Then we skip past the opening Brangäne/Isolde scene and the reprise of the sailor at the mast-head’s ditty to Isolde’s “Mir erkoren.” From there continuity holds through to a cutoff at Brangäne’s “Welcher Wahn!”, a stretch of about 21 minutes that takes in Brangäne’s mission to the afterdeck with Kurwenal’s song, and most of what we usually call Isolde’s Narrative and Curse. This is accomplished by inserting into the broadcast material the  passage beginning with “Doch nun von Tristan” from a 1928 studio recording (Berlin, Blech, with Elfriede Marherr-Wagner the Brangäne). There are several things to comment on here, a couple having to do with the 1933 Brangäne, Maria Olszewska. That she was an important artist, a pillar of the great traveling Wagner medicine show of those years, is beyond doubt. Also beyond doubt is that the real impact of her voice seems to have often eluded the microphone. She was a contralto, and when she sang Lieder (e. g., Schubert’s Aufenthalt or Der Tod und das Mädchen) in contralto keys, or essayed a true contralto role in opera, like Erda, she sounds quite marvelous.^(I) But like many other non-Mediterranean singers of that category in those years—even after the introduction of the microphone mitigated the distancing requirement for big voices on high notes—when music extended above the upper F or so and a sustained full-voice note was called for, she tended to either back off (we hear this on her version of the Ortrud/Telramund scene with Emil Schipper), or to release into a vague “pure head” adjustment that is not very satisfying, particularly in climactic moments. Now, Brangäne is designated by the composer as a soprano role. Its writing is predominantly lyrical, and its tessitura hangs fairly high; yet its compass does not reach beyond the top A (whereas Isolde’s goes to the B natural and C) , and extends no lower than Isolde’s at the other end. So although the role of Brangäne has mutated over to the mezzzo-soprano or even contralto vocality, to judge from the writing it would seem that two sopranos are intended, the Brangäne’s actually the lighter and more pliant of the two, rather as with another similar case, that of Norma and Adalgisa. I have seen this attempted only once, in Stockholm, when the Isolde of Berit Lindholm was tended by the Brangäne of Sylvia Lindenstrand, a nice singer but a quite light one. That did not work very well, because the voice seemed lost in the overall sonic framework of the opera. It serves better here, since Marherr-Wagner’s bright voice has some strength and thrust, and she takes an active part in the dialogue.

Jumping over the passage that would have given us the best read on Olszewska’s Brangäne (not a criticism, only an observation) we pick up again at her “Wähnst du, die Alles klug erwägt” and carry on through the bringing-forth of the fateful draught and Kurwenal’s intrusion to Isolde’s “Untreue Magd,” just short of Tristan’s entrance. We must work at it, but here we can at spots get an idea of both Olszewska’s deep-set timbre and a few forays to strong (though brief) top A-flats and As. And that’s where our Act 1 ends. Saving comments on the singing until the end, I’ll observe here that the sound, while hardly of the “sit back and enjoy” variety, is better than I had anticipated, and a great deal better than David Hamilton described from the sources he had at hand. The restorations have clearly improved the listening situation. As always with such material, the more familiar the listener is with the music and the singers, the more coherent the performance will seem, and following with a score will help. There are three noticeable changes of ambience in the course of the act. One, I assume, is from one broadcast source to another, the other two are in and out of the studio “Doch nun von Tristan.” With respect to this last, I must say that while I can see, mistily, the point of reconciling the 1936 excerpts to the limits of the ’33 transmission, I do not see the advantage of degrading a tolerable studio recording (i. e., making it sound worse than it needs to) for the sake of pretending we’re hearing a continuous performance.