Bach: “Sei Solo” — Sonata No.2 in A minor for Violin Solo, BWV 1003

Media Review / Comparison

2022-02-21 — Original posting

Table of Contents

Introduction — The Recordings

This posting is about the Sonata No.2 for Violin Solo in A minor, BWV 1003, which Johann Sebastian Bach (1685 – 1750) published under the title “Sei Solo” (see the title page above). I am comparing the 25 recordings in my collection:

WikiWeba' = 433ReviewArtist, Media
WikiWeba' = 440ReviewArtist, Media
WikiWeba' = 415ReviewArtist, Media

Weba' = 415ReviewArtist, Media
a' = 415ReviewArtist, Media

Weba' = 440ReviewArtist, Media
a' = 440ReviewArtist, Media
a' = 440ReviewArtist, Media
WikiWeba' = 440ReviewArtist, Media
a' = 415ReviewArtist, Media
WikiWeba' = 440ReviewArtist, Media
WikiWeba' = 440ReviewArtist, Media
WikiWeba' = 440ReviewArtist, Media
WikiWeba' = 440ReviewArtist, Media
a' = 433ReviewArtist, Media
WikiWeba' = 415ReviewArtist, Media
WikiWeba' = 443ReviewArtist, Media

Weba' = 415ReviewArtist, Media

Weba' = 440ReviewArtist, Media
WikiWeba' = 415ReviewArtist, Media
a' = 415ReviewArtist, Media
1967HenrykSzeryng19181988WikiWeba' = 440ReviewArtist, Media
a' = 440ReviewArtist, Media
WikiWeba' = 440ReviewArtist, Media
a' = 440ReviewArtist, Media

Explanations on the Table

  • You can sort the table by any specific column (in ascending or descending order) by selecting the respective title field.
  • The first field is the year when the respective recording was completed (not necessarily identical to the ℗ or © years).
  • The birth year is not known for all artists.
  • The fields “Wiki” and “Web” are links to the respective artist’s Wikipedia entry and/or personal Website.
  • The highlighted column “Review” contains links to the respective entry in the comparison section (The Interpretations, Detail) below.
  • The green column “Summary” contains links to the respective entry in the comparison summary, featuring detailed Media information, as well as notes on artist, instrument, recording, etc.

Media Information

Details about the media (CDs) are available as part of the Comparison Summary posting on Bach’s Sonatas and Partitas for Violin Solo. That information includes cover image, title, artists, technical media information (label, label-number, booklet info, barcode, amazon link, where available, plus additional information, as deemed relevant). That summary also features an overall comparison table.

About the Sonata No.2 in A minor for Violin Solo, BWV 1003

Within the “Sei Solo“, the three Sonatas for Violin Solo formally are baroque “church sonatas”, following a four-movement scheme slow — (fast) — slow — fast. In all three Sonatas, the second movement is a fugue, with the complexity growing towards the third sonata. I don’t need to give a detailed introduction to these pieces, as they are all well-known. However, you do find some additional information on the Sonata No.2 in A minor for Violin Solo, BWV 1003 in several of my concert reviews:

The Movements

Bach completed his “Sei Solo” around 1720 in Köthen (Anhalt). Bach’s original manuscript (see the three images in the header section) survived to this day. There is also a beautiful manuscript, now identified as being a copy dating from 1727–32 by Bach’s second wife Anna Magdalena Bach (1701 – 1760). The facsimile of Bach’s manuscript (as well as of Anna Magdalena’s copy) can now be downloaded from IMSLP.

In lieu of basic explanations on the individual movements, I’m just including short excerpts from that document, showing the first 2 – 3 lines of each movement. Additional comments are added to point to critical aspects in the comparison.

I. Grave

Bach: Sei Solo, Sonata No.2 for Violin Solo in A minor, BWV 1003, I. Grave, score sample
Bach: Sei Solo, Sonata No.2 for Violin Solo in A minor, BWV 1003, I. Grave, score sample

Slow movements can be really tricky! The problem is that the score is full of small notes—demisemiquavers and even shorter values. Yet, the movement is in 4/4 time, and these small notes are ornaments, not melody. Albeit Grave, I think the artists should make at least an effort to let the listener feel the underlying crotchet foundation.

A Look at the Ending

The last two bars in the Grave deserve some extra attention. Preceding to the final bar, Bach writes two crotchets with parallel sixths, f’/d”f’♯/d”♯, that resolve into an octave e’e” in the last bar. There are wavy lines above both notes of the first crotchet. In the d”♯ (second crotchet, upper voice), that wavy line changes to tr, i.e., a trill. The first sample below is Bach’s original, the second one is the copy by Anna Magdalena Bach:

Bach: Sei Solo, Sonata No.2 for Violin Solo in A minor, BWV 1003, I. Grave, score sample, final bars, JSB
Bach: Sei Solo, Sonata No.2 for Violin Solo in A minor, BWV 1003, I. Grave, score sample, final bars, JSB (source:; public domain, CC-BY 4.0)
Bach: Sei Solo, Sonata No.2 for Violin Solo in A minor, BWV 1003, I. Grave, score sample, final bars, AMB
Bach: Sei Solo, Sonata No.2 for Violin Solo in A minor, BWV 1003, I. Grave, score sample, final bars, AMB (source:; public domain, CC-BY 4.0)
Trill, Regular Vibrato, or Bow Vibrato?

In more recent, in particular historically informed interpretations, artists realized that for a plain trill, the tr mark would be on the first crotchet, which it isn’t. Some of the artists therefore limit the trill to the second crotchet (and typically to the upper note, d”♯, only). Across the 25 recordings, the wavy lines are interpreted in a variety of ways:

  • As plain trills on both crotchets, usually on the upper note only. My least favored option: to me, it does not follow the notation.
  • As bow vibrato, typically a single-tone, slow tremolo (tremblement, slow “shuddering”), by periodically varying the speed of the bow.
  • Some do a bow vibrato by leaving the bow on the lower string, but periodically altering the pressure on the upper string.
  • Others resort ordinary vibrato. Especially in (historically informed) interpretations which otherwise use minimal vibrato only (or none at all), I find this a legitimate and viable option.

Some may find the bow vibrato “archaic”, or at least “pre-baroque”. However, Bach wasn’t a “modernist” in many aspects (his frequent use of fugues is just one prominent example). Hence, the occasional use of bow vibrato is at least worth considering.

I’m indicating the artist’s choice of option in the comments below. That choice is present some surprises!

The Artists’ Choice of Options

For the reader’s convenience, let me add a chronological list of these choices here:

  • Menuhin, 1936: trills (d” – d”♯) on both crotchets.
  • Szigeti, 1956: trills (d” – d”♯) on both crotchets.
  • Grumiaux, 1961: trills (d” – d”♯) on both crotchets.
  • Szeryng, 1967: vibrato on the first crotchet, trill on d’♯.
  • Kremer, 1980: vibrato on the first crotchet, trill on d’♯.
  • Zehetmair, 1982: vibrato on the first crotchet, trill on d’♯.
  • Schröder, 1985: bow vibrato on the first crotchet, trill on d’♯.
  • Paternoster, 1995 (cello): trills (g – g♯, on the violin: d” – d”♯) on both crotchets, combined with bow vibrato on B (second crotchet, f’♯ on the violin).
  • Huggett, 1997: trills (d” – d”♯) on both crotchets.
  • Podger, 1999: vibrato on the first crotchet, trill on d’♯.
  • Kremer, 2002: oddly, plain trills (d” – d”♯) on both crotchets (see 1980 above!).
  • Tetzlaff, 2005: bow vibrato on the first crotchet, trill on d’♯.
  • Mullova, 2008: bow vibrato on the first crotchet, trill on d’♯.
  • Baráti, 2009: trills (d” – d”♯) on both crotchets.
  • Ibragimova, 2009: bow vibrato on the first crotchet, trill on d’♯.
  • Khachatryan, 2009: vibrato on the first crotchet, trill on d’♯.
  • Beyer, 2011: trills (d” – d”♯) on both crotchets.
  • Pietsch, 2011: bow vibrato on the first crotchet, trill with mordent on d’♯.
  • Busch, 2012: bow vibrato on the first crotchet, trill on d’♯.
  • Faust, 2012: bow vibrato on the first crotchet, trill on d’♯.
  • Carmignola, 2018: bow vibrato on the first crotchet, trill on d’♯.
  • Pochekin, 2018: bow vibrato on the first crotchet, trill on d’♯.
  • Aldemir, 2019 (viola): bow vibrato on the first crotchet, trill on g♯ (the d’♯ on the violin).
  • Cotik, 2019: bow vibrato on the first crotchet, trill on d’♯.
  • Hadelich, 2020: trills (d” – d”♯) on both crotchets.

II. Fuga

Bach: Sei Solo, Sonata No.2 for Violin Solo in A minor, BWV 1003, II. Fuga, score sample
Bach: Sei Solo, Sonata No.2 for Violin Solo in A minor, BWV 1003, II. Fuga, score sample

Playing this fugue doesn’t just confront the artist with the technical challenges that all the triple- and quadruple-stop chords present to the left hand. In listening through these recordings, one realizes that the real challenge is not so much in these chords, but in musical phrasing, in maintaining the musical flow (and I don’t mean uniform flow, of course). And, at the same time, in letting the listener experience / follow the often hidden melodic lines, in particular the recurring theme(s) of the fugue. Plus, ideally, the artist should probably make the listener unaware of the technical challenges.

The fugue ends with a short cadenza, consisting of a descending sequence of four demisemiquaver figures in bars 286/287. Note that Bach does not write legato slurs:

Few of the artists nevertheless perform the cadenza with legato articulation: Menuhin (1936), Szigeti (1956), Szeryng (1967), Schröder (1985—strange!), and Paternoster (1995, understandably, on the cello).

III. Andante

Both the Andante, as well as the closing Allegro consist of two parts with repeat signs, i.e., an AA’BB’ scheme:

Bach: Sei Solo, Sonata No.2 for Violin Solo in A minor, BWV 1003, III. Andante, score sample
Bach: Sei Solo, Sonata No.2 for Violin Solo in A minor, BWV 1003, III. Andante, score sample

The main challenge is with the apparent conflict between the Andante annotation, the 3/4 meter, and the persistent quaver structure in the accompanying second voice. Performing Andante on the quavers bears the danger of the 3/4 meter getting lost. To me, that comparison offered a revelation when I realized how Isabelle Faust is reading Bach’s notation: she offers a brilliant solution to the challenge I mentioned.

IV. Allegro

Bach: Sei Solo, Sonata No.2 for Violin Solo in A minor, BWV 1003, IV. Allegro, score sample
Bach: Sei Solo, Sonata No.2 for Violin Solo in A minor, BWV 1003, IV. Allegro, score sample

Strictly speaking, also this movement leaves some challenges in the annotation. The movement is Allegro, but written in split time (₵ = 2/2 = alla breve). With the entire movement being a chain of semiquavers or shorter notes, it is impossible to perform an Allegro based on half-note beats. Throughout the movement, all notes in beamed together in groups of four semiquavers (or equivalent). Therefore an Allegro based on four beats per bar seems appropriate, and should be doable.

The digitized autograph is available from the Staatsbibliothek zu Berlin, Preußischer Kulturbesitz. This document is in the public domain and shared under a Creative Commons (CC-BY 4.0) International License. It is free of known restrictions under copyright law, including all related and neighboring rights.

The Interpretations, Overview

In order to provide a rating overview, as well as an idea about duration relations between the recordings, I have prepared the table below. Note that the color coding for the duration (blue = longer/slower, red = shorter/faster) refers to the average between the recordings.

