided between strings and woodwind, Walter makes nonchalant and rather dreamy, as if escapism, until the strings running quavers resume, yet even then the wake-up climax is ambivalent and the woodwind’s focus on the head motif of the opening theme is shared with first violins’ sighs that seem to remember a happier past. So, Walter’s end of the exposition, for all its rigorous descending quaver runs, retains a fighting spirit. The development immediately makes that more of a struggle. The first theme now comes downcast and the fight is in a more straitened atmosphere. The recapitulation, normally a relief, now confirms ever present gloom and the recap of the second theme, also as normally now in G minor, is one of exhausted desolation, yet nevertheless has to summon up some steel for the following climax. After Walter’s urgent, rising chromatic scale in the next and final climax, the coda’s layering of the strings, in turn second violins, first violins and violas, in a gazing, reflective first phrase of the opening theme from 5:52 is particularly expressive and poignant.
I compared this with middle-period Walter conducting the New York Philharmonic in 1953 (Andromeda ANDRCD 5048). That is a more stylish, extrovertly dramatic, flourishing and full-toned account, fluent and appreciably shaped but also less personal. It doesn’t capture the inward sense of grieving that the 1929 account does. Nevertheless, the climax after the second theme is heartening. The theme in the development is more forlorn but is more set aside by the vigour of the following contrapuntal struggle. The first theme recapitulation here is delivered with enough nuance to create relief and the second theme recap is duskily subdued rather than wan. The final climax’s chromatic scale is powerful but not as focussed as in 1929 and the 1953 coda’s strings’ layering is smoothed over.
Walter’s account of the Andante slow movement makes me think of a lullaby. Its opening theme has three aspects. First a head motif, a measured tread sustained by repeated notes and layered between the upper strings; second, a three-note sighing fall and then rise; third, a perky rising of two demisemiquavers followed by two falling quavers. On the repeat of the first aspect Mozart adds a gentle but expressive descant in the first violins, exquisitely done by Walter, after which the third aspect’s demisemiquavers are extended. In a second phase of the first theme (tr. 6, 1:22), a loud chord makes sure mother isn’t nodding off, but soon tails off into more demisemiquavers. I see these as a maternal concern to keep everything tidy. Anyway, those demisemiquavers now combine with the head motif in the strings and are then especially winsome when exchanged between flute, clarinet and bassoon. A second theme (2:24) is little more than an intermezzo-like relief headed by a sighing figure, so you can consider it a kind of variation of the first theme’s second element and, like that, it’s gate-crashed by the demisemiquavers; but it’s also, in Walter’s account, affectionate contemplation which grows into rich satisfaction. The development (3:36) casts a dark shadow over this. The head motif and demisemiquavers now combine and alternate between strings and woodwind, Walter showing them standing firm with immense resolve, the determination more tensile than the dynamic. And suddenly this all melts into a now appreciated serenity of the recapitulation, Walter’s focus just a touch softer than before with the second theme now presented with the beauty of a clear, pearly focus. This Andante isn’t easy to judge: too fast and the demisemiquavers seem skittish, too slow and the head motif chugs: for me, Walter gets it just right.
Walter’s tempo in 1953 is about the same. His opening aspect is less tender than in 1929, but still expressive. His violins descant in 1953 is more musing, his opening of the second phase more finessed: no waking up with the loud chord. His 1953 demisemiquavers are more exquisite with a hint of caprice. His second theme is more reflective and the rich end of the exposition more shadowy, which anticipates a more sinister, dramatic approach to the development which is a threat that nevertheless quickly passes, brushed away by those capricious demisemiquavers. His second theme recap is still beautiful, but the grace notes at the end of its phrases; compare 7:17 and 7:20 in the 1929 recording, and generally are treated in a more mannered fashion in 1953.
I think the purposeful, rugged cast of the opening of the Minuet really needs more strings than in Walter’s 1929 recording, but in its second part the taut interplay between strings and upper woodwind is well conveyed, with much of the tensile quality shown in the development of the slow movement. The Minuet’s coda (tr. 7, 0:55) is another Mozart conjuring trick in which the friction melts away and the woodwind are used to celebrate melodic beauty as Walter emphasises the cheer with a flute part which adds an extra G, the first time at 0:59, to bring more zip. Walter takes the Trio a touch slower, not marked, but this does bring to it a feel of innocence rather more than serenity.
