Tuesday, November 30, 2004

Ella's Kern Song Book: lost masterwork

I've come to this one late in life. Today was the first time I've ever heard this recording.

It was the next-to-last installment in the song book series, from 1963. It and the last (the Mercer) were one LP each, the briefest of the series. The concept was clearly running out of steam.

Oh yeah?

This is the great lost masterpiece of the series--I'd go so far as to call it one of the great lost masterpieces of American popular singing. The arrangements are by Nelson Riddle--some of his best, most unusual, gentle, understated, buoyant work, perfectly attuned to Ella's voice and style. All that work with her on the five-disc Gershwin set didn't go for naught.

One supreme example: A Fine Romance. Riddle scores this not unlike his famous chart of I've Got You Under My Skin for Sinatra, which is to say as a slow crescendo. It's a great comic song from Swing Time, as I'm sure you know, and Astaire and Rogers did it to a 'T' in the film. Ella's is totally off the map--it bears very little resemblance to the movie version, other than the lyrics and melody. Interpretively, it's a thing of such artistry (both subtle and grand) and majesty that you've got to hear it to truly appreciate it. Words, especially my feeble ones, can't begin to describe it.


Monday, November 29, 2004

Gone With the Wind: DVD comes of age

The new Warner's four-disc special edition DVD of Gone With the Wind is a triumph. I'll omit discussion here of the tons of extra material included, other than to point out that the feature-length documentary is one of the best ever made about a film--the screen tests alone are worth the price of admission. There's so much other stuff that it would really require a review of its own.

All that would be for naught if the visual reproduction of the film was lacking. Fear not. This is by a wide margin the best-looking Gone With the Wind ever seen on home video. Prior to this one, the CAV laserdisc edition took the honors; the previous DVD came early in the format's history, and was ruined by over-enthusiastic application of edge enhancement. What's edge enhancement? If you don't already know, I'll spare you, because it's one of those things that will bother you for the rest of your life every time you see it. Ignorance in this case can be bliss.

The good news is that I can't detect any of that here. Nor can I see any evidence of DVNR (digital video noise reduction). The techniques for "cleaning up" old films have obviously progressed beyond those, and what we have here is a video reproduction of a movie that even projected to over 100 inches (on the Plus Piano DLP) looks gloriously film-like. No phony "sharpening," no goosing of contrast or color, no elimination of film grain in the interest of "clean" reproduction. Cleanliness may be next to godliness when it comes to transfer of video material, but where film is involved, eliiminating grain is tantamount to throwing out the baby with the bath water.

The color scheme of the film is on the warm, creamy, brownish, pastel-ish side, which I believe to be true to the film-makers' intentions. GWTW is not Singin' In the Rain--eye-poppingly garish color wasn't the aim here. That said, the eye is constantly seduced by small, vivid details--the emerald green of Scarlett's Parisian bonnet, the brownish red of her hair (I find hair color a reliable gauge as to color accuracy), the baby blue of Bonnie Blue Butler's eyes. The photography is stunning, and on a big screen--which is really the only fair way to watch this film--the whole thing conveys a sense of hugeness that has virtually no equal in film.

The package is worth the price for the extra material alone. (On laser, the documentary was actually sold as a separate set.) For the movie, it's worth twice the price--that's how good this transfer is.

A truly sensational piece of work. I can't imagine this film looking any better in the here and now.


Sunday, November 28, 2004

Solti and Tristan: Match made in hell

Wagner: Tristan und Isolde. Nilsson, Uhl, Resnik, Krause; Solti, cond. Vienna Philharmonic Orch. & Chorus (Decca 470814: 4 CD's, originally released in 1960)

The conducting on Solti's 1960 Decca recording of Tristan is of course silly--Boehm, to name just one, understands this music whereas Solti kind of slugs along from moment to moment--the opening to Act II is almost humorously jumpy. Too bad, because Nilsson is in spectacular early voice. That said, her later singing for (there's that name again) Boehm is far more inflected and passionate, though the tone isn't quite as sweet--it's a tradeoff I'm willing to make. Uhl isn't quite up to the task of Tristan, though he's not much worse than, say, a Rene Kollo, except that the tone isn't as pretty. Resnik's Brangaene? More secure than in her slightly later and distinctly wobblier and more sour-toned Decca Carmen, but none too pretty either.