Bach: Sonata No.2 in A minor for Violin Solo, BWV 1003 — comparison table
Bach: Sonata No.2 in A minor for Violin Solo, BWV 1003 — comparison table (© Rolf Kyburz)

I have not corrected the timings for trailing or leading blank time (the former especially with last movements, the latter sometimes with the first movement). Therefore, the timings in the above table should be read with a grain of salt.


Not all artists perform all repeats. In the cases where repeats were omitted (movements III, IV), the track durations can’t be used as indicator for the tempo. Therefore, I corrected these durations in the table (adding the time for the repeat parts to the track duration). In that sense, the overall duration (second-to-last column) is to be read as “if the artist had performed all repeats“. For the actual track and overall durations please see the section below. These may differ from the numbers in liner notes: I’m ripping the recording into Apple Music and use the times in the player software, which may use different rounding algorithms.

A Note on Ratings

First and foremost: all my ratings reflect my personal opinion, hence are inherently subjective. I use a 1 – 5 star rating scheme—simply because that’s what my player software (Apple Music) offers. I use the same scheme for concert reviews. You will note that for these, there are rarely reviews below a 3.0 (★★★) rating. That’s largely because I try to avoid concerts where I anticipate a marginal performance. And I stick to an “absolute” scale, where results below 3.0 are negative.

Ratings in Media Comparisons

In media comparisons, especially reviews involving a large number of recordings, I tend to use a relative scale covering the full range of (close to) ★ … ★★★★★, in order to achieve more differentiation among the many ratings. My rating criteria are similar to the ones in concert, such as

  • does the performance reflect the notation, i.e., the composer’s (perceived) intent?
  • does it present the character of the piece (e.g., in the dance movements in Partitas)?

Personal Views

My ratings also reflect how much a recording offers to me, particularly as a listener with interest in historically informed (HIP) performances. With this, I tend to give preference to HIP recordings. I do not mean to devalue the achievements of historic recordings by the great artists of the last century. However, time has moved on, and it is my belief that the in-depth encounter with HIP performances makes it hard(er) to enjoy some of the traditional recordings, especially romantic ones with heavy vibrato, etc. Again: this is my personal view, and I don’t mean to spoil the pleasure that the fans of past great violinists draw from their recordings.


I should also mention that audiophile arguments play a secondary role in my ratings. My primary focus is on the interpretation, not perfection in recording technique. The latter comes into play mainly where it affects the audibility, clarity and transparency, e.g., through excess reverberation. And for newer recordings, blatantly dull, “muffled” sound should also have an effect on the rating.

The Interpretations, Detail

The review comments below are sorted by recording year, from the oldest (1934) to the most recent one (2020). Note: for the artist’s life data, Website and/or Wikipedia entries please see the first table above. Note: in the artist segments below, the pitch is mentioned only where it deviates from a’ = 440 Hz.

Procedure, Technical Aspects

I listen to all recordings in full, typically even more than once. Note that the sequence of recordings below is not the sequence in which I listen to them. I have written about my comparison approach in an early blog post. In essence:

  • I go though the collection movement by movement, i.e., I start with listening to the first movement with all recordings before progressing to the next movement.
  • I try to choose a sequence that does not put subsequent recordings at a disadvantage. Typically, I start with slow performances, progressing to faster ones. At the same time, I try using a suitable sequence of historic vs. “conventional” vs. HIP interpretations.
  • Especially in large comparisons, such as this one, the sequence will typically vary from movement to movement.
  • In the sequence in which I listen to the tracks, I typically “just” move forward. If I relate to other interpretations, I refer to recordings I listened to previously, irrespective of the time of the recording. In other words: for older recordings I may may use comparisons to interpretations of artists who may not even have been alive at the time of the early version. That may occasionally sound strange. However, in the interest of efficiency, I can’t risk “jumping around” to amend comments that I have already written.
  • Naturally, my comments will mostly refer to the recordings immediately preceding the one I’m writing about—in the listening sequence for that given movement (it is impossible to memorize all performances in detail). However, I try my very best to make the ratings absolute, not relative.


It is a few months since my last review in this series. I try not to delay these reviews too much. However, the delay has the advantage that I can approach this review without bias from the preceding ones. I apologize for duplication with earlier reviews. One benefit of the text duplications is that they help making each review readable by itself, without an excess of cross-links to other postings.

J.S. Bach, Sonatas & Partitas — Yehudi Menuhin (CD cover)

Pitch: a’ = 433 Hz

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (4’29”)

Not nearly as slow as Joseph Szigeti, but still slow, with way too much emphasis on the ornaments, which are all “fully played out”. Traditional, of course, with some similarities to Kremer’s 1980 recording. Everything is emphasized, ornaments are melody—and featuring Menuhin’s nervous vibrato. About the final bars (see above): plain trills (d” – d”♯) on both crotchets.

II. Fuga (7’35”)

Menuhin’s articulation typically is a broad, accented portato, again filled with the expected, somewhat nervous vibrato. The latter does less harm than in slow movements, though, except for the expressive climaxes. The performance seems to state that agogics were not invented at the time of the recording. The artist relentlessly moves forward, often leaving a slight impression of pushing forward (already in the first bars), as if there was a need to make up for time lost in the occasional, slight rallentando.

There is dynamic variation, but (apart from the few f / p echo segments) essentially limited to long crescendos after a mf at the beginning of an exposition. In the mid-30s, meticulously following the composer’s notation wasn’t a priority. So, it’s no surprise that Menuhin’s demisemiquaver cadenza (bars 286/287) is legato.

III. Andante (4’40”, first repeat not performed)

This is one of two instances (the other one is in the Partita No.3 in E major, BWV 1006) where Menuhin omits the first repeat. No other artist does that anywhere. An accidental omission, caused by the artist performing by heart? People must have realized this, but repeating the recording may have been regarded not worth the effort.

It’s tough to judge a 1936 recording. The intonation isn’t always quite flawless—this may happen to today’s artists as well, but these have the opportunity to correct mishaps in a recording. Menuhin’s vibrato is nowhere nearly as nervous as Sergey Khachatryan‘s, nor as obtrusive and heavy as Joseph Szigeti‘s. But as the latter, also Menuhin performs the melody voice quasi-legato, almost throughout. Slow, romantic-traditional.

IV. Allegro (4’17”, second repeat not performed)

Well, it’s the time when many felt that fast movements in baroque music—Bach in particular—were to move along with the regularity a sewing machine. That’s almost what Menuhin’s interpretation feels like. He moves along almost like a steam roller—the only tempo variation in the entire piece is an excessive, broad allargando at the end. And Menuhin’s articulation is a monotonous, broad, resolute détaché. The one variation is that it is a little shorter in the p segments. There are no agogics, and essentially just two-level dynamics. I’m tempted to say that in comparison to today’s performances (HIP or not), this sounds like a MIDI player.

Total Duration: 21’01”

Rating: 2 / 3 / 2 / 2 = 2.25

Comment: An important historic recording, for sure—even though it can hardly compete with today’s performances.

J.S. Bach, Sonatas & Partitas — Joseph Szigeti (CD cover)

Joseph Szigeti, 1956

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (5’22”)

Sorry, no. This recording should not have been released. There’s the dreadfully slow tempo, which makes the listener lose track of the melodic line. And Szigeti makes no distinction between melodic skeleton and ornament. There is no recognizable, persistent overall musical flow or harmonic / rhythmic scheme. This way, the movement makes no sense at all. Plus, the slow, exceedingly heavy vibrato, combined with slow portamento, which makes some of this sound like the meowing of a tortured cat. About the final bars (see above): plain trills (d” – d”♯) on both crotchets.

II. Fuga (8’30”)

Not surprisingly, Szigeti’s fugue is very resolute, rigid, often almost “military”, sometimes really extreme and noisy. In complex chord sequences, the artist slows down: the tempo often is inconsistent, the broad rallentando at the end.

And of course, there is vibrato throughout—though this is most noticeable in the segments with broad portato articulation. And there, it sometimes obscures the intonation. Not surprisingly, the demisemiquaver cadenza (bars 286/287) is legato. I must say, though, that despite these critical notes, the interpretation shows that in 1956, the artist still performed with very respectable technique.

III. Andante (6’19”)

Very slow (luckily not nearly as slow as Sergey Khachatryan, though), calm, and an Andante on the quavers at best. I could live with that, but why this obnoxious, broad, heavy and strong vibrato on the melody voice, and why all the glissando transitions in the melody? More cats meowing … hard to listen to.

IV. Allegro (5’40”)

Joseph Szigeti’s bow technique still worked, as does his left hand. Except that the intonation in this movement is often approximate at best. The articulation alternates between portato (e.g., at the beginning of a phrase, or in the f motifs) and a fairly “woody”, noisy spiccato / sautillé (for the most part, but most prominently in the p responses in echo segments). Why? Just to make the performance (or the piece) sound virtuosic? And then there are a few rather prominent tempo instabilities. Yet, overall, it does not sound all that bad—it just can’t have been Bach’s intent.

Total Duration: 25’52”

Rating: 1 / 2 / 1 / 2 = 1.5

Comment: Probably intended as a testimony of Joseph Szigeti’s legacy—unfortunately, it was recorded at a time when the artist’s technique and physis showed clear signs of decline. A historic document for people interested in Joseph Szigeti as a great teacher, or in the history of violin teaching. However, it should not have been released to the general public.

J.S. Bach, Sonatas & Partitas — Arthur Grumiaux (CD cover)

Arthur Grumiaux, 1961

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001. Arthur Grumiaux consequently leaves out all second repeats.

I. Grave (3’43”)

In my listening sequence from “slow” to “fast”, this is #21 out of 25—and based on the impression, we have definitely entered the field of the “faster” performances. There is Grumiaux’ beautiful tone, his broad articulation almost throughout, and his permanent, prominent vibrato (luckily, it’s not quite as nervous as others).

My main objection is not with the vibrato, but with Grumiaux’ almost pedantically, obstinately metric performance. This completely blurs the distinction between rhythmic / harmonic foundation and ornament. Moreover, it makes me wonder why the artist claims to play from Bach’s manuscript. And if so, what he read from the manuscript—certainly not the moment, the swaying motion in Bach’s handwriting! About the final bars (see above): plain trills (d” – d”♯) on both crotchets. f’ and f’♯ aren’t even clean.

II. Fuga (7’48”)

Not far from Henryk Szeryng‘s 1967 interpretation, though less obstinate and lighter in the articulation, and slightly faster. Also old school, uniform in the tempo, no agogics to speak of, except for occasional, slight broadening around highlighted notes / chords, compensated by occasional, subtle rushing.

III. Andante (3’30”, second repeat not performed)

Beautiful sonority. Unfortunately, Grumiaux keeps the melody line quasi-legato almost throughout, and filled with vibrato, needless to say. The latter is moderately fast, but fairly (too) prominent. The rallentando at the end of the first part is unnecessary, the one at the end really excessive. The demisemiquaver ornaments are much too metric, almost boring. And one can’t really feel the 3/4 meter.

IV. Allegro (3’58”, second repeat not performed)

Compared to Yehudi Menuhin and Henryk Szeryng, Arthur Grumiaux offers the fastest tempo (close to the average among the 25 recordings) and the smoothest, most harmonious articulation. I would call the latter an “accented portato“. It’s actually détaché, but that is not so conspicuous. There is dynamic differentiation / phrasing. However, there are no agogics to speak of (except for a pronounced ritardando at the end). The artist runs over the score, as if the prime goal was to keep the tempo at all costs, unstoppable. Technically flawless, of course—but where’s the music?