Walter has more strings in 1953, but the opening of his Minuet then has a little more swing than rigour. Curiously, the entire Minuet is tauter when it returns after the Trio. The Minuet’s coda in 1953 has less colour than 1929 and no extra flute note spurs. The 1953 Trio is presented at the same tempo as the Minuet, lucidly and liltingly with a pleasant open tone, but the horns over dominate the second part.
The Allegro assai finale is excellently paced and articulated by Walter in 1929. It falls short in nothing except fullness of sonority and venom. Unfortunately, these shortcomings blunt the closing impact of the movement and symphony. The opening is good, with its continual contrast of the light, gracious and smiling manner with loud, curt rebuttal. Because of the speed of interchange this feels not like two characters in dialogue but a schizophrenic and, once the quaver runs start, the torment is inescapable. The second theme (tr. 8, 0:54) which Walter here makes snaky, coy and cajoling, is more complex than its equivalent in the first movement, also in B flat major. He vamps it up by adding appoggiaturas to the first two notes as well as the third and fourth. The running quavers latch onto this theme too and the mood now becomes one of manic devil-may-care. The development (1:38) starts from Walter here in comic opera fashion, after whose rhetoric the first theme entries are first whimsically treated, until the serious intent of the loud sustained notes on the horns from 1:56 and later all the woodwind, all of which should be louder here, and the combination of first theme and running quavers, can’t be ignored. The second theme recap, now like that in the first movement in G minor, is bereft of cheer and Walter’s earlier extra appoggiaturas. When in the coda the flute, oboes and bassoons’ rising scales at 3:52 and 3:55 sequence those of the violas, cellos and double-basses the effect should be the scare of a wave-crest emerging; but alas it’s hardly more than a mellifluous ripple.
Walter’s 1953 finale has an abundance of venom that 1929 lacks but misses out on its charm, so of the two I prefer 1929. At a timing of 4:46 in comparison with 1929’s 4:04, the flow, particularly of the quieter aspects, has less character in 1953, while the loud passages in 1953 are brutal and abrasive. The beginning of the development is all spleen, but the sustained notes on the horns and woodwind strike home. On the other hand, the coda’s wave crest is submerged by the whale like presence of the horns.
Symphony 41, the Jupiter, in the set under review was recorded by Walter with the Vienna Philharmonic in 1938. It begins with a real conversation: a rather formal, dogmatic man here from Walter to start, then a delightfully unfazed, cheerful lady who herself becomes more assertive and, by her second theme (tr. 9, 1:34) positively merry. When the man has a later attempt at domination, her cheeky third theme (2:33) is like an extended titter. Click this link to the first movement and you can hear it all for free, courtesy of Pristine Audio. Enjoy the sneaky, quiet and false recapitulation of the first theme (4:04) and, after the real one, the further development of the themes from 5:22 which feature more conciliatory, then again assertive, positions as the couple agree to disagree.
I compared this with late Walter, conducting the Columbia Symphony Orchestra in 1960 and in stereo (only available in the UK as a download, Sony G0100040942846). He’s rather steadier for an Allegro Vivace twenty-two years on, taking 8:51 to 1938’s 7:54. Walter’s 1960 gentleman at the opening is beefier yet also more urbane and his lady responds with class and style, though I miss her 1938 counterpart’s greater sense of fun. But in 1960 the lady’s second theme is sleek and silky in turning on the charm, though her third theme is little more than a decorous response to the gentleman’s splendid preening. I like the greater weight and clarity of the timpani, but the power here is just for show.
I suggest the couple’s conversation continues in the slow movement but its qualification cantabile to the tempo Andante suggests the lady will triumph here. The use of muted violins also ensures that it’s the quiet manner that celebrates and secures calm. There’s always an issue with this movement regarding how loud the forte at the end of the first theme’s first and second phrases and the beginning of the fourth phrase should be because this is the start of much interplay between loud and soft, the former of which you can think of as the man trying to have his say. Personally, I like a gentle nudge rather than a poke in the ribs and Walter here kind of half does this, though his end of phrases’ fortes don’t really sound as short as they’re marked. What he rightly makes more memorable is the musing violins’ and sometimes woodwind elaboration. The ongoing drama of the movement is that, while lingering in the beauties of relaxation, there are also clouds on the horizon to be faced, but it’s the beauties that hold centre stage. The second theme (tr. 10, 2:07) is a hymn of stalwart consolation, yet its second part (2:27), first heard on the first violins, later echoed by the flute, is the most intimate and delicate of caresses. The exposition codetta (3:00) interchanges rising violins and falling flute and oboe, the latter backed by rising bassoons, all of which creates a sense of aspiration followed by fulfilment. All this is sensitively caught by Walter. In the returning threats of the development he underplays the forte markings and tension, but I like the poignant sadness that this reveals and paves the way smoothly for the return of the exposition codetta exchanges (4:20) to cast away all cares. This movement, like the first, has a false recapitulation (4:44), because the violins’ elaboration now in demisemiquavers turns more alarming. This could well be interpreted as a Tchaikovskian sweep, but Walter consistently understates because the second theme is now to be recapped trouble free before the true first theme recap heads the coda. The density of the movement reinforces the secure nature of its contentment.
