Richard Caniell, Immortal Performance’s doughty producer, does what many collectors tried to do in the 78-rpm era. Though there had been complete opera recordings since the second decade of the 20th Century, and in growing number from the advent of the electrical process, many of the greatest individual performances were represented only piecemeal; the devotee’s ambition was to assemble the pieces. By this means one could painstakingly set in proper array, for instance, most of Chaliapin’s Boris (as well as his Pimen and Varlaam) or, closer to the present case, of Melchior’s young Siegfried^(I)—all from their commercial releases, of course, which came and went in the catalogues. Private off-the air recordings remained private and under copyright proscription, and the universal availability of broadcast material, in however uncurated condition, via the internet and free of charge—and, n. b., of any compensation—was no more than fantasy.) With the gradual emergence of the aforementioned Pirate Underground, it became possible to hear such performances as close to complete as their stage originals allowed, or in some instances to at least extend significant stretches of them by intermixing studio and broadcast sources. By this time, I think, everyone has had a chance to decide for him- or herself whether or not the mixing of a variety of sources to constitute a continuous performance that didn’t ever happen quite that way is an acceptable procedure. It seems to me that so long as the sources are faithfully disclosed, there’s no truth-in-advertising problem; as to its effectiveness, I’m taking it on a case-by-case basis. (Readers will recall that, great admirer of Ezio Pinza as I am, I was not happy with IM’s insertion of his Father into the otherwise integral (with cuts) Columbia Louise, in place of the perfectly solid and stylistically manor-born André Pernet—see “Louise,” 4/16/21).

The situation with regard to this Tristan is different. IM’s release presents more of the Leider/Melchior material than has ever been gathered into one sequence, restores and equalizes the sound to the extent possible, and sews the passages together wherever feasible. We have two Brangänes and two Kurwenals, but the eponymous protagonist couple is kept on track throughout. There are big chunks of the opera here, but no question of a complete version, so I will describe only the more extensive cuts that are taken within the excerpts. We start with the Prelude, in surprisingly present sound, even considering that it is taken from the 1936 broadcast.^(II) I have found with these Artur Bodanzky-led Wagner performances that all the purely orchestral interludes are impressive, with the exceptions of a few of the more excitable ones like the hysterically rushed Lohengrin Act 3 Prelude. That’s certainly true of this—allowing for the inferiority of the sound when set alongside any of the late versions, it’s as gripping a reading as you’ll hear, very slow—the fermatas over the rests extended almost to the snapping point—weighty and sinewy, imbued with the tension of a pressing forward against a retaining force, each of the harmonic and orchestrational destinations along the way arrived at with seeming inevitability, and the mighty climax released with all its tragic force. Conductor and players had been working together for nearly two decades through an intense period of Wagner performance, and knew exactly where they were going.