In addition, Grumiaux strangely deviates from the score in a few places, for no obvious reason. And he is the one who claims always to play from Bach’s manuscript! I could agree with variations, if this was in the repeat—but the second repeat is omitted.

Total Duration: 18’59”

Rating: 2 / 2 / 2 / 2 = 2.00

Comment: Grumiaux may sound more harmonious, offering the smoother playing than Henryk Szeryng. However, in the end, both recordings are outdated by now—definitely nothing I would consider passing on as a gift.

J.S. Bach, Sonatas & Partitas — Henryk Szeryng (CD cover)

Henryk Szeryng, 1967

Instrument: 1744 violin “Leduc” by Giuseppe Guarneri del Gesù (1698 – 1744), Cremona.

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001. Henryk Szeryng consequently leaves out all second repeats.

I. Grave (4’26”)

Not surprisingly a traditional interpretation: slow (same tempo as Menuhin, Kremer 1980 and 2002, Khachatryan), and devoid of Klangrede, a tendency towards “flat” dynamics within long notes, except for some “belly notes” and the occasional Nachdrücken. For the phrasing, Szeryng resorts almost exclusively on dynamics, rather than agogics and rhythmic “swaying”. The playing feels “celebrated”, and definitely “old school” / traditional. About the final bars (see above): vibrato on the first crotchet, trill on d’♯.

II. Fuga (8’08”)

Szeryng does differentiate dynamically, but his articulation is very uniform—typically a broad, accented, often resolute portato. And the performance is virtually devoid of agogics (let alone Klangrede): the interpretation appears to move ahead relentlessly, like a steamroller. Old-fashioned. And the demisemiquaver cadenza (bars 286/287) is legato. No surprise here.

III. Andante (5’42”)

Calm pulsating lower voice, slow—maybe a tad too slow for the artist to keep the pace? No, it’s not running away—but there are occasional traces of acceleration, which have a minimal, but noticeable effect on the calm pulsation. The melody voice is virtually legato, but very harmonious overall, with beautiful, gentle dynamic arches. Also the vibrato (especially on the melody) is harmonious and unobtrusive. Romantic, but still the best movement in this sonata performance. One quibble, maybe: the (written out) ornament in bar 25 is way too metric, feels rather clumsy.

IV. Allegro (4’11”, second repeat not performed)

Not too far from Yehudi Menuhin‘s performance in terms of uniformity in tempo and articulation. The main difference is that Szeryng’s articulation is less détaché, closer to portato, i.e., less strict/rigid. The few instances of four successive demisemiquavers in bars 47 and 54/55 are the main thing (“little oddities”?) that disrupt the uniformity of the piece. And the ritardando at the end isn’t as extreme as Menuhin’s. Split time? Rather an Allegro in 8/8 time…

Total Duration: 22’27”

Rating: 2 / 2 / 3 / 2 = 2.25

Comment: It seems that 34 years after Szeryng’s passing (and 55 years after his recording), his interpretation has fallen behind the “reality” in concert halls and recent recordings. Even as a historic recording, I don’t see this as significant, except perhaps for people specifically interested in the artist’s recording legacy.

J.S. Bach, Sonatas & Partitas — Gidon Kremer, 1980 (CD cover)

Gidon Kremer, 1980

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (4’30”)

A little more fluent than Alina Ibragimova, but rather resolute, especially in the f chords. I get the impression of an artist who wants to make sure no single note “falls under the table”. Every note (except for the hemidemisemiquavers) receives attention and emphasis. With this, every motif, every ornament is fully “formulated out”—as melodic element rather than ornament, of course.

Bach writes not dynamic annotation in this movement. So, it is up to the artist to add dynamics, as appropriate. To me, Kremer’s strong sforzandi on key chords and (particularly) phrase beginnings are a little too prominent / conspicuous. This sounds like an interpretation that has grown on the soil of the big, Russian violin tradition. And doesn’t Kremer want to make this sound big, even huge, with his grand musical gestures? Needless to say that the sound is about as far from a performance on period violins as it can get. About the final bars (see above): vibrato on the first crotchet, trill on d’♯.

II. Fuga (7’06”)

Gidon Kremer’s 1980 fugue performance features the same overall duration as Thomas Zehetmair‘s (1982) and Viktoria Mullova‘s (2008). It is far from the latter’s calm, transparency and differentiation, though. And, among the three, it is the one with the least amount of differentiation in agogics and articulation. The latter changes in larger blocks (exposition vs. episodes). Kremer’s technique is of course superb—despite occasional superficialities in intonation, and some excesses in dynamics / expressive playing. 1980, Kremer was still quite far from a HIP performance. Rather, one can sense the influence by his “forefather” David Oistrakh (1908 – 1974).

III. Andante (5’09”)

I like the principal attitude, the pace, the calm pulsation, the phrasing arches, the dynamics. Sadly, that favorable impression is destroyed by the ubiquitous, very nervous vibrato. It isn’t exceedingly strong in amplitude, but all the more irritating through it’s high frequency. How is this supposed to fit the calm nature of the piece?

IV. Allegro (5’00”)

The same, fast tempo as Christian Tetzlaff in 2005, virtuosic, expressive, outgoing, technically excellent. Gidon Kremer is less “technical” in the articulation—more harmonious, more “in a flow”. And despite the fast pace, there aren’t any substantial superficialities. Yet, compared to performances at a more moderate pace, including Kremer’s own from 2002, there is far less differentiation in articulation, dynamics and agogics at the level of motifs. Too much focus on tempo and virtuosity.

Total Duration: 21’45”

Rating: 3 / 3 / 2 / 4 = 3.00

Comment: In 1980, Gidon Kremer’s interpretation was still following the traditions of his teachers, and of the big, traditional performances of the time. Expressive, virtuosic, “typical Kremer”, of course, already back then. I prefer Kremer’s 2002 recording.

J.S. Bach, Sonatas & Partitas — Thomas Zehetmair (CD cover)

Thomas Zehetmair, 1982

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (3’11”)

An interesting approach! Distinctly faster than all others. Initially, that feels amazingly OK, despite all the life in the ornaments. There is some over-articulation, some excess in “micro-articulation”, though. Zehetmair takes a lot of rhythmic liberty in the ornaments. When there are prominent figures in hemidemisemiquavers, such as in bar #5, these turn into sudden, surprising (almost shocking) whirling motions that can hardly fit the Grave annotation. Somewhat inappropriately (given that it’s a slow opening movement in a “church sonata”), the interpretation turns this into a movement full of life, with sparkling, joyful, even jubilant moments. It’s fun to listen to, nevertheless. About the final bars (see above): vibrato on the first crotchet, trill on d’♯.

II. Fuga (7’07”)

Thomas Zehetmair performs at virtually the same pace as Viktoria Mullova 16 years later. However, he can’t match the latter: for one, there is an ever so slight sense of pushing forward. More importantly, the articulation—albeit excellent—is by far not as detailed, diligent and differentiated. Zehetmair (still) uses vibrato. And for strong expression, he does not shy away from occasionally driving the articulation into areas where it starts to sound rough (without feeling irritating, though). It’s certainly never a monotonous, let alone dull performance!

III. Andante (3’16”)

“Conventional” performances in this comparison perform this movement at ♪=48 (Sergey Khachatryan) up to ♪=72 (Viktoria Mullova). I wrote about the revelation that Isabelle Faust‘s performance at around ♪=80 or ♩=40 offered to me. In his 1982 recording, Thomas Zehetmair took this to the extreme, by selecting a tempo of ♪=108 or ♩=54. Taken by itself, the result is quite convincing, though not as compelling as Isabelle Faust’s. In particular, in the context of BWV 1003, the sonata now lacks the “relaxing point” between a musically demanding fugue and the virtuosic, fast final movement.

In “sneak previews”, Zehetmair’s interpretation felt “way too fast” compared to traditional performance. However, after having listened to Isabelle Faust’s “revolutionary approach” extensively, I do concede that Zehetmair’s 1982 recording now feels much more acceptable: I definitely enjoy it!

IV. Allegro (5’01”)

A bit too fast, probably. However, technically excellent, differentiation in articulation, dynamics, agogics, phrasing, even at this fast pace (better than Kremer’s 1980 recording at that!). An occasional, slight tendency to accelerate (runaway tempo?), and occasionally—typical Zehetmair—somewhat on the rough (expressive) side. Slightly overdone, but still excellent. In the group of very fast performances—including also Tomás Cotik (2019), Gidon Kremer (1980), and Christian Tetzlaff (2005)—this is the one I like the most.

Total Duration: 18’35”

Rating: 4 / 4 / 4 / 4 = 4.00

Comment: Technically excellent, very fast (often the fastest), as always, strongly expressive (expression above polished technical perfection)—typical Zehetmair! Very good, but not at the very top within these 25 recordings.

J.S. Bach, Sonatas & Partitas — Jaap Schröder (CD cover)

Jaap Schröder, 1985

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (3’47”)

By now, one can feel that this is the interpretation of one of the pioneers in historically informed performances. I occasionally sense a somewhat demonstrative attitude. Newer HIP performances allow for more metric freedom in the ornaments. That said, there are also phrases with beautiful agogic swaying. And I like the sound of the gut strings, the sonority of the violin, the natural tempo, the articulation, the dynamics. One should value see this as a pioneering HIP performance, a cornerstone recording (along with others), probably. About the final bars (see above): bow vibrato on the first crotchet, trill on d’♯.

II. Fuga (8’39”)

Jaap Schröder’s tempo is almost identical to that of Giuliano Carmignola. However, the articulation of the fugue theme is much more mellow, more pleasant—and that holds true throughout the fugue. And Schröder is much better at integrating (soft) arpeggiated chords into the musical flow. The flow in general is much more harmonious. The downside is that very often the intonation is fairly marginal—certainly far from perfection. Schröder performs demisemiquaver cadenza (bars 286/287) legato, which I believe is against Bach’s notation.

III. Andante (5’10”)

Excellent flow, keeping the focus on the 3/4 meter. Traces of occasional unrest, maybe—but that is inherent in this piece with its quaver-dominated texture. The occasionally somewhat “spicy” sound of gut strings and baroque bow are fitting well into the interpretation. However, in that context, the occasional vibrato isn’t really needed, it is bordering on “too much”. There are a few extra ornaments in repeats, which I think are well-placed and fit the character of the piece.

IV. Allegro (5’44”)

Some rare instances if marginal intonation, occasional superficialities in the articulation, especially in demisemiquavers. Good, but not top.

Total Duration: 23’21”

Rating: 4 / 3 / 4 / 4 = 3.75

Comment: Still a good, historically informed performance, excellent, “baroque” sonority. In terms of intonation and cleanness in articulation, though, this performance has (in my opinion) been superseded by more recent HIP recordings.

J.S. Bach, Sonatas & Partitas — Vito Paternoster, cello (CD cover)

Vito Paternoster, 1995 (Cello, D minor)

Instrument: 1792 cello by Lorenzo Carcassi (1737 – 1775), Florence. Pitch: a’ = 443 Hz

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (4’04”)

Vito Paternoster’s tempo is identical to Amandine Beyer‘s—definitely a good tempo. Contrary to the latter, though, Paternoster, uses a vast excess of Klangrede, overloading all of the ornamentation with articulation, expressive agogics and dynamics. It’s all beautifully done, really, and Paternoster’s technical potential is astounding! I just don’t think it’s the music that Bach wrote? The listener gets some glimpses at the “backbone”, the harmonic progression—but overall, this rather feels like a very slow dance movement. About the final bars (see above): plain trills (g – g♯, in the original: d” – d”♯) on both crotchets, combined with bow vibrato on B (second crotchet, f’♯ on the violin).