Again the 1960 Walter is more leisurely, taking 9:02 against 8:03 in 1938. The cantabile is now more beautiful, but also more manicured. Take the third note: it’s a double-dotted crotchet and, within that first phrase of the opening theme, Walter is arguably more accurate to give it more space, but I find his warmer flow in 1938 more effective. Some things are better in 1960: those ends of phrase fortes are a crisper nudging, the clouds on the horizon in the exposition are more of a presence. But generally, it’s the 1938 account that has the more direct and emotive violins’ elaboration and more warmth and focus to the second theme. While in 1960 the key passage of those clouds in the development has more edge and drama, it lacks the personal impact of sadness of 1938.
The conversation continues in the Minuet and Trio with the man trying more subtlety. The lady begins with a sleek drawl that the man politely acknowledges, then he’s more, but still genially, assertive in the tutti second phrase and still more in the second strain. Walter brings an attractive, buoyant lilt to everything. The Trio is started quietly by the man and the lady then dances cheekily around him. As the man parades his strength again in the second strain (tr. 11, 2:31), the first four notes of his loud theme become, with a change of key, the soft, benign opening motif of the finale. The lady had won the Trio’s discourse by echoing the man’s last two notes and then sequencing them a tone lower (3:03) and all is play again. Walter handles all this delightfully.
To the Minuet’s Allegretto in 1960 Walter again brings more measure, timing at 5:04 to 1938’s 4:23. This makes for a rather more considered conversation: the lady graceful yet more distanced, her Trio’s dance more neatly formal, while the man is ever more flamboyant. Walter’s dynamic contrasts are fuller, so the man’s Trio second strain is more stimulating, while the lady smoothly brushes it aside. 1960, then, is classier, but I prefer the wiry physicality of the 1938 Vienna Phil violins.
For the finale the conversation expands in effect into an operatic quintet, as it’s built on the interaction of five motifs. I’d suggest you think of motif 1, the opening phrase, quietly transformed from its appearance in the preceding Trio, as the lady, while motif 2, the livelier second phrase (tr. 12, 0:05) is the lady’s young daughter. The swaggering motif 3 (0:19) is the man and soon works into a triumphant tutti. The lady, however, finds equal satisfaction in the musing treatment of her motif begun by second violins (0:34), with the other string parts entering with it in turn as their colleagues weave counterpoint in echoing phrases around one another. Walter makes this a lovely interlude of sportive reflection amid the surrounding high jinks. In the next tutti motif 4, the shortest with a trill in its centre, appears (0:52), I suggest this is the man’s young son and at the end of it comes motif 5 (1:08), a quiet, nonchalantly flowing one in the first violins which cunningly combines aspects of motifs 1 and 3, so perhaps a peace-making elder daughter. So, I can leave you to enjoy and identify the mixing of these motifs. If you want to get into the complexity more, the participants sometimes wear masks: for example, at 2:03 motif 3 appearing in the lower strings as originally is dovetailed at 2:04 by its appearance upside down, rising rather than falling an octave. This is all done with great clarity and zip by Walter, balancing rhythmic flexibility with cohesion and bounce.
The development (2:23) spotlights our lady and gentleman in the minor, Walter’s lady pensive and probing while his gentleman is irritated and blustering, but also from Walter with an efficiency which suggests problems may readily be overcome. We’re quite soon back to recapitulation in the major (3:25), yet while not quite the false recaps of movements 1 and 2, this proves just the start of a growth of a steely element in motif 1, to which eventually motif 4 and then motifs 3, 5 and 2 come on board. In the coda (5:26) motif 1 enters upside down at its most reflective, against what proves to be a salve of chromatic woodwind contributions to give it the final steel to appear in the manner of a double fugue with a now assertive motif 5 (5:37). Motif 4 joins in a little later and finally motifs 3 and 2 tag along. So, the lady wins this contest, however much of a flourish the gentleman makes at the end, but the pep of his theme you remember longer.
