16 October 2011

The new EMI Wilhelm Furtwängler box-set -- equivocal sound improvement in Beethoven and Brahms symphonies

In a previous blog entry, the mistakes in documentation of the newly issued EMI box were mentioned. As the Beethoven and Brahms symphonies within this box-set were remastered based on the remastering for SACDs issued in Japan, it was hinted that there would be an improvement in sound compared with previous remastered versions. This arouses my curiosity, so I went to my friend Savio's home for A-B tests as he has high-end gears, quite unlike my own at best mediocre system. Savio's system is shown below.


CD transport : Marantz PMD331

D/A converter : Weiss DAC 2

Pre-amp : Primaluna Prologue Three

Power-amp : Primaluna Prologue Five(bridged) x 2

Speakers : Dynaudio '30th Anniversary' Sapphire

We gathered different versions for comparison as shown below.

Versions for comparison

Brief notes on the sound difference:
For Beethoven 1&3, the new remastering, compared with the old References version, has better sound in terms of details and richness, but Savio has the impression that the old version sounds more Furtwängler-like in line with his impressions of "Furtwängler-ness".

For Beethoven 5, the new version has warmer sound but the Japanese ART version (TOCE-59006) has a more distinct sound stage particularly where the brass and string pizzicato are concerned. The Japanese 2088 version has much surface hiss with a limited dynamic range, really disappointing compared with the other 2 versions.

It is interesting to compare the old References version of Brahms First Symphony with the new version. The difference in details and sound is apparently not much, but the overall impression tells quite another story. The old version has a tympani-dominant opening while the new version has a string-dominant one.

Where Brahms 4 is concerned, although the old References CD has muddy and turbid sound compared with the new version, oddly enough the brighter and more limpid new version sounds disordered in terms of orchestral balance, which gives the listener an umcomfortable feeling. The best sounding in this group is the Japanese ART version (TOCE-59003) with a firmer bass, more beautiful strings and woodwinds.

The overall impression is that the new version sounds more 'modern' -- cleaner and brighter, with little background tape hiss -- but not necessarily more satisfying musically.

P.S. Thanks are due to Savio's wife Julie for her patient help in putting the CDs into the player without our knowing which is which, thus allowing a blind test.

5 comments:

  1. Impressive comparison!
    Very helpful.

    ReplyDelete
  2. This is really disappointing to me (a professional sound engineer.) What does 'sounds more like Furtwaengler' mean? And saying that a transfer is 'tympani dominated' while another is 'string dominated'?? Or that something is "not necessarily more satisfying musically"?

    This is so right brained, elusive, and vague that one has very little to go by.

    And, nobody mentions this, which I noticed:

    In both the August 1954, and the Nov 1943, live performances LvB Fifth, WF has the tymp/doublebasses actually anticipate the beat by a FRACTION of a second at this same point (track 7 - 4:01 elapsed time in the new SACD version), leaping over the end of the last phrase. In the 1950 version it's displaced in the other direction--after the beat--by the same amount. It HAS to be an edit point, the startup of a new take I suspect.

    Now it gets really interesting. I just compared the EMI References issue with the new one, at this point in the closing of the third movement. The difference in fidelity is, here, rather astonishing: the new one has MUCH more in the high end. And, the OLD one has that passage with the basses RIGHT on the beat! The editor, Hardwick, must have overlapped the takes, or used a different one! There is a microscopic difference in the new one, which is why it stood out when I heard it. I've heard the Lp issue, and the References CD, many, many times stretching back to the late sixties--and I never heard this phenomenon before. So, Simon Gibson must have put in a re-take insert, but not QUITE on the beat. I would not have done that!

    Further proof that it is, indeed, an entirely different transfer from the older EMI CD issue.

    EMI producer/engineer Simon Gibson explains in his YouTube video that though the master is on a 10" tape reel, the individual takes were long enough just for one side of a 78 rpm disk, the intended release format. So they were all under about 4'20" in length. This makes sense--because obviously this was the start of the NEXT side--and WF & the orchestra were not 'up to speed' as it were. They were apparently starting that spot in the score 'from scratch'.

    Indeed, in the old EMI References transfer, the passage IS OVERLAPPED so that the double basses do not come in slightly after the beat -- as they do on the new transfer -- but they come in ON the beat. In the live performances, in fact, WF brings them in slightly ahead of the beat. But, WF also does this in Berlin--and how! He even hits is HARDER, ahead of the beat, at that point, with a huge SMACK on the tymp.)

    So, Simon Gibson introduces an odd effect--since WF does not play the passage that way on either of the live performances, surrounding the 1950 recording.

    That means that Keith Hardwick is a more musically-cognizant restoration producer though I certainly can't complain about the sheer SOUND of Gibson's transfer.

    Now, I wonder how many of the alleged "Furtwaengler experts" out there--they must be legion--will notice this, and comment? Not one?

    ReplyDelete
  3. Hi(who?),
    Thank you for you critical comments and shrewd observation and detailed analysis of the minute difference you mentioned.
    Savio said his impression of "Furtwangler-ness" (his choice of the term) is obviously right brained.
    The "tympani dominated" and "string dominated" terms refer only to the opening of Brahms 1st. The impression is a result of the different EQ applied to the transfers resulting in more prominent strings or tympani in each case.
    Musical experience is always a subjective plus objective thing. I've never pretended to be "scientific" in presenting my subjective impressions. I don't listen to music with my brain only, I also listen with my heart. This duality makes listening to music so fascinating an experience.
    Horace

    ReplyDelete
  4. In both the August 1954, and the Nov 1943, live performances LvB Fifth...

    From the dates,it should rather be LvB Seventh...

    ReplyDelete