II. Fuga (7’56”)

Performing this fugue on a cello is a huge challenge. For one, there is the technical side—which Vito Paternoster masters very well. His interpretation often sounds playful, light-weight, with added trills every now and then. The articulation is light and certainly historically informed in the best sense of the term, and even complex polyphonic chord sequences never sound forceful. Still, one can hear the challenge in limitations in the intonation purity in complex chord sequences. No, I would not call it “completely off”. Most of this may have to do with the string’s slow response to the bow, which does not leave enough time to deploy its full sonority and clarity of the pitch.

The other challenge is inherent with the transfer to the dark sonority of the cello, maybe in particular with the extra full-bodied sound of Vito Paternoster’s 1792 Carcassi—and despite the modern pitch of a’ = 443 Hz. The sonority as such is not the problem. However, relative to strong, singing quality of the d and a strings, the C and G strings mainly sound “deep” and “dark”, often lacking tonal clarity. It is hard to keep track of polyphonic voices at the bottom of the instrument’s range.

The demisemiquaver cadenza (bars 286/287) is legato—though here, this is understandable, given the heavier sonority of the instrument. To be fair: this could also be a change that the 17th century transcriber added. Remarkably, Vito Paternoster’s overall timing is the same as Gidon Kremer‘s (2002) and Isabelle Faust‘s (2012)!

III. Andante (4’43”)

Most of what I stated about Atilla Aldemir’s 2019 viola performance also applies to this performance on the cello. The two recordings share a very similar tempo / overall timing. Differences: Vito Paternoster performs the melody with less legato. This enhances the transparency. On the other hand, this is also needed, as the occasional outreach onto the C and G is at a disadvantage in terms of tonal clarity and definition. Compared to the viola, the intonation purity is a tad inferior (but certainly never seriously off). And Vito Paternoster’s occasional extra ornament compensates for the “bulkier” nature of the cello sonority.

IV. Allegro (5’30”)

Vito Paternoster performed at the same tempo as Isabelle Faust some 17 years later! Given that he is performing on a cello, this alone is a strong achievement. However, obviously, a cello is not a violin. As agile as Paternoster’s playing may be, many of the fast demisemiquaver figures sound somewhat superficial, and the intonation isn’t always as good as it should be. The characteristics of the instrument (for the most part, the slower response of the strings) don’t permit being as detailed in articulation (Klangrede) and dynamics as the top HIP performances on a violin. A slightly slower, more adapted pace would help making this substantially better.

Total Duration: 22’13”

Rating: 3 / 3 / 3 / 3 = 3.00

Comment: An interesting experiment and astounding achievement. Even though it can’t quite compete with a HIP performance on the “proper” instrument. However, if you are looking for peculiar performances / configurations in Bach’s “Sei Solo“, why not give it a try?

J.S. Bach, Sonatas & Partitas — Monica Huggett (CD cover)

Monica Huggett, 1997

Instrument: 1618 violin by Antonius & Hieronymus Amati, i.e., Antonio Amati (1540 – 1607) and Girolamo Amati (1561 – 1730), Cremona. Pitch: a’ = 415 Hz

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (3’37”)

The tempo is fluent (if not very fluent, #23 out of 25). And indeed, there are ornaments that flow with a beautiful, swaying movement—true ornaments, indeed! That unfortunately is more than offset by the fact that Monica Huggett plays in miniscule phrases, hereby breaking the overall flow with prominent détaché gaps, leaving the impression of a fragmented piece. There is also an irritating tendency towards occasional highlights through sudden “belly notes”. Too “demonstratively historic”? Not really happy with this. About the final bars (see above): plain trills (d” – d”♯) on both crotchets; both crotchets are distinctly détaché.

II. Fuga (9’35”)

A disappointment, throughout. Yes, there is the sound, the sonority of a period violin—but that’s about it. I don’t claim the fugue is easy to play (it definitely is a big challenge!). However, I don’t need an interpretation that tells me how difficult it is to play.

For one, the intonation is often marginal, and then, the articulation is often limited to a rough staccato, with the exception of the semiquaver passages, maybe. Then, whenever the polyphony involves triple- and quadruple-stop chords, the artist disrupts the flow (and it’s the slowest among the performances already!). She is taking extra time for a broad arpeggio with notorious belly dynamics on the final note/interval. It’s like a statement “That’s difficult! The violin isn’t made for polyphony!”. Overall, it all distracts from the music, and the performance doesn’t “breathe”.

III. Andante (5’05”)

The artist’s persistent “belly accents” on every single quaver soon feels (almost) obnoxious, unfortunately. With this, the 3/4 meter is lost entirely. And there are occasional instabilities in the pulsation. Overall, there is an unrest which—to me—does not fit into a calm Andante movement.

IV. Allegro (6’56”)

At first, the tempo (the slowest among the 25) seems appropriate for an Allegro. However, for one, Bach writes groups of four beamed semiquavers—four per bar, and the performance really is barely Allegro, at ♩=66. Moreover, the movement is in split time—so, strictly speaking, the pace is (half note)=33. Tempo aside: Monica Huggett’s performance features numerous irregularities in the pace—I don’t see why p echo instances are very often a tad faster than the initial f instance. The differences are enough to be slightly irritating (to the careful listener, at least).


The above, however, are minor issues. My main objection is with the ornamentation. Monica Huggett leaves the first passes alone, which is fine. In the repeats, she adds a large number of extra ornaments. To some degree, ornaments are a matter of the artist’s personality and personal taste. And in a live performance, they may even be improvised. In a recording, however, they really should fit (and serve) the music. Here, I have several objections:

  • Some of the arguments are fairly peculiar, if not occasionally odd (are they just meant to be different from anybody else?).
  • There are also ornaments that alter the melodic and rhythmic structure of the composition—to me, this is a strict no-go. They should be well-adapted in style. And they should fit into the flow.
  • In my opinion, ornaments should be inconspicuous, or at least enrich the listener’s experience.
  • There are instances of ornaments that are exact replicas of previous instances. For one, this immediately sounds odd (lack of fantasy? lack of imagination?) and distracting. It may also cause the listener to lose interest. Lastly, it may look like the composer has written these ornaments—which I think is pretentious (especially if the quality of such ornaments is questionable, as here).

Total Duration: 25’13”

Rating: 2 / 1 / 2 / 2 = 1.75

Comment: The idea here is that of a historically informed performance. It is so in terms of sonority and instrumentation. That’s about all the positive arguments that I can think of. I can’t recommend this recording.

J.S. Bach, Sonatas & Partitas — Rachel Podger (CD cover)

Rachel Podger, 1999

Instrument: 1739 “re-baroqued” violin by Pesarinius, Genoa; baroque bow. Pitch: a’ = 415 Hz

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (3’56”)

Hard to believe that Mikhail Pochekin’s and Rachel Podger’s recordings feature identical overall timing; the impression is vastly different. That’s not really because of Podger’s period violin and bow (I like the sonority, the sound of the gut strings, the sparing use of vibrato). The key differences are in the artist’s tendency for fairly metric ornaments. This defeats some of the “ornament impression” by giving the ornaments extra weight. Inappropriately, I think.

More importantly, Rachel Podger shows a tendency to perform all longer notes (say, quavers and above) and every motif with a crescendo. It’s not Nachdrücken, and it’s typically not “belly notes” (that occasionally happens as well, though), but merely a soft beginning followed by a crescendo across the note—prominent, overblown, and soon feeling (close to) obnoxious. About the final bars (see above): vibrato on the first crotchet, trill on d’♯. The two crotchets are linked via glissando.

II. Fuga (8’36”)

Also here, there are some idiosyncrasies, such as a tendency towards crescendo on slurred quaver pairs, or on some “big arpeggio chords”. And not all arpeggi are harmoniously fitting into the flow. Also, to me, the very mellow (legato) articulation in the comes is a bit too much of a contrast to the fugue theme itself. And as in the first movement, one can again feel a tendency towards crescendo articulation. It’s much less conspicuous, though. Nevertheless, these idiosyncrasies are far less conspicuous than in the Grave, and they have much less of an impact on the overall impression. The same holds true about the intonation, which occasionally is very slightly marginal.

Just like Giuliano Carmignola, and at virtually the same pace, Rachel Podger performs the demisemiquaver cadenza (bars 286/287) détaché—though here, it is much lighter, more natural, and not noisy at all.

III. Andante (4’54”)

In coming from the slowe(er) performances, this is #19 for this movement. The first one below 5 minutes duration, the first one that feels distinctly more fluent. And it’s one of the first that make the listener feel the 3/4 meter—without creating unrest. And the Andante impression remains in quaver accompaniment, of course. There is some of the note swelling that I noted in the first movement, and also a rare Nachdrücken—but by far not as conspicuous as in the Grave.

IV. Allegro (6’13”)

Virtually the same (moderate) tempo as Giuliano Carmignola. Rachel Podger’s performance is slightly smoother, overall, with somewhat less Klangrede, less focus on the articulation of small motifs. This makes it appear somewhat more fluent. There are a few, subtle superficialities in the articulation of demisemiquaver motifs (but that’s really a minor, negligible issue).

Total Duration: 23’39”

Rating: 3 / 4 / 4 / 4 = 3.75

Comment: A good, solid HIP performance—with some idiosyncrasies in slow movements.

J.S. Bach, Sonatas & Partitas — Gidon Kremer, 2001 (CD cover)

Gidon Kremer, 2002

Instrument: 1730 violin “ex-David” by Giuseppe Guarneri del Gesù (1698 – 1744), Cremona.

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (4’28”)

22 years after his 1980 recording, Kremer’s tempo hasn’t changed (still on the slow, traditional side). There is still a certain scent of “big” from his earlier interpretation. However, the focus is now more on the crotchet “backbone”. Small notes have much more of an “ornament feel”. The vibrato is much less conspicuous, often even absent. Kremer does not shy away from the raw sound of empty strings. Clearly, the HIP trend has had its effect on the artist! About the final bars (see above): oddly, plain trills (d” – d”♯) on both crotchets (note that in 1980, the first crotchet was done with vibrato).

II. Fuga (7’56”)

In his 2002 interpretation, Gidon Kremer starts of with a fairly “robust” détaché articulation—much more robust (with far less elasticity) than newer HIP performances, such as Amandine Beyer (2011), Christine Busch (2012), Isabelle Faust (2012), Alina Ibragimova (2009). However, he is of course miles apart from people such as Henryk Szeryng (1967), and in comparison to traditional interpretations, Kremer’s vibrato is virtually absent. In softer segments, the articulation turns gentler, more mellow.

It’s primarily in polyphonic fugue parts, with its frequent arpeggiated chords where the interpretation seems to remind of the artist’s prevenance from the great “Russian School” of violinists—his last teacher was David Oistrakh (1908 – 1974). It’s an interpretation full of contrasts: robust vs. mellow / gentle, broad strokes vs. light détaché in the extended semiquaver segments, or in the demisemiquaver cadenza (bars 286/287).

III. Andante (5’33”)

In his second recording, Gidon Kremer steps away from perfection in the sound—he leaves the tone slightly rough, “grainy”. And he keeps the absolute calm throughout the piece. Every quaver in the lower voice receives a little accent—not percussive, but gentle, and not delayed too much into the note. It doesn’t really feel like a “belly note”, but is noticeable. The downside is that he does this with every quaver in the lower note. This not only feels a tad uniform, if not monotonous, but it also shapes each associated quaver and semiquaver pair in the melody voice. With this, there is too much focus on the quaver structure (“6/8”), and the 3/4 meter disappears almost completely.

However, overall, I like the artist’s attitude which obviously does not mean to produce perfect beauty, but rather wants to serve the music, the composer.