This Allegro molto, as the other movements, receives more measured treatment by Walter in 1960, timing at 7:04 against 6:26 in 1938. The 1960 version is incisive and vivacious but lacks the bounce of 1938. The entries and interchange of the motifs are very clear but at the cost of an element of deliberation which for me makes the presentation over didactic at times. At the beginning of the development the contrast is still strong but a little different: the lady’s presentation gazes at the mysterious, the man’s is bullish.
CD3 begins with the best-known of Walter’s pre-war recordings, the Eine Kleine Nachtmusik made with the Vienna Philharmonic in 1936. It’s a de-luxe presentation. I’ll just focus on the first movement which has vivacious loud and velvety soft passages. The third theme (CD3, tr. 1, 1:03) offers in turn graceful finesse and vibrant might while Mozart’s felicitous modification of themes in the recapitulation is sunnily apparent, for example that of the second theme putting its second phrase up a fourth (3:05). Yet this Allegro is already a little slower, at 4:08, than the 3:50 of Walter’s 1931 recording with the British Symphony Orchestra (Pristine Audio PASC 492). The earlier account has more piquancy and sense of fun, the second theme glides more cheerily; but with a smaller orchestra, the dynamic contrasts are less vivid and the third theme is less sweet.
The remainder of CD3 features the recording of a live performance in 1937 by Walter of the Requiem which he rejected. The question that then arises is, if the artist himself didn’t wish the recording to be made available, should we be listening to it? Or do its merits outweigh this consideration? I’ll give you my verdict later. Fortunately, there’s a benchmark to assist, a ‘studio’ recording Walter made in 1956 (Urania WS121232) which he didn’t reject. It features the Westminster Choir and New York Phil with Irmgard Seefried (soprano). Jenny Tourel (mezzo), Léopold Simoneau (tenor) and William Warfield (bass).
Musically the Requiem is problematic because the dying Mozart never completed it. Walter uses Mozart’s pupil Franz Süssmayer’s completion of the orchestration and some of the sections, which many interpreters still do. The only section that’s all Mozart is the Introitus Requiem aeternam, though ironically here Mozart himself took much of it, including the orchestral introduction and first theme, from Handel’s Funeral anthem for Queen Caroline, The ways of Zion do mourn. Walter’s 1937 orchestral introduction is expressive in the woodwind opening, after which the agitated strings’ accompaniment and beefy trombones’ signal for the chorus to enter are clear. But the basses’ first entry isn’t really forte. Elisabeth Schumann’s solo, ‘Te decet hymnus’ from Gregorian chant is fervent but rather too operatic in this context. The choral take-up, ‘Exaudi’, is rather heavy. The fusion of the opening and second theme, the later first heard in the violins and bassoons (tr. 6, 3:48), is clear except the orchestra masks a rather weak ‘Requiem’ entry by the basses (3:57) before stronger sopranos’ ‘Dona’ (3:59). The sopranos’ top A (4:41) is reached more laboriously than exultantly and their intonation isn’t ideal.
I compared this with Walter 1956 with a leaner orchestra and choir. The opening choral entry has every part forte, crisp and dramatic. Unfortunately, the soprano soloist, Irmgard Seefried, is rather woolly. The best I’ve heard in Walter’s Requiems is Lisa Della Casa in his 1956 Salzburg Festival live account (Orfeo C430961B): she has a glowing spirituality without being over projected. The 1956 chorus ‘Requiem’ return entries are good and the whole movement strives forward in a satisfyingly purposeful manner.
