Liszt's Grande Etudes de Paganini are basically transcriptions of five
of Paganini's Twenty-Four Caprices, opus 1. Liszt also made an arrangement
of the rondo theme from Paganini's second violin concerto, and this study
is known as La Campanella (the third study in the set).

The first study is based on the fifth of Paganini's Caprices. To this
Liszt added the Introduction (Prelude) and Coda from Caprice No 6. The
first part is a study for the left hand only (tremolo and melody), but
then the right hand enters and dominates after about 25 bars.

The second study is based on Paganini's No 17. Liszt inserted Paganini's
introduction after the trio section and he also added a coda. The study
contains many descending scales as well as Liszt's famous alternating
octaves.

The third study, La Campanella, is concerned with leaps, where the
accompaniment lies on fixed notes above the melody. Maybe Liszt found
inspiration in the 2nd Caprice, where Paganini employs the same treatment
to a similar melody.

The fourth study is a straightforward transcription of Paganini's No 1.
It should however be mentioned that Liszt replaced Paganini's Andante
indication with Vivo.

The fifth study (Paganini's No 9) is also a very straightforward
transcription with just some clarification of the harmonic progression
in the middle section.

Finally, the sixth study is based on the famous 24th Caprice with it's
Theme and 11 Variations. Liszt basically leaves Paganini's variations
unchanged and only underlines the hidden theme in each variation. Again
the harmonic progression is more worked than in Paganini's original,
just as in the middle section of the fifth study.

Liszt's first version of the Paganinis studies was called Etudes
d'execution transcendante d'apres Paganini, and was written nearly
fifteens year earlier (in 1838).  This early version contains many
difficult passages, and are very difficuly to play on a modern piano.
In fact the piano developed quite a lot during Liszt's lifetime, and
this was one of the reasons why Liszt rewrote some of his early pieces.
Liszt's revision of the Paganini studies also resulted in piano pieces
that are more faithful transcriptions of Paganini's caprices.

Comparison of recordings: Hamelin vs Howard (both Hyperion)

Let me first state that Hamelin and Howard belong to slightly different
schools of pianism. Hamelin is more of a legato pianist, while Howard
strives to make every tone audible. In this review I will try to focus
on the handling of details in the score, phrasing and the structural
coherence.

Study No 1: Here Hamelin plays the Prelude in tempo, ignoring the quasi
cadenza indication. Howard plays more cadenza-like and emphasizes the
resting points.  In the main section of the study, Howard delivers a
convincing interpretation which hardly can be bettered: three different
types of accents and a fine build-up towards the fortissimo. Furthermore,
there is also a fine balance between the left hand and the right hand,
and every tone is audible.  Hamelin is of course not bad, but he does
not stand a chance in this study.

Study no 2: A more balanced battle, where Hamelin's brilliance stand
against Howard's clarity.  In the middle section (poco piu animato)
Hamelin chooses a brisker tempo than Howard, wherease the ideal tempo
probably is somewhere between Hamelin's and Howard's. I would probably
have chosen Hamelin's version for his outstanding interlocking octaves,
but unfortunately Hamelin chooses to insert an ascending scale just
before the choral ending in the coda. This may be the 'crass flourish'
Howard mentions in his sleeve notes (written four years earlier). Indeed
this flourish connects quite poorly with the choral ending.  Instead,
the price for the best recording I have heard (and still remember) goes
to Tamas Vasary on DG.

Study no 3: A very close call again. Hamelin, Howard and Wild (on
Quintessence) are all equally fine. Hamelin is the best in the coda, but
on the other hand he plays the last statement of the melody (just before
the coda) too loud. In one bar he adds a little embellishment  to a
broken Gsharp minor chord.

Study no 4: Here Hamelin is closest to Liszt's Vivo indication, but
Paganini's original Tempo indication is Andante, so it is not entirely
clear how fast this study should be played. Hamelin chooses to 'rest'
on some of the semi-quavers (fermati), which distorts the rhythm in an
otherwise very good performance. Howard delivers a solid performance.

Study no 5: just as in the fourth study, Liszt emphasizes the non
legato character of the study already at the beginning of the piece.
Nevertheless, Hamelin plays the flute imitation with a slight legato
touch (the semiquavers), while Howard plays non legato in the flute
imitation as well as in the French horn imitation. Howard has got the
edge in this one.

Study no 6: Just as in the first study,  Howard succeeds in getting all
the accents and dynamics right. Hamelin takes a different (and interesting)
view on some of the variations, but in Variation 5 he plays a heavily
distorted version of what Liszt actually wrote down. It is strange that
Jed Distler (Classics today) did not notice this.

Mikael Rasmusson