Recording of January 1984: Debussy: Three Nocturnes; Jeux

rotm184.pjil.jpgDebussy: Three Nocturnes; Jeux
Concertgebouw Orchestra of Amsterdam, Bernard Haitink conducting.
Philips ACD 400-023-2 (CD).

This is the first classical CD I have heard that was originally mastered on analog tape, and the sound is quite different from what I'm accustomed to hearing from the silver discs.

I had read so many critics' complaints about excessive background (tape) hiss from analog-mastered CDs that I was fully prepared to be appalled. I wasn't. Perhaps my speakers (Watkins WE-1s as of now) are smoother than what some other critics listen to, perhaps I prefer a more subdued high end than some, but I did not find hiss to be a problem with this Philips disc. Yes, it is audible at high listening levels, but it is not a ssss, it is a hhhh, like the sound of a very gentle rain far off in the background. I have heard worse hiss from microphone preamps.

This is one of the best orchestral recordings I have heard from a major label. The chorus is placed, concert style, behind the orchestra, and sounds suitably ethereal and distant (no chorus mikes, maybe?). And the orchestra/chorus balance is perfect. There is no question but that, by comparison with digitally mastered material, the analog tape did round off some of the edges to the sound, but that sound is so suavely rich and effortless that it is easy to become completely immersed in the music while forgetting the sound—most of the time. My only real complaint concerns the violin section, which is not only too prominent, it also has a slightly hard, roughish edge riding on it.

Despite the multimiking, the recording has a nice feeling of depth and ambience, and instruments are located sensibly across the stage without being bunched together at the sides, although the channel balance is continually biased towards the left because of the overly prominent violins. This recording also suffers from something which bothers me about most multimiked recordings; it gives me the feeling of hearing too much inner detail in the orchestra. Instead of sonorities I hear the individual instruments which should be blending to create those sonorites (footnote 1). And I doubt that your average concertgoer is accustomed to hearing the flutist inhaling before an entrance or the conductor whooshing before eliciting a forte. I had never noticed either from my old (ca 1980) analog copy of this, but the lower noise and enhanced detail of the CD made these incidental noises noticeable enough to draw attention to them. (Now that I know where they are I can hear them on the analog disc, but barely.)

The short ballet Jeux was commissioned by Diaghileff and premiered in 1913 with choreography by Nijinsky. It was coolly received and has rarely been staged since, although there have been a few recordings through the years. Based on the view of love as merely a game, like tennis, the music represented a turning point in Debussy's style. Instead of languid dreaminess, Jeux is quixotic, highly rhythmic and at times awkwardly angular.

The performances? My mind's ear tells me I have heard somewhat better ones of the Nocturnes, but since I cannot recall which they were—an ancient one on 78s and an early LP with Ansermet and this same orchestra were two, but they are hardly relevant in 1984—I will simply say that Haitink's readings are among the most persuasive of those currently on records, and almost certainly the best-recorded. For Jeux I have little frame of reference, having heard it only once many years ago in concert. But I find this performance completely captivating, and could not imagine it being done any better. What else can I say? Despite my cavils, then, this deserves inclusion in our growing list of "Recommended Recordings."—J. Gordon Holt

Footnote 1: In the early 1950s, when Sir Adrian Boult was engaged in the formidable task of recording all of Ralph Vaughan-Williams' symphonies for English Decca, the composer sat in on some of the recording sessions to lend spiritual and artistic support to the undertaking. During the setup tests for one recording, the recording director reported that in the opening passage the English horn was not audible through the rest of the orchestra. Before the director could move to correct the problem, Mr. Vaughan-Williams spoke up. "That's just a warmup for the player," he explained. "It isn't supposed to be audible." Further proof, I maintain, of the fundamental inappropriateness of multimiking.

volvic's picture

What a great review, brings back a lot of memories. The performance as Holt says is great and after many years I cannot think of any other version that betters this. I will pull my disc out and listen tonight.

hollowman's picture

Yes, the SQ is very good. And the performance is decent as well (Haitink can be a bit dull in some recordings, but not this one).
For comparison, I also pulled out my copy of Boulez/Cleveland SO/Deutsche Grammaphon ... an all-digital recording from the early 1990s.

....I was surprised that sonics on the Philips is better [ and that is almost entirely due to engineering -- i.e., mike set up, pre-amping, mixing levels and hall acoustics -- rather than "analog vs digital"] .
I somewhat disagree with JGH about multi-miking and too much inner detail. I feel that indiv instruments (and or instr. groups) sound WAY better if closely- and/or multi-miked ... virtually all aspects of an instruments sonics, IMO, diminishes as the mike is moved further away from it.
OTOH, an engineer can adjust levels on a closely-miked instruments w/o much loss of quality/detail.
There are a few exceptions to this ... e.g., audibility of oboe valve "clicks" can be downright distracting ... (here's an example of a poor engineering) ... the engineer has to be especially careful of mike placement with double-reeds.