The Kyrie, the Requiem’s second section, like most of the remaining movements, has the vocal parts and bass by Mozart but the instrumental parts by Süssmayer. It’s a double fugue, the basses beginning with the first fugue subject ‘Kyrie eleison’ with the contraltos almost immediately adding the second fugue subject, ‘Christe eleison’ (tr. 7, 0:04). Both themes run simultaneously throughout, spread across all vocal parts, an exciting mix of voices determined to present their different texts, but also the danger of a melee. The Kyrie fugue is sturdy in nature, the Christe one, with its semiquaver runs, more supple but also incisive. Walter in 1937 avoids the melee by compromising on tempo. The marking is Allegro, his is a slowish Andante. This works well for the Kyrie fugue, but makes the Christe one less supple than ideal, even that first ‘Christe’ entry could be better articulated. That said, the fugue subject entries are clear, there’s much spirited singing, even if the ladies hoot a bit, and the sopranos’ top B flat at 2:10 is good. The prominent presence in this account of the timpani added by Süssmayer keeps everyone together at the end. Walter’s tempo in 1956 is only 8 seconds faster, say Andante. The ‘Christe’ entries are better articulated and more incisive but I prefer the 1937 account’s fervour and weight to 1956’s grim, dutiful determination.
The Requiem’s third section is a Sequence of six movements beginning with Dies irae, for the first time in the work the stark message of a largely homophonic text spat out with venom. Walter in 1937 is equal to this and the Allegro assai tempo. Mozart’s variations of homophony to add to the excitement, such as the sopranos’ entry a half beat before the other parts at “teste” (tr. 8, 0:08) and the tenors (0:15) echoing the sopranos “Quantus” (0:13) are clear. My only quibble is that the antiphony from 1:00 between the basses’ “Quantus tremor est futurus” with its thrice repeated semitone illustrating trembling and the other parts “Dies irae, dies illa” with their own fearful semiquaver repeats is too stoic in its speedy strictness. It needs to be more petrified. In 1956 Walter achieves this at the same tempo because the chorus is throughout more tense and the clearer recording of the strings’ splenetic semiquavers gives the whole section more bite.
The second movement of the Sequence, Tuba mirum, presents all the soloists in turn from bass to soprano. It opens with a trombone obbligato solo with the bass, depicting the wondrous sound, “tuba mirum”, gathering all to judgement. It’s a really mellow, attractively lyrical invitation. Unfortunately, the trombone soloist in 1937 isn’t mellow, sounding uncertain and lacking the richness of the bass Alexander Kipnis. Walter was unhappy with the trombone. The tenor Anton Dermota now sings, in a suitably more emotive manner, of the book of judgement being brought forward; but Walter’s strings’ accompaniment in quavers could be clearer and more pressing here. Mezzo Kerstin Thorborg brings a bit of a hoot to the judge taking his seat and everything being revealed. Finally, soprano Elisabeth Schumann offers a wonderful mix of plaintive and sweet, pearly tone when showing the isolation felt by a just person and such a one’s quest for mercy, that final plea repeated by all four soloists together, an operatic companions’ ensemble. This latter I find emotively and movingly presented. The 1956 account has a better trombone, while William Warfield is a less rich but well-articulated bass, Léopold Simoneau arguably an over-dramatic tenor, not for me identifying with the situation as Dermota does. Jenny Tourel, I find a somewhat haughty mezzo but soprano Irmgard Seefried conveys the pathos of her solo well, though her voice is drier than Schumann’s. Better throughout is Walter’s fusion of singers and orchestra, especially in the closing ensemble.
The third movement of the Sequence, Rex tremendae, is the whole community responding with outcries of awe addressing the almighty and then elements within the community eager to show their commitment by the fusion of two double canons. The contraltos sing “Rex tremendae” with the sopranos echoing, while the tenors, one beat after the contraltos start, sing “Qui salvandos” with the basses echoing. The two groups then exchange each other’s texts and music. The end point, as in the previous movement, is the desire for salvation, here in the quiet prayer, “Salva me, fons pietatis” for all voices in homophony. Walter’s chorus in 1937 opens with heartfelt obeisance, though not quite the element of terror present in awe. There are strong entries within the two sets of double canons, the contraltos and sopranos in the first (from tr. 10, 0:40), the basses and tenors in the second (from 1:11), but all four parts don’t emerge clearly in the maze of sound. The closing quiet prayer is delivered in an imploring, creamier manner. Walter’s chorus in 1956 opens more crisply and with more sense of awe. Its entries in the double canons are clearer, but the distribution between all four parts still isn’t ideally apparent. The closing prayer is less successful, tailing off in a rather limp, dejected manner.