Sonics: this was one of the great Deccas, but here they've de-noised it, which is a crime against humanity. They call it CEDAR de-hissing, but it's EQing by any other name, and no matter how you slice it, it takes some music--in this case fine detail and upper string sound--with it. The lack of hiss gives the whole thing an oddly dead noise floor--it's simply not meant to sound this way; something's seriously out of whack. Better they should've left it alone and done a straight transfer, which at least would've suffered only from being digitized and having all its frequencies above 20kHz being thrown out. Sound good? It doesn't. If I had it to do over again I'd go straight to the older CD transfer, or far better yet, to the LPs.

That close to Act I is still a Culshaw spectacular, though you have to take a great deal of it on faith here. Not only is it the production and engineering, it's--at the start of Act II as well--the stunning brass of the VPO, which sound like no other. Too bad Decca didn't enlist Kna for this set; Solti was their big marquee name at the time, especially right after Rheingold, and thus the obvious choice.

Played at a healthy volume, incidentally, the voices are not really recorded too distantly at all. Culshaw was no idiot, and he knew this would force listeners to play the recording loudly, thus enjoying the full impact of the climaxes. If you can't accommodate that, well, don't blame him--blame your pathetic little squawk-box.


Wednesday, November 24, 2004

RCA/Decca's "lost" Gioconda unearthed

Ponchielli: La Gioconda. Milanov, di Stefano, Warren, Elias, Amparan, Clabassi; Previtali cond. Orch. & Chorus of L'Accademia di Santa Cecilia, Rome (RCA LSC-6139; 4 LPs. Issued 1958)

After a long separation, this recording and I were finally reunited yesterday when I gave a listen straight through to the original RCA Living Stereo pressings. These you generally don't see around very often: the recording was kind of a dud commercially. It was issued on four rather than three LPs, there were numerous typos in the booklet (including the misspelling of the name of the original play and the omission of the side change markers), it was negatively reviewed, and a better-received competitor came out from Decca almost simultaneously. It soon was transferred to RCA's budget Victrola label, and eventually the rights reverted to Decca, who've had it in and out (mostly out) of print ever since.

Was it deserving of this bummer of a fate? Well, yes and no. On the debit side: This is just the sort of opera that cries out for the kind of old-style Italian repetiteur represented by the likes of Previtali, but here he seems on autopilot. The thing obviously *needed* eight LP sides, thanks to his deliberate, occasionally leaden tempi. (As a guidepost as to how it can and should be done, there's that roughly contemporaneous Gavazzeni performance for Decca; and for the Toscanini-ites among us--not I, though in this work I might be persuaded to budge a bit--there's Panizza in the '39 Met broadcast, who conducts the Dance of the Hours like it's the opening scene of Otello.) Strike one.

All three--repeat, *all three* of the principals are past their best. This (and Leonora in Forza, I suppose) was Milanov's signature role, and what a pity that she (or RCA) got around to it a tad too late. Thankfully, there are several Met broadcasts to give us a better idea of what she was capable of in the part--first and foremost that Panizza of '39, where she's partnered by the great Martinelli--but of course none in this kind of sound, about which more anon. While there are moments, none more memorable than The Note, which she sings beautifully here (but releases a bit awkwardly--though only in comparison to her best, er, "Notes" in the broadcasts), there are overall a kind of blowsiness and loosey-goosey approach to the high register typical of Milanov's late-career performances. (And to think she sang 8 years after *this!*)

Warren's best years were the '40s, so all his LP recordings of the complete operas find him varying degrees of "past his prime-ness." But that doesn't mean there's not a lot of power and brilliance to be heard; I find it heard here to greater advantage than in any of his studio recordings of the '50s, save for the Macbeth. More below.

Di Stefano is the real problem child of this recording, which is frequently cited as one of the prime examples of the self-destruction of his voice. I can't really argue with that. When he slams into and holds any of those loud, open high notes (go ahead--you pick 'em), you get this mental image of him standing on his tippy-toes, neck strained, eyes bulging, face turning a deep shade of purple. Really, it apparently wasn't all that bad (the neck thing I mostly made up), but this was still the kind of singing that did his voice absolutely no good. It was, as we now know, the beginning of the end, this in his mid-'30s, an age when most great singers are just getting their feet wet. Very sad.

So with bad marks for all three of the principal singers and the conductor, that makes four strikes, and in baseball, as we know, that's one too many (actually two too many, but you know what I mean). But because this is my review and I'll do as I please, I'll give the singers one strike combined, which gives us only two strikes total and allows this review to continue (as the crowd jeers in unison).

