Bolero

[Deleted]

Are the following 1966 Boleros all the same?

DG Originals
EMI "Les grands classiques"
DG "Karajan Collection" 2 CD
DG Galleria

[charlesb]

I doubt that the EMI CD contains the 1966 performance. It is probably the 1977 recording he made for EMI.

As you probably know, Karajan recorded Bolero 3 times for LP or CD. The timings for the 3 recordings are:

1966 16:08 DG Analogue
1977 16:07 EMI Analogue
1985 15:46 DG Digital

The recording on the 3 DG CDs you mention is the 1966 recording. However, the timings listed on different CDs seem to vary slightly.

It is amazing that 2 recordings 11 years apart should be so close in duration.

[Deleted]

Thanks. Yes, the EMI is the 1977 version.

However is there possibly an additional version? Karajan - Complete Emi Recordings 1946-84 Vol 1 (88 CDs at USD $2 per).

According to arkivmusic, that Bolero comes from 06/1971, at 16 Minutes 17 Secs.

[charlesb]

I checked the booklet for the box set. The time for the Bolero is listed at 16:07 . However, there is a mistake in the booklet. It lists the Bolero as being recorded in 6/1971, implying that the Bolero was recorded at the same sessions as La Valse and Alborada del Gracioso. The box set booklet also wrongly states that the orchestra was the Orchestre de Paris, not the Berlin Philharmonic.

The disc in the boxed set is an unchanged reissue of the EMI Studio Plus CD, which correctly lists the date and orchestra for the Bolero as 1/1977 with the Berlin Philharmonic.

[Deleted]

A Life in Music, p. 487:

"A game that Karajan liked to play on long flights. With someone timing him, Karajan would mentally play thru a section of a piece. He would then do it again: the trick being to get the thing identical next time round to the nearest millisecond."

Considering Bolero equals 339 measures for timing purposes (340th is a single quaver), I used a stopwatch to time measure #2 of different versions, as well as the overall timing up to the last quaver. (For reference, 16 minutes = 960 secs)

According to docs, may be 66 or 72 quarters per minute, so for example (60 sec / 66 notes) (3 notes / measure) = 2.727 sec / measure.

The results are as follows. The format is:

Year of Recording / Conductor / Measure2 (sec) / Measure2 times 339 (sec) i.e. theoretical overall / Actual Overall (sec)

1928 / Ravel / 2.78 / 942.42 / 934.50 (7.92 secs faster than theoretical)

Mar 1966 / HvK / 2.94 / 996.66 / 961.65 (35.01 secs faster than theoretical)

Jan 1977 / HvK / 2.94 / 996.66 / 961.19 (35.47 secs faster than theoretical)

Dec 1985 / HvK / 2.72 / 922.08 / 938.97 (16.89 secs slower than theoretical)

Dec 31 1985 (live) / HvK / 2.72 / 922.08 / 941.84 (19.76 secs slower than theoretical)