The fourth movement of the Sequence, Recordare, is the third and most extended and personal plea for salvation. You wonder how can a petition be as serene as this. Well, it’s in F major and imbued with a faith, directly conversing with Jesus, that it will be answered. Walter in 1937 also has an ideal, smoothly flowing tempo. The orchestral introduction sports mellow woodwind and clear strings with benignly falling sequences. A plea from contralto quickly echoed, another canon, by bass, followed by soprano and echoing tenor becomes more urgent as the four voices are layered together at “Ne me perdas” (tr. 11, 1:47). Throughout there’s an eloquent projection but also finesse in the finishing of phrases and elaboration of the orchestral part. These latter elements are, however, underplayed by Walter, which I like because it avoids fussiness, while his accompaniment remains ever sympathetic. This is also a good counterbalance to the very emotive, heart-on-sleeve singing, partly because there’s more vibrato than would be used in such music today. There’s just one stark, clouded passage, the simple, wan homophony of “In gemisco” (3:26), but the companions come out of it and the final repeated phrase, “statuens in parte dextra” (5:46), headed by Schumann’s diaphanous soprano is upliftingly positive. In 1956 Walter’s tempo is a shade faster, taking 6:00 to 1937’s 6:24, but this is enough to make the opening and returning melodic passages seem a little thrust forward and less serene. Walter’s orchestral opening is gentle, but later his accompaniment isn’t as delicate as in 1937. The 1956 quartet sound is cleaner and its dynamic and dramatic range wider, though soprano, Irmgard Seefried, doesn’t quite have the poise of Schumann. Where the 1937 voices sound unashamedly contrite on the edge of tearful, the 1956 ones offer a refined solicitousness.
The fifth movement of the Sequence, Confutatis, returns us, as in the third, to the juxtaposition of clamorous music and pleas for salvation. The basses begin in great alarm describing the horrors of the damned, the tenors dovetailing their lines in canon have the effect of continual, relentless pursuit. The sopranos and contraltos chant together sotto voce in comfortable harmony with the altos’ echoing of the sopranos a supportive one. However, after both sections are repeated, it’s the basses who lead a prayer for Jesus’ help, joined by the others. The tone is bleak but humbly submissive. Walter in 1937 doesn’t for me succeed. His opening men are too hearty, as if from the Triumphal March in Aida. Walter’s ladies are suitably beatific but over trance-like. The second time the men are serious and strenuous, the women more dejected in manner and straggly in voice. The basses’ prayer entry, “Ora”, beginning the final section (tr. 12, 1:08) is a solemn conversion with the overall effect fittingly desolate. Timing at 2:35 to 1937’s 2:57, in 1956 Walter adopts a faster Andante to advantage. The men throughout bring more sense of strain and awe, the ladies in contrast are ethereal at first, yet keep their composure in their sadder second response, after which the basses enter tamed in the final section in which both groups are more truly piano and accordingly moving than in 1937.
All that was composed by Mozart of the sixth and final movement of the Sequence, Lachrymosa, is the first two bars of strings and six of voices and bass. So, from tr. 13, 0:49 in Walter’s 1937 recording the rest is by Süssmayer, developed and extended from Mozart’s opening. Of course, it’s a good opening: a broken, sorrowing strings’ introduction, from all voices an anguished leap, then gentle, but chromatic fall followed by a slow, hesitant rise, as it were of fragments of broken limbs, to a searing crescendo and chromatic ascent depicting the guilty man coming to judgement. Walter in 1937 is best at the softer passages, performed quite affectingly, which is also the case with Süssmayer’s additions, in particular the plea “Huic ergo, parce Deus, pie Jesu Domine” (1:25) rather than the louder passages which I find overblown, so that the closing “Dona eis requiem” sounds more like a command than a heartfelt supplication. Walter in 1956 is better, with a more cleanly articulated start and softer passages, but the louder ones remain overcooked.
The Requiem’s fourth section is the Offertorium, comprising Domine Jesu and Hostias. For these Mozart composed the vocal and bass parts. Domine Jesu is a prayer to Jesus from the outset, to liberate the souls of the faithful departed from hell and its feel is now intrinsically calmer and the pulse more flowing, even though it’s shot through with alarm. So, it’s full of dynamic contrasts, from the soft beginning to loud affirmation of “Rex gloriae” (tr. 14, 0:08), heartily done by Walter in 1937, from the loud sopranos’ lead “de poenis”, straight to the soft “inferni” from all. This all requires crisper articulation and a smaller chorus than here to keep the piece flowing. The loud (f) passages are too loud (ff), as if they’ve strayed in from Verdi’s Requiem. Yet the more difficult chorus passages are better done, such as the combative plea of the fugato “ne absorbeat eas” (1:21) and the determination of the closing fugue “quam olim Abrahae” (2:42), with the entries of the key word “promisisti” belted out. The soloists’ contribution, “sed signifer Michael” (2:02), requesting the appearance of that saint to lead them out, sounds rather strained. Walter’s 1956 account, timing at 4:31 against the 4:51 of 1937, has a better flow and sense of purpose, notably in the soloists’ contribution, but suffers from the same extreme treatment of loud passages. Nevertheless, the “ne absorbeat eas” fugato is given a remarkable, frenzied energy.