IV. Allegro (5’16”)

Gidon Kremer uses a “percussive détaché” at a very fluent pace. It does not feel excessively fast, not rushed, but it leaves the listener somewhat breathless. And it is at the point where many demisemiquavers are losing sonority and definition. Some even sound superficial, occasionally noisy, and some “low-mid-high-mid” semiquaver figures across three strings are slurred, sound like arpeggiando. Sound esthetics is not Gidon Kremer’s goal—rather: expression, at all costs.

Total Duration: 23’13”

Rating: 4 / 4 / 4 / 4 = 4.00

Comment: If you are looking for a performance that puts expression and liveliness in sound above cleanness in articulation and perfect sonority, then this is your recording! Recommended, with some reservations.

J.S. Bach, Sonatas & Partitas — Christian Tetzlaff (CD cover)

Christian Tetzlaff, 2005

Instrument: 2002 violin by Stefan-Peter Greiner (*1966), Bonn.

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (3’52”)

After Thomas Pietsch‘s performance (same duration, almost), what a relief! Gone is the excess in noise, the demonstrative attitude. Christian Tetzlaff steps back behind the composer, letting the music “work”. His approach is devoid of excesses in dynamics, agogics and articulation are natural.

The ornaments are performed with sufficient rhythmic freedom to make it clear what is ornament, and what isn’t—the perfect balance. The ornaments never prevail, letting the listener follow the calm, basic harmonic progression. This is about as close to “historically informed” as a performance on modern instrument and bow can get. About the final bars (see above): bow vibrato on the first crotchet, trill on d’♯.

II. Fuga (7’24”)

Basically a good, fluent tempo. However, unlike Viktoria Mullova (2008) who makes an even faster pace feel relaxed and natural, Christian Tetzlaff’s 2005 interpretation gives an almost constant sense of pushing, of restlessness. Too bad, as otherwise I like the general approach: dynamics, articulation, agogics. With this, it feels as if I didn’t have the time to listen through the polyphony. And: a little less vibrato at climaxes would not hurt! The density & intensity at the climax in bars 157ff is maybe a little overdone.

III. Andante (5’01”)

Warm, mellow, expressive. The vibrato is fairly ubiquitous, though fairly moderate. I’m only irritated by it where it reinforces expressive, dynamic excursions. In these moments, the performance feels a little too romantic. I prefer interpretations that leave the simplicity of the melody as is. On the bright side, I really like the calm pace on the quavers, which still allows for the presence of the 3/4 meter in the melody.

IV. Allegro (4’56”)

Technically excellent playing, clear, precise, despite the very fast tempo. Careful, expressive dynamics, differentiation in the articulation between f segments and p echoes. To me, the tempo is a little too fast. It is controlled (some slight accelerations for little good reason), but rather Presto than Allegro, and somewhat breathless. The main drawback of the fast pace is that it leaves little or no time for agogics, for detailed articulation within motifs, for “local Klangrede”.

Total Duration: 21’14”

Rating: 4 / 4 / 4 / 4 = 4.00

Comment: I would characterize this as a modern interpretation “with HIP ambitions”, technically excellent, maybe occasionally too much focus on technical excellence? A very good performance, no doubt. I prefer a “proper HIP” performance, though.

J.S. Bach, Sonatas & Partitas — Viktoria Mullova (CD cover)

Viktoria Mullova, 2008

Instrument: 1750 violin by Giovanni Battista Guadagnini (1711 – 1786), Milan; gut strings; baroque bow by W. Barbiero. Pitch: a’ = 415 Hz

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (3’34”)

Only very few of the 25 performances in this comparison leave me completely happy. Not surprisingly, this is one of these few! Despite a very fluent tempo (the fastest, apart from Thomas Zehetmair), Viktoria Mullova is able to keep her calm. Nothing feels rushed, one can feel the Grave foundation, the overall, solemn rhythmic and harmonic progression. And, of course, ornaments are—ornaments, not melody. They form wonderful, flowing, swaying arches, picking up momentum, relaxing again. Such natural breathing, beautiful phrasing arches. Vibrato is used very rarely—and only as (kind of) ornament. My favorite performance, no doubt! About the final bars (see above): bow vibrato on the first crotchet, trill on d’♯.

II. Fuga (7’08”)

Despite a rather fluent pace, nothing in Viktoria Mullova’s interpretation feels pushed or rushed. Quite to the contrary! Among all the interpretations, this is the one with the most detail and richness in “speaking” articulation within the fugue theme. And the artist maintains this throughout the fugue. Within the theme, that articulation ranges from a staccato to portato. But despite the latter (and the fluent pace), the performance never loses transparency. One way the artist achieves this is by keeping accompanying voices in a very short, but gentle staccato. However, this always feels controlled, differentiated, never rough, crude, or gross.

That said, Viktoria Mullova never sacrifices expression for perfection or the beauty of sound. The sonority of the Guadagnini violin is marvelous. Another, true landmark performance, indeed, hard to beat!

III. Andante (4’30”)

This performance and the one by Tomás Cotik share the same overall timing. But, as mentioned there, the two recordings could hardly be more different. Where Cotik’s tone is rough and grainy, Mullova’s performance sounds so warm, harmonious, rounded,, mellow—without sacrificing the character of a historically informed performance.

Viktoria Mullova manages to keep the quaver line pulsating, while at the same time keeping the melody voice clean, singing, with a natural legato—and devoid of unnecessary pulsation. She obviously doesn’t do the pulsation by modulating the speed of the bow, but by periodically moving the bow to and off the lower string, leaving the melody line alone. This makes such a huge difference! It’s really the highest art of violin playing. Artful, but the simplicity and harmony of the result makes the listener believe that this is simple, if not easy!

IV. Allegro (5’45”)

Technically absolutely superb, very clear, as expected—though without aiming for perfection in articulation. Flawless in the intonation. Slightly more fluent than Giuliano Carmignola, Rachel Podger, Christine Busch, and Amandine Beyer. Viktoria Mullova sees no need for unnecessary extras, such as additional ornaments, or extravagances in articulation. Simply Bach’s text, as written. A negligible exception is in rare instances of light staccato semiquavers: these serve to de-emphasize notes in weak parts of a phrase. Harmonious, simple, beautiful playing without show effects—more is not needed: it’s all there, in Bach’s music!

Total Duration: 20’58”

Rating: 5 / 5 / 5 / 5 = 5.00

Comment: Along with very few others, my favorite recording, for sure: highly recommended!

J.S. Bach, Sonatas & Partitas — Kristóf Baráti (CD cover)

Kristóf Baráti, 2009

Instrument: 1703 violin “Lady Harmsworth” by Antonio Stradivari (1644 – 1737), Cremona.

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001. Kristóf Baráti consequently leaves out all second repeats.

I. Grave (3’50”)

Kristóf Baráti, as I have encountered him in concerts. Big sound, perfect execution, intonation and sonority. But also largely a conventional interpretation, with vibrato essentially everywhere. The vibrating is not irritatingly fast, nor excessive in amplitude. Just conventional. The interpretation doesn’t just “talk big sound”, but it also seems to say “this is big music“.

As for the ornament / backbone balance: some of the ornaments appear as such, with natural, fluent / flowing agogics. Others, though, appear rather metric, seem to say “I’m melody, too!”. The overall impression is somewhat ambivalent. About the final bars (see above): plain trills (d” – d”♯, rather fast and nervous) on both crotchets.

II. Fuga (7’42”)

Perfect sound and articulation, straight portato, big tone. Rather expansive dynamics, focus on instrumental perfection. I sense a concentration on one voice at a time, i.e., it seems more important to the artist to present a perfectly shaped and articulated melody line, than to let the listener experience the complexity of the polyphonic texture. My main criticism is about the lack of “local agogics” (a.k.a. Klangrede). The relentlessness and stubbornness in Henryk Szeryng‘s 1967 performance comes to mind.

III. Andante (5’18”)

Carefully played, melodious, calm (with a trace of unrest, though). Fairly metric in the demisemiquaver ornaments. On the other hand, it is amazing how the artist manages to “shield off” the quaver accompaniment from the long notes in the melody. To me, the one, major downside in this interpretation is in the constant (harmonious, but always quite dominant) vibrato throughout the piece. I find this rather distracting. The music doesn’t need it, nor is there a need for the artist’s few extra ornaments.

IV. Allegro (4’05”, second repeat not performed)

Perfect tone, perfect articulation—but sadly, always the same: virtually the same détaché strokes throughout, no agogics to speak of, little in dynamics (apart from the p and f annotations in the score, and highlighting the peak note in phrases). Maybe perfect, but uninteresting. Why no repeat in the second part, and why the doubling of the last note (e”) in the first pass of first part?

Total Duration: 20’55”

Rating: 3 / 3 / 3 / 2 = 2.75

Comment: It may be perfect in the execution—however, in comparison to today’s HIP performances, it feels pale, if not boring.

J.S. Bach, Sonatas & Partitas — Alina Ibragimova (CD cover)

Alina Ibragimova, 2009

Instrument: 1738 violin by Pietro Guarneri (1695 – 1762), Venice.

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (4’35”)

Alina Ibragimova is a tad faster (i.e., less slow) than Augustin Hadelich, but still very (too) much on the slow side. I like the fact that the artist plays without vibrato. And Indeed, some of the fast notes are feeling like ornaments (unlike with Hadelich), though the base “melody” remains elusive, mostly. Oddly, the artist shows a prominent tendency to perform strong (excessive) crescendo on long notes. With very few exceptions, it’s not Nachdrücken, but probably an attempt to “let the note grow/evolve”. To me, this “feature” is far too prominent / conspicuous, and neither natural, nor relaxed and harmonious. About the final bars (see above): bow vibrato on the first crotchet, trill on d’♯.

II. Fuga (7’47”)

A tad faster than Isabelle Faust‘s 2012 approach, but also keeping the calm through the agogics and the slightly more expansive dynamics. Alina Ibragimova does leave a little extra space for 3- and 4-string arpeggiated chords—though that does not disrupt the overall flow. She also apply slightly more expressive agogics, where appropriate—her interpretation is a tad more expressive overall. At the same time, Alina Ibragimova’s dynamics never exceed a simple f. Very rarely, the performance reminds if the swelling on long(er) note values—this was far more prominent in the Grave, though.

III. Andante (5’34”)

Calm, with gentle rhythmic swaying and harmonious, broad phrasing arches. I like the character of the interpretation, the pace, the simplicity of the music, the consequent absence of vibrato. And I love how in the end the piece subtly retracts into ppp: heavenly!

Unfortunately, as with the Grave movement, there is this somewhat unfortunate tendency towards applying crescendo to big(ger) note values, sometimes bordering on Nachdrücken.

IV. Allegro (5’35”)

Careful, diligent articulation, dynamics, and phrasing in general. However, Alina Ibragimova’s tempo is at a point where the articulation of some of the demisemiquavers is about to suffer, i.e., start to sound hasted, lacking sonority, almost a bit superficial. Good, but not top (one should consider that this recording is already 13 years old).

Total Duration: 23’31”

Rating: 3 / 4 / 4 / 4 = 3.75

Comment: Very good, historically informed performance, recommended.

J.S. Bach, Sonatas & Partitas — Sergey Khachatryan (CD cover)

Sergey Khachatryan, 2009

Instrument: 1702 violinLord Newlands” by Antonio Stradivari (1644 – 1737), Cremona.

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (4’29”)

Virtually identical in tempo to Kremer (1980), Menuhin, Kremer (2002), and Szeryng (see there for remarks concerning ornament vs. melody). As in most of Khachatryan’s movements (especially the slow ones), there is this exceedingly nervous, prominent vibrato, which really hurts. Particularly with the artist’s irritating tendency towards “belly notes” and Nachdrücken. Not much joy. About the final bars (see above): nervous vibrato on the first crotchet, trill on d’♯.

