I love music.

I write about the music I like and have purchased for the benefit of better understanding it and sharing my preferences with others.

Corelli Opus 5, featuring Imaginarium Ensemble and Enrico Onofri

This is a review of the two volumes of Corelli sonatas by Enrico Onofri with Imaginarium Ensemble, emphasizing the new, second volume, but I also compare their playing with a few others in my collection to provide a larger context of this set of sonatas.

Corelli - Bridge Between Early and Late Baroque

My first exposure to Corelli was his opus 6 Concerti grossi, as performed by Trevor Pinnock’s English Concert. During those years, likely around 1990, I bought up all the DGG Archiv recordings as they came out as I couldn’t satisfy my thirst for all the baroque music I was discovering. I listened to that 2 CD set over and over and fell in love with Corelli’s writing and his harmonic progressions. It’s precisely those progressions that I think of when I consider Corelli’s art.

Volume 2 of Corelli's opus 5 sonatas is presented by Imaginarium Ensemble on the Pasacalille label

Volume 2 of Corelli's opus 5 sonatas is presented by Imaginarium Ensemble on the Pasacalille label

Corelli however was more famous in his day not for his counterpoint or his harmonic progressions. It was his violin playing that gained him fame, and no doubt, helped sell his music once it had become published. And yet even for those who never may have heard him play, it was this set, his solo violin sonatas, that kept his name alive. In terms of fame throughout music history, often it was this set that helped define the cross-over point between “early baroque” and “late baroque.” And his set became the new standard for the solo violin sonata with bass for at least fifty years.

Performance Practice

As our knowledge of performance practice has evolved, I believe my original exposure to Corelli may have been performed in a somewhat conservative manner with Pinnock. Comparing that recording with a later one by Fabio Biondi’s ensemble, Europa Galante, exposed a lot more nuance to the music which perhaps could be expressed through the style adopted by the ensemble. I saw Biondi blowing dust off the music. But I wasn't always sure that Opus 111 recording was solid on historical terms, but in terms of enjoyment, I found it more enjoyable than Pinnock.

And “style” in the Biondian sense, accenting off- and on-beats, providing rhythmic exaggeration, and providing dynamic contrasts that can’t be ignored, is part of it. But there’s also the role of the performer with the text. If you go to school (or to a masterclass, let’s say) to learn about baroque interpretation, a rather loaded word, there’s more to playing this music than simply reading notes off the page. This isn't just style, as a general term. What I’m getting at is the matter of what notes that get played that aren’t on the score. One set of rules or customs we might call improvisation. The other--related--are ornaments. Ornaments by some composers are indicated in the score with symbols as part of the musical notation. The problem here is deciding which notes to play with which symbols (and if we are being picky, certain ornaments may have had regional dialects, or preferences by the composers). Beyond printed ornaments, there's a performance culture of adding your own ornaments at the ends of phrases and at cadences. So if we think of ornamentation as decoration, there’s another matter entirely of improvisation. This is where we treat what’s on the page as a mere guide, sort of a basic skeleton of music, and as the confident performer, we can go on and about our business of adding to the music. Critics would tell us the boundaries are bolstered by the style (you wouldn’t necessarily play something in the style of a 19th century violin sonata with a Corelli skeleton, if to give a gross comparison), but then we start to poke at the intention of the published score. We look at Bach’s fifth Brandenburg Concerto as a fitting example of improvisation, where Bach has written-out a cadenza for the solo harpsichord. In almost every recording we hear Bach’s cadenza. But did Bach intend it to always be the one we hear, or is the spirit of a cadenza to invent it yourself? When philosophy becomes involved, it makes for very interesting conversations and some interesting musical results.

Corelli’s Opus 5

So Corelli’s collection got published twice, and in the second publication, it was joined with his own improvisatory versions of slow movements. Performers then always have a conundrum of sorts when performing this set:

  • Do I play it as originally written (more or less verbatim)?
  • Do I play it as written, with the so-called Corelli improvisations (with no way to authenticate these “bonus” lines as authentic to Corelli, despite the admission of the publisher who says they are (and noting that publishers during this time were not always truthful, they’d stretch things to be able to sell music))?
  • Do I recognize that this music is too simple for reading off the page, and that improvisation is likely apt for this literature, and therefore, add my own improvisations as part of the recording?

That’s all something for the violinist to consider. The other consideration is who gets to play alongside you with the second part? The title page suggests two options (violone (a stringed bass instrument) or keyboard), but almost everyone only treats that as a light suggestion.

That’s what’s fun about collecting this opus 5 from different performers. They all don’t choose the same solution at addressing the conundrums. Let’s look at three comparisons: sonatas 9, 12, and 8.

Sonata no. 9 in A Major

This piece is organized in four sections: Preludio - Giga - Adagio - Tempo di Gavotta.


The opening movement presents material twice with repeats, which opens the opportunity for improvisation the second time around. This is a practice born in dance pieces, where musical material might be repeated multiple times, and performers would change it over time. When we see repeats in Baroque music, it is often an opportunity to do something “different” the second time around. I am not expert enough to know if this is always the case or if its always encouraged, but given the context of Corelli’s music, it seems appropriate. Think about the purpose for repeating the music?

I have to go back to my first purchased recording, that on the Hyperion label with Elizabeth Wallfisch and the then-named Locatelli Trio. They used harpsichord and cello for each sonata for the bass part. In this sonata, she reads the publisher’s “Corelli improvisations” in the repeats. I like the result.

Imaginarium also uses harpsichord and cello, but the balance for me is better; there is more emphasis in the mix to the violin and harpsichord. Onfori too adopts “variations” in the repeats, but unlike Wallfisch, he presents his own which are less florid.