Hostias is mainly a calm, smooth prayer affectionately commemorating the departed souls so that they pass from death to life, in which occasional loud fervour of appeal, as at “Hostias” (tr. 15, 1:26) is immediately contrasted with humble supplication in the soft “et preces” (1:32). Unfortunately, in Walter’s 1937 account the same problem as in Domine Jesu of excessive loudness recurs and even the opening, though affectionate, is rather mushy in tone. At the end Mozart repeats the “quam olim Abrahae” fugue of Domine Jesu, sung here with welcome animation. The timing contrast here is even more marked with Walter’s 1956 account than in Domine Jesu, in this case 1956 timing at 4:51 against 1937’s 5:40. In 1956 the tone at the start is lighter, yet still gentle. The dynamic contrasts are here better tempered: the loud passages are firmer without being over dramatic. The repeat of the ‘quam olim Abrahae” fugue is a resolute claim of just rights.
The remaining sections, the Sanctus, Benedictus, Agnus Dei and Communio: Lux aeterna, were all composed by Süssmayer, except that he did in the Lux aeterna adapt music from Mozart’s Requiem aeternam and at its “Cum sanctus” repeated Mozart’s music of the Kyrie fugue. Whether, had he lived, Mozart himself might have done this is a moot point, but it does end the work with more gripping music than that immediately before it. That’s not altogether fair; the solo quartet Benedictus begins gracefully, nicely topped by Elisabeth Schumann in 1937, but it outstays its welcome and the other soloists sound a bit frayed. The Vienna State Opera Chorus does offer a lighter, more festive Osanna section than its turgid first attempt when ending the Sanctus. In Walter’s 1956 account the choral singing is better focussed and the dynamic contrasts are better. In the Benedictus the New York Phil playing is more radiant and the soloists as a group sing with more feeling.
To sum up, I find this set a mixed bag. The 1937 Requiem is its lowest point. It does have some merits which I hope I’ve shown but, on the whole, I think Walter was right to reject it and courageous, given that at the age of 60 he mightn’t have thought he would ever get the opportunity to record the work, let alone 19 years later. But the only other slightly weak account in this set is the 1928 Die Zauberflöte overture. And when it comes to the symphonies, there are fascinating revelations even from my selective survey. If you want Mozart playing that focusses attention on melody and drama, Beecham is preferable. But if you like more illumination of melody, Walter is your man. Interestingly, in his New York Phil period he focussed more on drama, but for me this wasn’t his forte. In his Columbia Symphony period, he achieved a good balance of melody and drama, but for me too much at the cost of the melody. This set convinced me that Walter’s earliest recordings of the late Mozart symphonies are his best.
Die Zauberflöte overture, K588 (1790) [6:45] rec. Paris.
Le nozze di Figaro overture, K492 (1786) [4:08] rec. Central Hall, Westminster.
La clemenza di Tito overture, K621 (1791) [4:19] rec. Musikvereinsaal, Vienna.
La finta giardiniera overture, K196 (1775) [2:14]. rec. as tr. 3.
Three German dances, K605 (1791) [5:35]. rec. as tr. 3.
Piano Concerto No. 20 in D minor, K466 (1785) [27:37] rec. as tr. 3.
Symphony No. 38 in D major, K504, Prague (1786) [22:32] rec. as tr. 3.
Symphony No. 39 in E flat major, K543 (1788) [25:37], rec. Abbey Rd Studio 1, London.
Symphony No. 40 in G minor, K550 (1788) [23:10], rec. Beethovensaal, Berlin.
Symphony No. 41 in C major, K551, Jupiter (1788) [27:21] rec. as CD1, tr. 3.
Eine Kleine Nachtmusik, K525 (1787) [15:03], rec. as CD1, tr. 3.
Requiem, K626 (1791) [59:23], rec. Théȃtre des Champs-Élysées, Paris.