So what's so good about it? Let's take the points one-by-one. There are some names from the first part of this essay that will reappear in this section. Milanov: this was indisputably *her* role, and her performance, even at this late date, cannot be casually dismissed with a wave of the hand. The voice is still clearly of the right weight and timbre, the sound is still beautiful much of the time, she's got a functional chest register (though occasionally you wish she'd lean into it with a bit more gusto), and she knows how to sing--and above all, how to sing this role.

Warren practically walks away with the performance in his back pocket. First and foremost, there's the outsized, melodramatic characterization, including that anguished, hair-raising scream at the opera's close. (Can you imagine a Hampson or Terfel--to name just two who probably have never even come close to the role [or if they did, shouldn't have]-- "lowering" himself to do this kind of thing nowadays? But hey, friends: that's opera!) Although there is the usual lack of steadiness, particularly in faster passages, that tended to increasingly plague his singing at this point in his career, there is also the high register in its usual mind- boggling form, and also some of the most exquisite sotto voce, introverted soft singing you'll ever want to hear, particularly in Act I.

Funny how we remember Warren almost exclusively as a *loud* singer (and loud he could be, especially in the seemingly endlessly forte series of aria recordings in the '40s--the Romophone set, played in sequence, gives me a headache); but in reality, some of his best work was accomplished at the opposite end of the dynamic scale. Meaning "he sang soft real good."

The "Bit Players:" Alvise is hardly a comprimario role, but it gets something less than star billing on this set. Plinio Clabassi was a serviceable basso who came in handy, mostly for RCA's Rome projects, in roles like the King in Aida (both the Perlea and Solti recordings) and other assorted second tier parts. Here you get the feeling he was deputizing for someone--a Tozzi perhaps (who was the Guardiano in the Forza RCA recorded at around the same time with the same principals). He's not nearly in the same class as the big bad bassos of the era--a somewhat nondescript timbre and lack of solid tonal core (tranaslation: he's none too steady while never quite deteriorating into an outright tremolo)--but he earns his paycheck honestly enough.

The Spanish contralto Belan Amparan is a tad hooty as La Cieca, but aren't *all* the singers who take on this role? (Or so they sound to me; still, I've certainly heard worse.) Roz Elias' Laura, OTOH, is exceptionally good, and in the duet with di Stefano, she's something more than that: for once in this recording, someone's voice is in the first flush of youth. Very possibly her best recording, in a fairly extensive discography.

Di Stefano is at the end of his vocal tether, it is true, but what a tether it was! In the lyrical sections--as in most of the meat of "Cielo e mar," for example--his singing is extraordinarily beautiful, and he can still sing out in the soft passages: a genuine mezza voce, not the unsupported croon of his later years. In the "damning with faint praise" category, we can observe that his singing is steady at all dynamic levels and invariably in tune. His phrasing is as gorgeous as ever. Even compared to so great an artist as Gigli, in the "Enzo Grimaldo" duet the way he colors the words and phrases is fresh, imaginative, and totally natural.

The role, unfortunately, is about three or four sizes too big for him (I'd say it's a 9-7/8, and he's around a 6-1/2). In the aforementioned duet, the engineers have seemingly taken no measures to boost di Stefano's level to match that of Warren, and you can easily hear the way the poor little guy is swamped. And the ways in which he tries to boost his *own* level are unacceptable, audible, and ultimately destructive.

The sonics, in the original vinyl pressings, are miraculous. The voices are recorded with plenty of air around them, but also with striking immediacy. Same goes for orchestra and chorus; the ensemble sound has tremendous weight and impact, but also remarkably fine definition and detail.

The big shocker here? The names of the producer and engineer, respectively: John Culshaw, he of Solti Ring fame, and Kenneth Wilkinson, he of some of the greatest-sounding recordings of all time. The overriding suspicion is that they may have felt as if they were "slumming" here--just picking up the paycheck. But nothing of that nature is reflected in the first-rate production values and gorgeous sound, which shout "Culshaw/Wilkie" all the way.

Legally-speaking, the set was, like many of the Wilkie classics, a joint project of RCA and Decca. Decca now owns the rights to the recording. At any rate, this Gioconda makes my top ten list of best-sounding opera recordings of all time

In summation, this performance hardly makes the best possible case for the opera. Previtali is a serious debit--this in a work that practically conducts itself. But the orchestral and choral work is very good, the recording is sensational, the smaller parts are generally well-sung and -acted, and the principals, though all past their best, are still singers who demand to be heard (and in the case of Warren and Milanov, no doubt *were*--above orchestra, chorus, and even prompter).

