I love the music of the criminally underrated Polish composer Karol Szymanowski. He represents lush late romanticism at its finest. Of all of his extraordinary works, I can’t pick out a single favorite. But I can single out his Third Symphony, Song of the Night, as a work especially deserving of a wider audience, and so I decided to survey all of my recordings of the work.

This work, written between 1914 and 1916 after Szymanowski had returned from a trip to Sicily and North Africa, is exotic and contains music purportedly inspired by Near Eastern culture. The exoticism is accentuated by the poem he sets: the second Divan of 13th Century Persian poet Jalal ad-Din ar-Rumi provides the text. (Truth to tell: the poem is really not that great.) One can assume that Szymanowski was in search of his own voice, his earlier works being decidedly Straussian in nature, and he succeeded fabulously.

Enough background, what of the recordings? Well, there ain’t exactly a whole lot to choose from. A quick glance at Arkiv found but four recordings. Fortunately, I have three of them.

I’ll start with Karol Stryja’s 1988 version, formerly on Marco Polo, now on Naxos. He is lucky to have on hand Wieslaw Ochman as the tenor, as Mr Ochman had recorded this work before (which will be covered later). Mr Ochman definitely knows the piece. His first entry in the first movement is assured, powerful, and remarkably incisive. His command of his native language is obviously quite good, and he handles the text wonderfully. Nearly as good is the chorus. They are well honed, to be sure, and they act as a truly cohesive group. Whether singing forte or very carefully delivering the dreamy sound needed in parts, they do a splendid job. As different portions of the chorus sing different parts, the lush sound world is filled with the wondrous sound that only massed human voices can produce.

One of the things I enjoy about the choral writing here is its lack of any hint of religiosity and its outright sensuousness. Perhaps that’s to be expected given the poem and work, but it’s a common trait in all of Szymanowski’s choral writing. He is among the very greatest of choral composers, ranking right up there with Bach and Haydn, but his approach is decidedly different. But I digress: back to the work.

The second movement is more energetic than the first, and in this recording one hears what sounds a bit like Rimsky-Korsakov or Ippolitov-Ivanov inspired invention. But far more inviting than this safely exotic device is the use of wordless mixed chorus. Again, Szymanowski displays his gift for choral writing: the voices wax and wane with the music, in a sort of hazy, perhaps opium induced ecstasy. I suppose one could say it is reminiscent of Debussy’s Sirenes, but the sound world is even denser, and it tends more to the exotic than erotic. The third movement opens mysteriously, with low volume string playing that recalls the opening to Mahler’s First. Ochman returns in this movement, and again it is with authority. All of the elements combine quite well, but Stryja’s conception lacks that last bit of urgency. The big tutti near the end is not especially powerful or stirring, so the quieter ending loses some of its impact. The almost complete lack of organ output certainly does not help.

The orchestral playing is a tad subdued and lackadaisical, but I mean that in the best way. At no time does one get the sense that anyone has a desire to rush through this work; everyone just wants to revel in the glorious music. Stryja’s Polish State Philharmonic Orchestra of Katowice is hardly a top-flight band, but they prove up to the challenge. Mostly. The strings are rounded and warm, and the winds prove capable of floating a little line here and there as necessary, and the brass assert themselves, though they never sound blatty. (Brass aficionados may find this less than compelling, of course.) The hazy, less than perfectly clear sound actually doesn’t hamper things too much here; it rather adds to the allure. So, this is not a great performance, but it is good, and were it the only one I could live with it.

