After Jarvi's absolutely sensational 2012 Sleeping Beauty (also with the Bergen Phil on Chandos), I was expecting an equally illuminating and involving Nutcracker. But alas, this sounds like it's the last in his series to be recorded (it is) and he wants to get it done with as quickly and effortlessly as possible (he does).
I listened to this CD last December (2015) when it was first released. My initial reaction to it was an overriding annoyance with Jarvi's tempos and some insertion of extraneous, unnecessary dynamics not marked in the score (subito piano/crescendos and the like). But watching it collect dust, sitting on my shelves all year, I've been bothered by a nagging disbelief that Jarvi could do such a disservice to this beloved masterpiece. So, now that the cold weather has arrived, and daylight saving time has expired, and we near Thanksgiving (2016), I decided to give it another listen. And, well, in all truthfulness, it's really not THAT bad. And some sections of the score can be enjoyable.
Here's my take on this. I believe Jarvi recorded this ballet with Chandos hoping it would fit onto one CD. And when it was all in the can, it was SO close to being able to fit, but not quite. So perhaps - just perhaps - Chandos asked Mr. Jarvi to target a few sections of the score and re-record them at a faster tempo in an attempt to make it all fit. And I believe he did just that.
Of course, this scenario exists solely in my imagination and there is not a shred of evidence to support these imagined conversations ever occurred. But...let me explain why I think it just might be possible.
The first three tracks (the Overture, Scene 1 and the March) are absurdly fast. So much so that the Overture and March, especially, sound like a 45 record played at 78-rpm speed. (I'm revealing my age here, but I'm not exaggerating.) I can see why I was so put off by this last year. These three opening sections are completely unmusical and, frankly, absolutely ridiculous. (I'm surprised the Bergen players could manage all the notes at these speeds.) But then, beginning with the Entrance of the Parents (track 4), Jarvi settles down and begins to make music. And he continues to make music for much of the remainder of the score. Oh, there are sections that are still too fast. For example, the Waltz of the Snowflakes moves along like a blizzard, and at least two of the characteristic dances are absurdly fast (the Chinese Danse and, especially Trepak - which is taken at a breakneck speed). But Jarvi is known to fly through certain passages of ballet music, and these actually sound rather typical of him.
So it's those opening 3 sections (and few others) which don't fit the remainder of the reading. And they are not typical even of Jarvi. That's why I have dreamed up the imaginary scenario above to help explain it. Because, while it's impossible to ignore, if one were to forgive and forget and get past the opening three sections of the score, much of what remains is rather enjoyable. Realistically, taken out of context of the stage action, it's surprising how well this music benefits from swifter, more moving tempos. And, as usual, that's what Jarvi provides. But it simply goes too far here, and the sense of magic, imagination and, especially, charm are largely absent. And it must be noted, other than a few fluffed passages here and there (most notably from the piccolo player), the Bergen Philharmonic plays superbly, especially considering the very swift tempos they had to accommodate.
Aside from tempos, I was bothered by a few instances of inserted, unmarked subito piano/crescendos. But listening again, I was surprised to hear that Jarvi does it far less often than does Rattle in his EMI Berlin recording (which would have been a glorious reading of the score if it weren't ruined by this nonsense all through it). I am still not happy with Jarvi's sfz-p-crescendo on the last note of the 1st Act, but I'll forgive him that one sacrilege.
And there are other oddities. One big one is the insertion of TWO chiming of the midnight bells. Why, at the beginning of track 7, are there midnight bells just before the guests leave, and then again at the start of the dream sequence (4 minutes later), where it is notated in the score? Midnight came twice on this magical evening, apparently. There was a 30 second section they could have shaved off the playing time to save some room! And why is the tambourine so loud and forwardly placed in Arabian Dance?
I can't remember a time that I didn't enjoy listening to The Nutcracker (even played by amateur and community ensembles). Last December, I certainly didn't enjoy this one. But, hearing it again, I wondered if I was unfair to Mr. Jarvi. For comparisons this time around, I listened not to my usual favorites (listed below), but to two august staples of the recorded repertoire - those by Gergiev (Kirov Orchestra) and Bychkov (Berlin) - both for Philips. Jarvi certainly fares well in this company. Not that those Russian conductors are bad, but to me they have not stood the test of time well. Their performances sound rather perfunctory and routine. And Gergiev, who also manages to fit his recording onto one CD, sounds rushed and glossed-over. More significantly, however, both recordings suffer from what I call "early digititis" - bright, metalic upper range (especially metal percussion), some odd balances and spotlighting, and rather thin, artificial acoustic. They sound woefully "digital" compared to the full, warm, glowing (analog-like) Chandos sound.
In closing, despite what musical merits can be ascertained from Jarvi's recording, ultimately it cannot touch the magnificence of classic accounts like those from Dorati/Concertgebouw (Philips), Rodzinsky/RPO (Westminster) or Ozawa/Boston (DG) for life, color, magic, inspiration or imagination, much less quality of playing and recording. Year after year, these three recorded versions sound fresh and alive like none other I have ever heard. And while it's astonishing that Chandos managed to squeeze over 84 minutes of music onto one CD, sound quality and performance values suffered. It is impossible to ignore the opening three sections (and even a couple of the dances later on); they really sound like a compromise. How I wish Chandos would have taken the 2-disc approach, giving Jarvi a chance to breathe a little more in this score (and making music out of the first three sections), and adding a coupling. One such coupling comes to mind as being a perfect fit, and in keeping with this ballet series, that being Stravinsky's Le Baiser de la fee (The Fairy's Kiss) - which is comprised of his arrangements of Tchaikovsky's music. However, I understand the enormous expense involved in additional recording sessions (and a 2nd CD), and the fact that Jarvi has already recorded this ballet for Chandos (with the Scottish National Orchestra in 1984). So we are left with what we have. A mixed bag - not as bad as I first thought, but certainly not one to warm to year after year.