II. Fuga (8’20”)

Without doubt, Sergey Khachatryan is an artist with impressive technical skills and potential. However: needless to say that I have a strong aversity against the artist’s use of strong, very nervous vibrato. Particularly, whenever he wants the music to be expressive, intense. Especially with the vibrato, the interpretation occasionally sounds overblown.

With very few exceptions (where he wanted to achieve an echo effect by repeating f segments with p), Bach did not add dynamic annotations. Therefore, I don’t see why the artist plays the fugue theme always p when it is alone. Then, he (always) switches to f when other voices join in. The addition of the extra voices already increases the volume (inherent dynamics).

III. Andante (6’40”)

An attempt to make this movement sound tender, sweet—too much, too romantic, too slow. And these soft, trembling belly notes—that’s all excessive, and has nothing to do with baroque music at all! Besides: the movement is in 3/4 time, and Andante—this is around ♪=50, i.e., Andante in 6/8 time, at best! Yes, the pulsating on the lower (base) line is very calm, serene (though still a sequence of trembling belly notes). However, nothing in the notation indicates that the melody voice ought to be (near-)legato. And: how does the constant, nervous vibrato trembling fit into all this?

IV. Allegro (5’48”)

In essence, see my comments on Kristóf Baráti‘s performance. However, I concede that here, I even prefer Khachatryan’s over Baráti’s: there is more differentiation in dynamics and articulation, and clearly more agogics, far less mechanical playing. I would say that this is even one of the better non-HIP recordings.

Total Duration: 25’16”

Rating: 2 / 3 / 1 / 3 = 2.25

Comment: The excess vibrato in Khachatryan’s playing (in my view) often makes this recording hard to listen to. At the very least, in today’s HIP environment, this is a debatable, if not highly controversial recording. If I were a violinist with this much vibrato (that I don’t want to suppress), I would strongly hesitate publishing Bach recordings. Or baroque recordings in general. At the very least, I can say: sorry—not my taste!

J.S. Bach, Sonatas & Partitas — Amandine Beyer (CD cover)

Amandine Beyer, 2011

Instrument: 1996 baroque violin by Pierre Jaquier; 2000 baroque bow by Eduardo Gorr. Pitch: a’ = 415 Hz

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (4’04”)

A calm, contemplative base pace that permits following the harmonic progression. At the same time, the tempo allows for careful shaping of all ornaments and figures. Yet, this does not distract from the big structure. One might call this an ideal tempo that is in line with the Grave (heavy, serious) annotation. Beautiful phrasing, harmonious, well-balanced, peaceful, serene—superb! About the final bars (see above): just plain trills (d” – d”♯) on both crotchets—against my sincere expectations! The one little hair in the soup?

II. Fuga (8’28”)

Really amazing! Other artists make the fugue sound so difficult, so enormously challenging. In Amandine Beyer’s hands, it sounds absolutely natural, harmonious and effortless! There is that much differentiation in articulation, phrasing and dynamics, the arpeggiated chords are always completely integrated into the musical flow. There is Klangrede everywhere, from the fugue theme to the episodes, to the complex chord sequences, to the cadenza (performed détaché, of course), on to the subtle, vibrato-less decrescendo on the final tone. Exemplary—a benchmark performance in all aspects.

III. Andante (5’43”)

A remainder of a romantic view? Barely. Still, Amandine Beyer made the (maybe controversial?) decision to apply the Andante pace to the pulsating quavers (ca. ♪=56), not to the 3/4 meter. She isn’t alone in this, I should say, so I won’t rate this. The key point is: she just about manages to make the listener still feel the 3/4 meter. And the melody does not feel exceedingly slow, but turns into a heavenly serene reflection, full of harmony, warmth and inner beauty. All unexcited, flexible in the agogics, mostly quasi-legato singing, free in the rhythmic shaping of Bach’s ornaments (e.g., bar 25): a dream, really!

IV. Allegro (6’00”)

I love every detail in this interpretation! Amandine Beyer combines the best aspects of Giuliano Carmignola‘s and Christine Busch‘s interpretations: richness and detail in the articulation, down to small motifs, keeping the ₵ in mind and perceptible for the listener, clarity in every detail, agogics, an enthralling flow. The latter without rushing, but compelling throughout (her tempo is slightly above Carmignola’s, Podger’s and Busch’s. Moreover, I really like Amandine Beyer’s “percussive” articulation, be it in portato, or in her light “gut string staccato“: simply excellent, as good as it can get!

Total Duration: 24’15”

Rating: 5 / 5 / 5 / 5 = 5.00

Comment: Clearly among my very top favorites—strongly recommended!

J.S. Bach, Sonatas & Partitas — Thomas Pietsch (CD cover)

Thomas Pietsch, 2011

Instrument: 1672 violin by Hannß Khögl (1614 – 1680), Vienna; bow by Pierre Patigny (after an anonymous bow, early 18th century). Pitch: a’ = 415 Hz

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (3’53”)

I don’t think I want to hear how every finger touches the string, all the (not always clean) transitions between notes, the occasional buzzing of strings. The overall impression is that of a noisy performance. At times, also the intonation isn’t perfect. The performance of the ornaments is fairly metric, sometimes too metric. The articulation is too demonstrative, almost pedantic in the observation of Bach’s slurs. It lacks the notion of ornament, and the harmonious agogic swaying that Bach’s handwriting appears to imply. About the final bars (see above): slightly irregular bow vibrato on the first crotchet, trill with mordent on d’♯.

II. Fuga (9’06”)

Within this comparison, this is almost the slowest performance. That certainly helps the clarity, and it may also be a concession towards the church acoustics. However, at least in the context of the comparison, it makes the interpretation sound fairly static, especially in combination with the resolute staccato-like articulation. The latter also makes the interpretation sound a bit too didactic, if not sometimes schematic.

The beautiful sonority of the violin and its interaction with the reverberating acoustics are offset by subtle disturbances in the intonation, and by occasional, slight superficialities in the articulation. This and the relatively static performance may make it hard(er) to some people to “listen through” the natural roughness of the gut strings, the sound of the interaction between bow hair / rosin and the string. A “more distant” recording (one that focuses on the sound that projects farther into the audience) would help inexperienced listeners to ignore minor deficiencies, or what may sound like one.

Besides my critical remarks: Thomas Pietsch is absolutely right when he describes his violin as being particularly well-suited to polyphonic pieces, such as Bach’s fugues. The low pitch (a’ = 415 Hz) and the warm sound of the gut string are a great help in balancing the sound across the range of the violin.

III. Andante (5’10”)

This may be the movement where the reverberation hurts the least—it actually harmoniously fits into atmosphere of the piece. Similarly, I like the extra ornaments that the artist adds to the repeats. There are some downsides, still. The absence of vibrato (which I really, really welcome) amplifies the slightest issues with intonation, like a magnifying glass. And I don’t think this is “just” a consequence (downside) of Pythagorean intervals. Moreover, the proximity of the microphones reveals the slightest imperfections in the articulation, the transitions between notes, etc.

IV. Allegro (5’48”)

I like the general concept in terms of dynamics, agogics. Strangely, there aren’t just occasional, slight superficialities in demisemiquavers. Or, at least, the close microphone setting makes these demisemiquavers sound noisy.

In this piece, there are many split crotchets (4 x semiquaver) where the second semiquaver is split into two demisemiquavers. Very often, Bach puts a slur on all but the last note (there are also instances without slur). As far as I can tell, all performances follow this notation. Inspection of the manuscript shows that there are a few instances where Bach’s slur is a tiny bit longer than needed—but it does not extend over the last semiquaver. Oddly, Thomas Pietsch often, but not always (and not systematically), extends Bach’s slur over the last note. I don’t understand why.

Total Duration: 23’57”

Rating: 2 / 2 / 3 / 3 = 2.50

Comment: I hesitate recommending this recording—mainly because of the acoustic setting, i.e., the excess in reverberation. But also in terms of articulation clarity / cleanliness, the performance can’t compete with most other historically informed recordings.

J.S. Bach, Sonatas & Partitas — Christine Busch (CD cover)

Christine Busch, 2012

Instrument: 18th century baroque violin, Tyrol. Pitch: a’ = 415 Hz

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (4’07”)

Christine Busch’s tempo is certainly not too fast, and she makes a clear distinction between ornaments and the piece’s “framework”. However, similar to Giuliano Carmignola, she tends to be very articulate in the small motifs and figures. This distracts from the big structure, causing an excess focus on details. One can say the same about the artist’s very conspicuous (sometimes even irritating) tendency for belly notes and prominent swelling on longer note values. Hence, the listener easily loses track of the bigger structures and phrases. About the final bars (see above): bow vibrato on the first crotchet, trill on d’♯.

II. Fuga (8’27”)

Same tempo/timing as Amandine Beyer, but with fine differences: not quite as effortless and natural, slightly more expressive / expansive in the dynamics. I noted an occasional, very subtle tendency to accelerate towards complex polyphonic parts. However, in other places, the artist broadens the tempo around complex, polyphonic culmination points. With the artist’s very slight tendency to over-articulate, the movement appears somewhat less coherent, maybe a tad fragmented relative to Amandine Beyer’s interpretation (which I slightly prefer over this one).

III. Andante (5’12”)

Really calm and harmonious in the pace. Christine Busch is able to maintain a constant flow, without significant moments of unrest. At the same time, she manages to keep a focus on the 3/4 meter. The artist lets the melody voice prevail, keeping the quaver line of the second voice as accompaniment. The main “hair in the soup” here (as already in the Grave) is in a noticeable tendency towards conspicuous belly notes.

IV. Allegro (6’07”)

Very similar in pace and characteristics to Giuliano Carmignola, Rachel Podger, and Amandine Beyer. An excellent, very careful performance, with persistent attention to articulation down to smallest figures and motifs. There are no superficialities throughout this movement. One interesting detail: Christine Busch not only plays the p echoes softer, but also with lighter (gentle staccato) articulation. In general, along with the dynamics, she varies the articulation from portato (f) to staccato for softer parts, also within a phrase. Christine Busch also applies excellent agogic swaying.

Total Duration: 25’53”

Rating: 4 / 5 / 4 / 5 = 4.50

Comment: Among the historically informed recordings, this is an excellent one—among the very best; certainly worth a strong recommendation.

Isabelle Faust, Bach Sonatas & Partitas, vol.2 (CD cover)

Isabelle Faust, 2012

Instrument: 1704 violinLa belle au bois dormant” by Antonio Stradivari (1644 – 1737), Cremona. This is apparently the only Stradivari violin that has not undergone modernization, i.e., it still features the shorter and flatter neck.

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (4’10”)

From the timing, just a tad faster than Giuliano Carmignola. However, the impression is a much more fluent one, because Isabelle Faust puts far less emphasis on articulation and (excess) expression in the ornaments. With this, the movement appears much less fragmented, but the listener can focus on the “big picture”, the “backbone”, with a harmonious “breathing” in half and full bars. Ornaments are what they are, not melody. And there is Isabelle Faust’s calm, reflected approach, her beautiful tone, the clear intonation, virtually devoid of vibrato. And no, there isn’t a single moment where I miss the vibrating! About the final bars (see above): subtle bow vibrato on the first crotchet, trill on d’♯.