For the average human being who needs a Gioconda (and don't *all* average human beings?), this might not be my first recommendation: that would probably be--notwithstanding some not insubstantial flaws of its own--the Decca with Cerquetti, Del Monaco, Bastianini, and Siepi, conducted by Gavazzeni on Decca. But for the rest of us who need at least 20 or 21, this one should definitely be among the top dozen or so.


Tuesday, November 23, 2004

A "new" Flagstad-Melchior Tristan for the ages

Among Melchior-Flagstad Tristan recordings, between the Reiner and the Beecham, I've always preferred the Reiner for reasons I can't quite explain, other than that the conducting of Reiner seems more in touch, more of a mind with the interpretations of Flagstad and Melchior. Even on the Marston transfer for Naxos, however -- by all accounts the best sonic choice for the Reiner -- one will often have to listen very carefully to hear Melchior in Act I: he's almost constantly off-mike. The rest of the show is engrossing enough that I can deal with the sound; I've certainly heard plenty worse from La Scala in the '50s.

As for the *best* Flagstad-Melchior Tristan, I've got a new favorite: the March 9, 1935 Met broadcast -- an incredible performance. Even with all the time freed up by the healthy cuts, Bodanzky must've been rushing to catch the 11:30 to New Haven; that said, he actually makes the quick tempi work -- it's conducting of great fire and intensity, almost what I'd imagine a Toscanini Tristan would sound like, yet Bodanzky seems more flexible, more of a "singers' conductor." His sudden death must've been a tremendous blow to the Metropolitan, especially given the fact that his successor was the prosaic Leinsdorf.

Both Flagstad and Melchior are sensational, as great as I've ever heard them. Flagstad is even fresher-voiced than at Covent Garden, and Melchior simply outdoes himself with gorgeous lingering mezza voce effects in the lyrical passages and shattering power in the climaxes. The performance has a certain spark of excitement that the Covent Garden recordings lack. Perhaps it was the enthusiasm following Flagstad's spectacular debut just a few weeks before. I'd guess that nobody of our age or younger can possibly imagine the sensation she created. This is a souvenir of those exciting days.

The sound is what it is. One has to take a lot of both the singing and orchestral work on faith. The former is again affected by the singers wandering on and off mike, and the boxiness of orchestral sound makes for loss of color and nuance. Still, I find it more listenable than the Reiner: the singers are in general more closely-miked and can be heard more of the time; as for the orchestral sound, both recordings are fairly wretched. If you're familiar with the sound of other 1935 Met broadcasts -- the Ponselle Traviata or the Tibbett Rigoletto, for example -- the Met recording is about on par with those: excruciatingly bad sound by normal standards, but experienced listeners should be able to hear through the muck.

Flagstad and Melchior are of course the draw here. We can go another century or two and still not hear Wagnerian singing like this again. I'm very happy to have it all on CD in any form.

The set is the product of an outfit called the West Hill Radio Archives, the proprietor of which is one Don Hill. It has been intermittently available on eBay. The artwork and annotation are very professionally done. Considerable time, effort, and affection was clearly lavished on this project. Here's the label and catalog number:

West Hill Radio Archives WHRA 6001 3 2003

Assuming the prospective listener is acquainted with the sound of murky antediluvian broadcasts, this is very highly-recommended. Should this be one's first or only Tristan? Certainly not. I can't imagine a recording with so much of the score missing and with such poor sound serving as one's reference or introduction to the work.

For me, the near-ideal would be the Boehm on DG, which in additon to its other merits should satisfy in the soprano department (Birgit Nilsson). After that, perhaps the Furtwaengler/Flagstad, for the great conducting (albeit less intense than his live work) and the soprano in excellent late form (those famous non-high C's aside), all in good mono studio sound.

Of course, it's a faint echo of the Flagstad of 1935 vintage. So I do think one would want to get at least one of the Flagstad - Melchior recordings later, if only to hear how the music can be sung by two superhuman voices born to sing it, both at their absolute peaks.


Our opera recording legacy: shot to hell

Having just listened to Act I of the reel-to-reel tape edition of the glorious RCA recording of Aida (Price, Vickers, Solti, et al.), I decided, just for fun, to pop the CD's into the machine and give them a listen. This would be the 3-disc London set released in the mid-'80s--the full-priced set. (Decca has since re-released this as a budget-priced *twofer*--about one-third the price! And the sound of the twofer may be better. But I'll get to that anon.)