The next version is Simon Rattle’s 1993 recording. Szymanowski fans should rejoice that a major label artist has recorded so many of the Pole’s works, and certainly I am. Alas, they are not all they could be. Rattle uses his old band, the City of Birmingham Symphony Orchestra and its chorus, and the overall effect is technically proficient but somewhat lacking in the most important elements of the music. From the start, this is a more astringent, more forceful, and fleeter take. The strings aren’t cutting, but they are not lush, either. The winds sound decidedly clean and completely British. That’s not a putdown, just recognition that they don’t evoke the sound world they should. The brass fare better than in Stryja’s account, but that’s it. Well, the chorus is more precise and better controlled, too, but they, too miss the mark. Where is the exotic, mysterious sound so important in this work? The second movement is more vigorous than Stryja’s and sounds not a little playful. It is also stripped of some color, sounding more like lean Strauss than opulent Rimsky-Korsakov. The chorus again do a technically superb job, but it doesn’t really jump up and grab one’s attention. The final movement is far more driven than the Naxos offering, and Rattle really goes after the big tutti, producing a towering wall of sound, albeit in a non-Fluffy sort of way, and it really is quite exciting. If the color and exoticism is missing, one gets to enjoy something more exciting.

What Rattle’s recording does boast is the best sound of the three recordings. Indeed, it is something of a sonic showpiece. As with all CBSO recordings I’ve heard, the sound is remarkably clean, a bit devoid of color, but possessed of huge dynamic range and staggering power. The organ is readily apparent every time it is used, but it never overpowers the proceedings. For those lucky enough to have true full range speakers, this recording will pressurize a room, no doubt. Even with my two-way towers, the walls, well, rattle. For sheer sonic splendor, this is a fine recording. As a performance, it is actually quite good, certainly among Rattle’s better efforts, and were it the only one I could live with it.

But neither Stryja nor Rattle offer the only choices. Thank goodness. Jerzy Semkow’s 1979 recording is the one to own. It’s that simple. I still remember when I first heard his rendition of the Fourth Symphony. I was almost stunned at how much better it was than the other versions I’d heard. The result is nearly the same here. He manages to combine the strengths of the other two conductors - technically excellent execution and opulent playing - while also infusing it with an energy not to be heard in the other versions. His orchestra and chorus are both tighter and cleaner than Stryja’s and more flowing than Rattle’s. And he also benefits by having a younger Wieslaw Ochman as the tenor. Where to start?

The singing. Ochman is in better form here than for Stryja. His command is remarkable, his diction impeccable, his tone attractive and well suited to the music. The chorus is every bit his match. As technically adept as Rattle’s singers, they benefit from singing in their native language, and it helps. Whether they are nearly whispering, erupting from silence, or sumptuously undulating, they are decidedly easy on the ear. What is it about the Polish language? All of the works I’ve heard in this language, admittedly mostly by Szymanowski, have been alluring, the singing revealing this Slavic language to be especially singable. Anyhoo, the vocal contributions alone make this recording the one to have.

Fortunately, one gets the orchestral playing, too. As in the Fourth, Semkow proves to be in his element. He leads his Polish Radio National Symphony Orchestra to produce oodles of beautiful sounds, all while keeping the work under firm rhythmic control. He never loses the line, carefully and sometimes gently leading from one section to another, with everything blending together so well that one can only conclude that this is how it must sound. Detail abounds, and while he nearly matches Rattle here, he does one better by allowing for a more sumptuous and relaxed (but never sluggish or dull) background from which those details emerge. Listen to a little flute figuration waft thorugh the air, or perhaps to an especially lovely little passage in the violas, or maybe to the Scriabin-esque brass. (Oh, my, that sultry little horn figure at around 7’ into the first movement, how wonderful it is!) Listen to Semkow: listen to how he cajoles the strings, whipping them into a near ecstatic frenzy for brief periods; listen in the third movement as he so flawlessly moves between contrasting sections concluding in a massive, charged final tutti that fades away so delicately as the last notes linger until disappearing into silence; listen to how he leads the second movement, playing with dynamics and tempi so subtly as to create an entirely new sound world. Gone are Strauss and Rimsky-Korsakov. Here is Szymanowski.

To top it all off is superb sound. It certainly lacks the ultimate weight of Rattle’s recording, and may be a tad bright at times, but it will more than do. Of course, my reissue is based on the 1994 transfer, so perhaps the new Gemini reissue sounds better. In any event, this is the recording to own if you want just one. But who would want just one? Now, if only the Berlin Philharmonic or Philadelphia Orchestra would record it. I’m left thinking of but one word: More!