Der Rosenkavalier: The Classic Recording

by Horst Wahl

If ever in the history of opera there were an identification between a role and an artist it would be that of Lotte Lehmann as the Marschallin in Richard Strauss' Der Rosenkavalier. It was a lucky strike that both this role and that of Sieglinde in Die Walküre were recorded in Lehmann's prime. In 1932 it had been decided to produce the complete Der Rosenkavalier with the composer as conductor. But his price as well as that of Bruno Walter led to the selection of Walter's protogée Robert Heger.

The effects of the Depression were not yet over, so cost cutting measures had to be considered. This included the "cuts" made in this long opera. 12 discs were originally projected, but finally 13 were released (26 sides) in which the choices allowed for connecting scenes of the high points of the opera.

Lehmann was the obvious choice for the Marschallin; Elisabeth Schumann as Sophie and Richard Mayr as Baron Ochs were almost as inevitable. In the happpy choice of Maria Olszewska as Octavian, HMV found an aristocratic sounding mezzo whose sound was sufficiently different from that of Lehmann and Schumann, but who provided a complete blending in the duets and trios.

The recording was to take place in Vienna, thus the inclusion of the Vienna Philharmonic and Vienna State Opera Chorus. It was financially feasible to record in the "Mittlere Saal" [Middle Hall], although the 1935 recording of Die Walküre demonstrates the acoustic superiority of the large "Musikvereins-Saal."

The orchestra consisted of 10 first and 8 second violins; 6 violas, 5 'cellos, 5 basses, 1 piccolo, 2 flutes, 3 clarinets, English horn, 3 bassoons, 4 horns, 3 trumpets, 2 trombones, 2 bass trombones, 1 tuba, 2 harps, tympani, drums, celeste and piano.

Members of the chorus of the Vienna State Opera and the Vienna Philharmonic Orchestra were conducted by Robert Heger.

The dates, matrix and take numbers are valid only for published takes. Because there were two cutting machines working at the same time during the recording sessions, each take produced two wax masters. The recordings of the first machine were unidentified in the matrix numbers; those of the second machine are shown by an "A" after the take.

Catalog numbers DB 2060-72 (matrix 2WX 581-606) were issued in England in December 1933 and in Germany in June 1934. They were issued in America on Victor 7917-29 (set M-196) which included matrix numbers CVS 81418-81443. The auto coupling set was AM-196 (7930-42), auto drop sequence coupling DM-196 (17119-31).

The following list [not included: see complete discography] demonstrates what an incredible confusion existed in the order of the recordings. Our wonder and appreciation for the highest artistic standards are awakened when one hears the development and connection that is achieved. If one imagines listening to the four minute "takes" that we hear unbroken on LP or CD, then one has a better appreciation for the level of identification the arists had with their roles and the situations.

The recordings began on the afternoon of 20 September 1933 and ended 4 1/2 days later.

The great number of "takes" does't necessarily indicate artistic errors; the necessary combinations of solo and ensemble voices with orchestra demanded by the score forced the technicians to try various microphone placements [Remember only one microphone was used]. Both cutting machines produced the same number of possibly usable waxes, but far from all were saved and registered. Technical and musical reasons resulted in the destruction of many.

A total of 58 "takes" or 95 wax recordings, of which 2 were damaged during the trip to Engand, were quickly made into the album in time for Christmas sales.

Though it was disappointing that the "singer's aria" was missing, nevertheless there was world-wide acclaim for the recording. Many enjoyed Beniamino Gigli in the role of the Italian singer, but his fee would have been too high. Later a German recording with Helge Rosvaenge was welcomed. [At this point Horst Wahl recounts the story, told to him by Mme. Lehmann and later recounted by her in various interviews, of how she missed singing the famous "Ja, Ja" at the end of the opera. Herr Wahl provides the precise historic data to document the evident confusion that recording such a huge work in four minute scraps must have caused. In the end Schumann sang the words produced on the recording, although Lehmann also recorded them. It was because of Heger's decision of what length music should appear on the last disc, that Lehmann's "take" wasn't used.]