II. Fuga (7’55”)

It’s amazing to realize how effective Isabelle Faust’s clear, simple approach is, even without “gut string esthetics”! Never does she enforce the tone, her articulation is typically a simple, natural portato. The artist carefully shapes every melody voice, also in polyphonic segments. Her arpeggiated chords are simple, natural, never aggressive, almost inconspicuous, and she doesn’t let them disrupt the flow, the agogics. Never does the interpretation appear to say “Look how big, complex, and difficult a fugue that is!”.

I don’t mean to say that Isabelle Faust’s view is the only valid and viable way to perform this fugue. However, I believe this is maybe the best at allowing the listener to follow all voices throughout the piece, at all times, and without distraction.

III. Andante (4’07”)

I’m quite familiar with Isabelle Faust’s performances, and I have heard her perform the entire “Sei Solo” in concert. So, I expected a top rating here, as in most of her recordings and performances. I listened to this movement right after Viktoria Mullova‘s interpretation. The latter is the third-“fastest” among the 25 recordings of this movement, Isabelle Faust is the second-fastest, beating Mullova’s timing by 23 seconds (see Thomas Zehetmair for the “speed contest winner”). Hearing Faust after Mullova, made me stunned, and I hesitated about the rating. I realized that I first needed to gain some distance from 23 of the 25 recordings, in order really to appreciate Isabelle Faust’s performance. And indeed, after a night’s sleep, things looked different!

Stepping Back for a Fresh Approach

I’m sure that’s what Isabelle Faust did: she is an artist who does serious, thorough research. She doesn’t take existing performances and traditions for granted. Rather, she is seeking unbiased information on what the composer might have intended, given that there is no record (written or otherwise) of how Bach’s “Sei Solo” were performed in the 18th century, let alone by the composer himself.

So, unabashed by traditions, she must have looked at the facsimile, realizing that here, tradition has a problem with Bach’s annotation (see above). Most artists apply the Andante (walking) annotation to the pulsating quaver accompaniment. Also the melody line follows the quaver pulsation. With this, the Andante quavers make it hard, if not impossible still to sense crotchets of the 3/4 meter, even with the most fluent “ordinary” interpretations, such as Viktoria Mullova‘s. The solution seems logical: select a pace that applies the Andante to the 3/4 meter.

Some might think that this is not just “out of tradition”, but far-fetched. I think it isn’t! Just check how Bach uses beams to connect groups of quavers in the accompanying line: a clear indication that Bach was thinking in crotchets (if not occasionally entire bars).

The Result: A Revelation!

The result is so radically different from convention that barely anybody (here, 1 in 25 artists) dares to follow this path. At first, the reaction may be “How can one possibly play this solemn, calm piece that fast?” However, the annotation indeed is Andante, not Adagio, and 3/4 time. So, Isabelle Faust performs the crotchets at around♩=40—a slow, measured Andante. The focus is on the melody, which no longer has the traditional, solemn attitude, but turns into a narration full of serene happiness; wonderful!

The lower voice retracts into pure accompaniment. In the repeats, Isabelle Faust enriches the melody by an exquisite, fitting set of extra ornaments (without ever overdoing it, off course). This further draws the listener’s attention to the melody voice. Now, I hesitate returning to “ordinary” recordings of this movement, as these might feel boring in comparison to this one!


Things will definitely fall apart if you listen to this movement in Faust’s performance immediately before or after (any) other ones. It takes some unbiased listening into this—and only then one can fully appreciate it. I don’t mean to say that this is better than all others. And I’m not the person to judge whether this is “right”, or rather the vast majority of existing performances & artists. When I personally look at Bach’s manuscript (see above), the evidence tells me that indeed Isabelle Faust is right, though. That said, I can still enjoy and fully appreciate what artists such as Amandine Beyer, Viktoria Mullova, or Augustin Hadelich are offering: their performances may not be exactly Bach’s intent, but it still is wonderful music!

IV. Allegro (5’29”)

Maybe not quite as much of a revelation as the third movement, but still spectacular in all aspects: dynamics, agogics, articulation, phrasing, tempo, intonation, Klangrede in motifs, throughout. In this recording, Isabelle Faust did not (yet) use a baroque bow, and her pitch is modern (a’ = 440 Hz). However, her use of the modern bow gets her as close to a baroque bow performance as it can possibly get. The sound is close to that of pure gut strings (or perhaps she was using gut strings??). It simply can’t get any better than this!

Total Duration: 21’41”

Rating: 5 / 5 / 5 / 5 = 5.00

Comment: Strongly recommended, even spectacular—I would not want to miss this recording!!!

J.S. Bach, Sonatas & Partitas — Giuliano Carmignola (CD cover)

Giuliano Carmignola, 2018

Instrument: 1733 violin by Pietro Guarneri (1695 – 1762), Venice; 2007 bow by Emilio Slaviero, after Nicolas Leonard Tourte, 18th century. Pitch: a’ = 415 Hz

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (4’15”)

In moving from slow to “fast” performances among the 25 recordings in this comparison, this is #9, and the first one in a group of what I would call “mid-field tempi” (durations between around 3’45” and 4’15”). That said, it is still at the limit of being too slow, in that it is hard to perceive the rhythmic backbone, the fundamental harmonic scheme. The latter is also the consequence “a little too much” in “local expression”. I mean, somewhat of an excess in dynamic articulation in small phrases and motifs. This results in a somewhat “fragmented” impression.

However, of course, I like the “HIP articulation and sound”, the sonority. And I like the agogics, the rhythmic swaying, how Carmignola uses the scales and arches in the ornaments (yes, ornaments, of course!) to pick up / gain momentum for the next phrase / motif. Beautiful! About the final bars (see above): accelerating bow vibrato on the first crotchet, trill on d’♯.

II. Fuga (8’42”)

Giuliano Carmignola uses almost extreme staccato for the fugue theme, too terse and dry (if not sometimes “military”) for my taste. Also, this gives unnecessary extra emphasis to arpeggi that unavoidably are much broader in the articulation. And it does not help in deciphering the melodies that are hidden in such chords. Sure, the sound is that of a period violin—beautiful. And the intonation is infallible.

On the other hand, there are also a few disruptions in the flow—some in connection with arpeggiated chords, others in semiquaver scales (e.g., in bars 84 & 86) that don’t quite fit into the musical flow. The best parts are in the longer semiquaver passages, such as bars 205ff. On the other hand, I don’t see what the demisemiquaver cadenza prior to the ending (bars 286/287) must be such a noisy détaché?

III. Andante (5’38”)

The tempo is almost identical to Amandine Beyer‘s, but still somehow feels noticeably more fluent. One can attribute this probably to slightly richer dynamics, combined with the occasional, extra ornament. The latter is not a surprise with this artist. One might argue that some of these extra ornaments defeat the simplicity, the purity of the melody line. They are of course following the artist’s personal esthetics, preferences and taste. Not all of them fit my own taste, I should say. Finally, to me, some of the (“belly”) dynamics in the melody are bordering on being excessive, too conspicuous. I should also state that Giuliano Carmignola’s intonation is absolutely flawless.

IV. Allegro (6’14”)

Considering the ₵ / Allegro annotation, Giuliano Carmignola takes a rather moderate tempo at ♩= ca. 80 (only Monica Huggett is slower among the 25). The advantage of that moderate pace is that the artist can apply a lot of Klangrede at the level of motifs. His articulation and dynamics are highly differentiated, the agogics swaying. And Giuliano manages really to play alla breve, while at the same time applying care and diligence to the details at the level of motifs. All of course with the grainy sound of gut strings. The latter and the baroque bow in fact do impose tempo limitations, if the artist wants to avoid sounding superficial in fast motifs.

Everything appears totally natural—even what in traditional interpretations sometimes appears like little “stumbling blocks”: subtle jeu inégal effortlessly integrates the occasional groups of four demisemiquavers in the second part into the musical flow, such that they are almost inconspicuous. With very few exceptions, the artist resists the temptation to add extra ornaments in this movement.

Total Duration: 24’49”

Rating: 4 / 3 / 4 / 5 = 4.00

Comment: A very good, historically informed performance. To me, it isn’t always at the same level, though. Some the artist’s choices (e.g., in ornamentation) may feel controversial.

J.S. Bach, Sonatas & Partitas — Mikhail Pochekin (CD cover)

Mikhail Pochekin, 2018

Instrument: 1720 (modernized) violin by Francesco Gobetti (1675 – 1723), Venice, with metal strings; Tourte type modern bow by Eugène Sartory (1871 – 1946).

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (3’56”)

After the two D minor versions (Vito Paternoster / cello, Atilla Aldemir / viola), I welcome the “return of the ornament”! Although very similar in tempo, this restores the balance between “backbone” and ornaments—well done! Things I note in this performance:

  • an occasional, rather nervous and prominent vibrato (the biggest “stumbling block” to me, really);
  • frequent, somewhat excessive, expressive “belly dynamics”—not so much on single notes, but over motifs and phrases;
  • flawless intonation, also where there is no vibrato at all (e.g., in double-stop intervals involving an empty string);
  • Mikhail Pochekin performs the ornaments with considerable metric freedom. Some may call this excessive—I really like the idea, at least as long as it is applied to movements like this one.
  • Most importantly, Pochekin leaves no doubt about what is ornament, and what isn’t (and the metric freedom reinforces that distinction).

If only the artist dared playing with less vibrato! About the final bars (see above): distinct, accelerating bow vibrato on the first crotchet, trill on d’♯.

II. Fuga (7’33”)

A good, fluent tempo, excellent technique. Very expressive and lively in articulation, agogics and dynamics. As in the Grave, I note a tendency towards “belly dynamics” in highlighted notes, as well as climaxes in motifs and phrases. That’s a little too predictable, I think, and at some point, this becomes more of an idiosyncrasy than just a “feature”. This also contributes to the impression that there is a little too much focus on motifs, compared to big, melodic lines, and/or the larger scale, polyphonic fugue structure.

III. Andante (5’06”)

As already in the Grave, there is (to my taste) an occasional excess of vibrato, which adds “local nervousness”. And again, there is a tendency towards “belly dynamics”. These may be OK if used occasionally—but when used too often, they turn into a distracting idiosyncrasy. It’s a tricky movement, because of the quaver structure that seems to conflict with the Andante 3/4 meter. Here, one can feel this in the slight unrest that is present almost throughout the performance.

IV. Allegro (5’14”)

A fast non-HIP performance with very detailed, diligent, “speaking” dynamics, agogics and phrasing. Exceptionally clean in the articulation, down to the fast demisemiquaver figures. Key notes in a phrase are carefully highlighted by subtle broadening. Technically excellent, no doubt, among the top non-HIP performances. I giving this the same rating as Augustin Hadelich in the same movement—though I do have a slight preference for the latter.

Total Duration: 21’49”

Rating: 4 / 4 / 3 / 5 = 4.00

Comment: Congratulations to the artist for an excellent performance! I can certainly recommend this—with only minor reservations.

J.S. Bach, Sonatas & Partitas — Atilla Aldemir (CD, cover)

Atilla Aldemir, 2019 (Viola, D minor)

Instrument: 1560 viola by Pellegrino Micheli da Montechiaro (a.k.a. Peregrino Zanetto, ca. 1520 – ca. 1606), Brescia. Pitch: a’ = 433 Hz

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (3’58”)

Interestingly, both transposed recordings (Vito Paternoster‘s on the cello, and this one on the viola) perform at almost the same tempo. One can call both historically informed (apart from the alteration in pitch and key), and both use vibrato very sparingly. Also, in both cases, the lower pitch shifts the focus towards the ornaments, which is to the disadvantage of the movement, I think. However, that’s about where the similarities end—the experience is vastly different, overall.