According to my possibly faulty recollection, the CD's were grossly inferior to the tapes (of which I previously had the highlights), with a sharp attenuation of the high frequencies.

My CD player is now set up to perfection. I recently discovered that I had moved its AC plug from the wall socket into a power strip when I was working on a telephone jack in that vicinity; seems I had forgotten to move the CD player's AC plug *back* into the direct wall socket, where it always has sounded *distinctly* better--I kid you not. A hard-edged glare to the sound was thereby completely removed; I'd spent several months trying to figure out how to further tweak my CD player in order to get rid of the sonic annoyances, not realizing that a simple replacement of the plug to the wall socket would've solved the problem in two seconds. Live and learn.

Anyway, the CD player can now display any disc in its best light. If it's a good CD, it'll sound good on this system. If it's bad, it'll sound bad. If it's spectacular--and there are more than a few--it'll sound spectacular.

You know the punch line. The CD's of the Solti Aida were in fact shorn of their highest frequencies. This is not one of these subtle differences that only audiophiles with "golden ears" and $40,000 systems can hear. A 90-year-old guy with his hearing intact could hear this on his Victrola (assuming it plays CD's).

Why this is the case I do not know. My guess is that it was a "creative" choice: the Solti Aida was always known to sound extremely "hot" in the high frequencies; the brass was said to be over-recorded and obtrusive. Of course, this was *Solti's* creative choice (and/or possibly that of Lewis Layton, one of the greatest of the great engineers), which should have been -- but wasn't -- respected.

The other possibility is that something happened to the RCA mastertapes during their residence at Decca. All RCA issues that I've heard -- the open reel tapes, both highlights and complete, and the complete LP shaded dog set -- sound great, and have their high frequencies intact.

Same goes for the first Decca LP's. These are the ones with the cover art that featured Price in full costume with one of her '60s-style hair-do's, standing next to an Egyptian column. Unfortunately I was unable to find any scanned art for this set on the Internet. The bottom line is that these LP's sounded superb, in every way the equal of the RCA shaded dogs.

This set was manufactured by Decca/London until the early '70s, at which point it was replaced by LP's featuring the cover art featuring Price with '70s-style Afro and flowing dashiki.

This was the first issue of the Solti Aida where the high frequencies were attenuated. I've heard the LP's but not the cassettes, which were popular back then. I assume they were essentially the same as the LP's.

The next major issue of the Solti Aida was the first CD edition in the mid '80s--the full-priced three-disc set. These sound as if the same source material was used in their mastering, indicating that when Decca first issued the LP set described above (Price-with-Afro), they made a new set of masters which were then used for the first CD set, which shared the Afro cover art with the LP's.

The next set issued by Decca was the budget-priced twofer. (They were able to fit it on two discs by making a split during the Triumphal Scene; the earlier CD set was sequenced so that the discs broke at the conclusions of acts, which necessitated three CD's.) That issue has yellowish nondescript abstract cover art.

I've heard this set only via an .mp3 download. While it doesn't sound as bright as the original sets, it does sound more open and livelier than the "bad" sets. Of course, it's very difficult to make any definitive determination on the sound quality even via a fairly high bitrate .mp3 download. I'd have to rate the sound of the newest set as "inconclusive"

Believe it or not, both of the CD sets appear to still be in print! The three-disc set is far higher-priced than the twofer and may not sound as good. Go figure.

Just so you don't think I'm making a mountain out of a molehill here, let me just state that I consider this recording a very important document of some great singers and a famous conductor, and it should be heard in its proper sound. The same, unfortunately, goes for the Price/Karajan Tosca, which similarly lost its highest frequencies at Decca around the same time as the Solti Aida. There are further examples, some more blatant than others, but the fact remains that a huge number of recordings aren't being heard as they should be heard.

There have been some positive signs. The Living Stereo series on CD was a praiseworthy effort by RCA to present older recordings -- including opera -- in their best light. A few are so good (the Leinsdorf Tosca and Lucia) that they rival the original LP's. Unfortunately, not all of the recordings have been as worthy of the effort as the Solti Aida. Certain of the reissue programs of the other majors have not generally been of the same high standard, including the Decca Classic Sound/Legends series and the EMI Great Recordings of the Century. Conversely, Philips and DG have done some very nice work on their reissue projects, often limited only by the disfigurements of the CD format itself -- something worthy of another essay or two. While there are no hard and fast rules, the sad truth is that a very high percentage of our recorded opera legacy is not being properly presented to the buying public.