Aldemir uses less / more limited vibrato than Paternoster. As indicated, the viola keeps the bulk of the ornaments in the most intensely singing register of the viola, hence in excess focus. On the brighter side, compared to Paternoster’s interpretation, there is much less of an excess in Klangrede / dynamic and agogics, less over-articulation. However, compared to the violin, this can’t compensate the shift in focus. I wouldn’t call the viola version unattractive per se. However, the recording also gives the impression that this movement is challenging to perform on the viola—the intonation isn’t always flawless (especially close to the beginning, sadly). About the final bars (see above): subtle bow vibrato on the first crotchet, trill on g♯ (from d’♯ in the original).

II. Fuga (8’35”)

Same tempo as Rachel Podger, but without her idiosyncrasies. And the dynamics are careful and differentiated, the articulation is more natural here, and full of Klangrede. The intonation is OK—though one can sense the challenge in this piece! I certainly very much like the viola sonority in this fugue: the full-bodied lower voices have more clarity and relative presence, even though the viola puts the singing of the a’ string into the center of the listener’s attention. Impressive!

III. Andante (4’53”)

Again the same overall timing as Rachel Podger! As in the first movement, the downshift in pitch moves the top line into the center of the focus. With this, the attention remains on the melody, while the quaver pulsation in the lower voice clearly is accompaniment only. Here, this works to the advantage of the music, I think.

Atilla Aldemir’s melody line is a beautiful cantilena, mostly legato. And the absence of vibrato highlights the beauty, the inner harmony of the melody line. Besides the calm Andante quaver pulsation, one can (just about) feel the 3/4 meter. I should also say that the occasional intonation issues that I noted in the Grave are all gone here!

IV. Allegro (6’08”)

Tempo-wise, Atilla Aldemir is right in the middle of a group consisting of Giuliano Carmignola, Rachel Podger, Christine Busch, and Amandine Beyer—and that is certainly a promising indication! Well, the performance does not quite live up to that promise. In general, the articulation is detailed enough, one can sense the ₵ meter. Dynamics and agogics are maybe a little less detailed and refined than the top-class violin performances. However, Aldemir tends to lose a bit of momentum across the movement. His staccato feels somewhat rigid / stiff, “woody”. The intonation is fine—even though rarely, one can sense that it is a little harder to achieve than on the violin, due to the larger stretching in the left hand. But with that in mind, it still is a very good performance.

Total Duration: 23’33”

Rating: 3 / 4 / 4 / 4 = 3.75

Comment: In these pieces, the viola clearly presents the bigger challenges than a violin. However, if you like the warm, characteristic and characterful sound of the viola, this recording is certainly worth a strong recommendation!

J.S. Bach, Sonatas & Partitas — Tomás Cotik (CD cover)

Tomás Cotik, 2019

Instrument: 2000 violin by Marc de Sterke (*1948), Emmendingen bei Freiburg / Germany; baroque bow.

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (3’37”)

A little too careful, too conservative in using metric freedom in the ornaments, which blurs the distinction between ornament and the rhythmic / harmonic backbone. Too bad: one would expect that the relatively fluent tempo (#22 out of 25) makes it much easier to perform and leave ornaments as such! Also, the accents / highlighted notes are a tad too demonstrative. I do like the sparing use of vibrato, though, and the moderate use of dynamics.

There is also a slight sense of restlessness—in a Grave movement! That’s not because of the tempo, but because Tomás Cotik somehow avoids outlining big phrasing arches using agogics and occasional resting points. About the final bars (see above): bow vibrato on the first crotchet, trill on d’♯.

II. Fuga (6’41”)

Throughout the sonata, Tomás Cotik consistently is among the fastest performers, both in fast, as well as in slow movements. An attribute of doubtful value, I think. In the fugue, only Augustin Hadelich (2020) is undercutting his overall timing. Cotik does apply differentiated, careful dynamics. However, I’m not surprised to note a lack of agogics and differentiation in the articulation (a.k.a. Klangrede). Yes, the semiquaver segments in the episodes are light, gentle, “airy”, and contrasting to the expositions. In the latter, though, coarse and rather uniform articulation dominates, and with the push forward there seems to be no time for agogics and phrasing arches.

Cotik’s chords are virtually simultaneous, almost forceful, and often not arpeggiated at all. The tempo alone makes it hard to the listener to follow the polyphonic textures. At the end, the listener is somewhat out of breath. Even within an exposition, shouldn’t there be breathing, occasional (even just microscopic) resting moments?

III. Andante (4’31”)

After Isabelle Faust (2012) and Thomas Zehetmair (1982), both clearly faster than all others, Tomás Cotik and Viktoria Mullova both share the same, rather fluent (“fast”?) tempo. Yet, how different the two recordings sound and feel! Tomás Cotik’s tone is rather grainy, a tad rough even. Every quaver in the lower voice receives a little “belly accent”. One could be malicious in claiming that all this is slightly exaggerated, just so that the listener instantly notices that this is a “historically informed” performance?

The quavers are all separated (détaché) and rather uniform. And the melody voice (as it is done using the same bow) shares an identical dose of “belly accents”. For one, this makes the movement sound rather uniform, short-breathed, lacking larger structure / phrasing. And there is not much (if any) 3/4 meter left. I do sense a tiny, tiny bit of run-away tempo in the second part. This adds to the slight unrest caused by the monotonous chain of “belly accents”. A missed opportunity, in several ways…

IV. Allegro (5’02”)

Very clear in the articulation—albeit at a tempo (among the four very fastest within the 25 recordings) that does not allow much differentiation across the movement. The articulation is so pronounced in its switching between détaché and (percussive, noisy) staccato, that it turns into a “feature” which Cotik applies in a monotonous way. It’s as if the artist meant to “rub the listener’s nose” in this. The tempo not only feels breathless, but it also contributes to what I see as the biggest flaw in this interpretation: rhythmic distortion.

Tomás Cotik persistently alters chained dactylic motifs (semiquaver, followed by two demisemiquavers), such as in bar 3, and then throughout the piece, into triplets. If Bach wanted triplets, he clearly could (and would) have specified this in the notation. At the pace of this performance, triplets are of course easier to play than dactyls—but it’s not a technical issue, as there are isolated dactyls that the artist performs as written. So, this must have been a deliberate, conscious choice. One that I don’t understand and can’t agree with. Actually, it just makes the performance sound “rhythmically sloppy”, without benefit or enrichment to the music. Note: unlike in slow movements, these dactyls are not mere ornaments that the artist may (or may not) freely alter (in repeats only, preferably).

Total Duration: 19’51”

Rating: 3 / 3 / 3 / 2 = 2.75

Comment: A strong-willed interpretation (“enforced HIP”, I might call it) with a tendency for excessive (fast) tempo. And with some much debatable choices in how the artist is reading Bach’s score. I hesitate recommending this “for general consumption”.

J.S. Bach, Sonatas & Partitas — Augustin Hadelich (CD cover)

Augustin Hadelich, 2020

Instrument: 1744 violin “Leduc, ex-Szeryng” by Giuseppe Guarneri del Gesù (1698 – 1744), Cremona; baroque bow by Rüdiger Pfau.

For general comments concerning recording quality and settings, acoustics, as well as the artist’s approach to the “Sei Solo” see the posting on the Sonata No.1 in G minor for Violin Solo, BWV 1001.

I. Grave (4’43”)

Apart from Joseph Szigeti’s interpretation, this is slowest within the lot (Szigeti is about 12% slower). The tempo difference is noticeable—yet, Hadelich (to me) still is much too slow. Not only is the underlying meter (4/4) not recognizable, but also, at this pace, the ornaments receive far too much weight. Not the least because Hadelich spends utmost care and attention to “formulating them all out”. Yes, it’s beautifully done (and with technical perfection, needless to say), and Hadelich’s interpretation breathes in remarkable, broad phrases. But it’s phrases and melodies from ornaments—at the expense of the big, underlying melodic lines. About the final bars (see above): plain trills (d” – d”♯) on both crotchets.

II. Fuga (6’24”)

Augustin Hadelich offers the fastest fugue performance, by a large margin. A sport of sorts? I don’t think so—Hadelich is too serious a musician for this. The “why?” question is moot, but one can certainly ask: how does he do it? Well, without doubt, Hadelich is technically one of today’s best violinists—so, he does not face any technical hurdles, his playing is virtually perfect in clarity, articulation, tone / sonority, agility / virtuosity, and intonation. So, in all these disciplines, he is among the very best of all the 25 artists in this comparison. But is that enough to give him the crown overall?

While the tempo does not form a challenge to the artist, it still does so for the listener: with the fascination for all the perfection and virtuosity it is too easy to forget or ignore that this is a fugue, with themes to follow in a polyphonic texture. Viewing the piece as homophonic, Hadelich does apply diligent dynamics, agogics, and phrasing. However, what I miss is some extra focus on shaping the fugue theme through polyphonic structures, and maybe some more focus on transparency in the latter? All this is easier to achieve with a more considerate tempo. At the same time, I concede that the performance never feels rushed / exceedingly pushed.

Yes, the baroque bow helps with the lightness of the articulation, and it avoids Nachdrücken, but for a “proper” HIP performance, gut strings would definitely be a must. And these would very likely also moderate the tempo a bit, due to their slower response.

III. Andante (5’31”)

Augustin Hadelich’s pace is almost the same as Gidon Kremer‘s in 2002 (ca. ♪=62), though his articulation on the quavers is gentler and still permits feeling the 3/4 meter. A calm, serene, introverted interpretation—beautiful, melodic.

Interestingly, in the first part, there is a trace of a ritardando before returning to the beginning for the repeat (fetching breath, in a way). In contrast, in the two lead-back bars in the second part, the artist appears to let the pace slip into noticeable accelerando. I’m sure this is not accidental. I just don’t see why, what the artist’s intent was. It caused me to watch out for other “tempo instabilities”. I didn’t find any, but still left me with the impression of a subtle (hardly noticeable) unrest in the second part…

IV. Allegro (5’35”)

Technically absolutely perfect! However, thanks to the baroque bow, it does not sound overly polished. Of course, one can also easily tell that Augustin Hadelich is using modern violin strings. Nevertheless, articulation cleanness, differentiation and detail, as well as phrasing are on a par with the best “fully HIP” performances. The same holds true for the detailed and diligent dynamics. Fast / fluent, but never rushed, nor mechanical, and with subtle agogics. Excellent, indeed!

Total Duration: 22’13”

Rating: 3 / 4 / 5 / 5 = 4.25

Comment: Augustin Hadelich uses a modernized violin with modern strings. However, thanks to the baroque bow, one might call this a “largely HIP” performance. The one thing missing for “proper HIP” is the graininess of the sound of pure gut strings. Plus, maybe, the lower, baroque pitch (such as a’ = 415 Hz). Too bad (to me) that the tempo choice in the first movement is sub-optimal—otherwise this would be right “up there”, with the best of HIP performances. Still: a recommended recording, very much worth experiencing!


No surprises here: the outcome is very much in line with that in the preceding reviews in this series (Sonata #1 in G minor, Partita #1 in B minor), and the above comments should speak for themselves.

Other Review Posts on J.S. Bach’s “Sei Solo“, BWV 1001 – 1006


Some of the media / recordings were kindly supplied by agencies and artists for the purpose of this review:

The author would like to thank for these submissions. In fact, they motivated me (after years of pondering the idea), finally to tackle this major project in earnest.

AboutImpressum, LegalSite Policy | TestimonialsAcknowledgementsBlog Timeline
Typography, ConventionsWordPress Setup | Resources, ToolsTech/Methods/Pics/Photography

1 thought on “Bach: “Sei Solo” — Sonata No.2 in A minor for Violin Solo, BWV 1003”

Feel free to comment — feedback is welcome!