The Wallfisch recording again puts so much emphasis on the harpsichord and cello, especially the cello. The violin is lost. The gigue, another dance, gives us repeats and opportunities to add variation in the repeat. Wallfisch and Onofri both take advantage, but it is Onfori who adds more variation.

In his album Corelli’s Legacy, Riccardo Minasi with Musica Antiqua Roma perform the same sonata. This recording is almost the opposite of the Wallfisch in terms of balance; Minasi seems to be overly-represented in the mix. He uses organ and cello. In his reading of the gigue, I liked his improvisatory additions the best, likely, among the three: he diverges the most from the written text.


This is a short movement that joins together the two dances. I really think it’s an opportunity for the violinist to show off their tone and skill with vibrato.

Minasi plays it more or less as written; Wallfisch plays the improvised part published in the second publication of the work. Onofri’s interpretation is the quickest of the three, adopting the simpler style used by Minasi. Like Minasi, he employs a slow vibrato to the long notes which give them this beautiful quality.


Another dance in binary form with the opportunity to improvise on repeats. Onfori’s reading is very mechanical sounding to me, like a clock is driving the pulse. In the mix, many times he’s hiding behind the mechanizations of the harpsichord and cello duo. His improvisations on the repeats are inventive and stylistically sound. All around, his reading was enjoyable.

Wallfisch’s reading sounds less mechanical to me, but her second-time playing is less inventive and playful than Onofri’s. It works, though. I just can’t help get used to her sound being buried underneath the heavy cello of Richard Tunnicliffe.

Minasi attacks this last movement at a slower tempo than the others I’m comparing here, but this slower tempo affords him some more room in the repeats for adding extra notes and, seemingly, room to offer dynamic variation. I also like the continuo in this version; the cello seems to be playing with great space between the notes while the organ is the opposite, smooth and legato, and the two together are delicious.

Recording quality and balance aside, I think all three performances provide variation in how to perform Corelli. I have previously commented on all three recordings, and there is no real merit in continuing to discuss Imaginarium’s first volume.

Sonata no. 12, La Folia

This piece is a theme and variations on the tune of La Folia. It is performed first in the new Onfori recording, but appears last in the published set.

Andrew Manze and Richard Egarr

This recording on Harmonia Mundi was significant because the performers chose to play with just two instruments: harpsichord and violin. I have reviewed it already, but I really liked their reading of sonata 12. Egarr’s contribution really set it apart for me, taking his role of realizing the bass line to new heights. I do question how authentic all his choices were in terms of what we’re accustomed to hearing as historical practice, but he takes the spirit of his task to new aesthetic heights. Manze, who I found too many times earlier to miss opportunities to delight us, somehow recovers in the La Folia variations, offering an astonishing performance. What makes the reading really stand out for me is the arch they give the entire movement, starting simply and ending in full-throttle mode. As one of the last recordings Manze made as a baroque violinist, he finished “on top” for sure.

The other thing about this reading that stands out for me is how the performers linked the different variations. Instead of pausing after each one, they connect them in such a way that makes us feel they are strung together without missing a beat feeling any hiccup. They do pause between movements that really necessitate them, but I enjoyed the continuity. Especially apt was the transition of one particular section to the other, which I felt Imaginarium missed; the beat from the preceeding variation becomes the beat for the next, but the second pair is in 3 instead of a 2/4 and therefore with the same beat adjusted to 3, it naturally feels like it’s going faster despite having the same pulse. There’s a technical term for this--hemiola--and with Manze it’s so well-done. Imaginarium goes faster, but it’s not with the same pulse. This is described here as a metric modulation in general musical terms, but his article isn't specific to Baroque music or hemiola.

Enrico Onofri and Imaginarium Ensemble

The opening is very different than most, he adopts a somewhat connected, almost legato style to the presentation of the opening theme. He also adopts a very rich continuo team with lute, cello, harpsichord. (These players do not all play together throughout, which helps them deliver dynamic contrasts.) With so many continuo players sometimes the balance gets off, becoming too bottom-heavy. Compared to Manze, Onofri’s violin sound and tone is world’s apart, and to my ear, more agreeable.

I enjoyed Imaginarium’s reading of La Folia. But if I could only have one copy of the sonata, I’d take Manze. Manze and Egarr may have betrayed the baroque style that Corelli would have recognized, but it’s modern, flashy, and delivers goosebumps every time I listen to it on the back of my neck.

Sonata no. 8 in E minor

This sonata is presented in four parts: Preludio - Allemanda - Sarabanda - Giga.

Enrico Onofri and Imaginarium Ensemble


Onfori opens with just lute and violin which I think is an excellent combination. The music allows Onofri to shine in one of the things I believe he does best, perhaps better than any other violinist I’ve heard, and that’s to take these nice long notes and procure from them the most milky, creamy sound from them, using what I call a slow vibrato. He’s also open to adding his own ornamentation on the repeats which is nice.

He writes in the liner notes about something I do not quite understand. In speaking to the choices I mention above that performers have to make when performing these collection of sonatas, he takes off the table for the recording “improvisation.” Or to be more specific, the type of improvisation the ensemble employs in a live performance. I think it demands an explanation.

I believe there are practical reasons for staying close to the score for a recording, and it’s one that I can appreciate as a music student. When you’re listening to recordings in the library, following the score, you need recordings that adhere to the score to hear what the composer left us in notation.

Onfori does not cite this reason in the notes. There may also be the choice of choosing “which” improvisation is best, if something could be played ten different ways. And let’s say they did 5 takes of each movement, then chose the “best” out of all these takes. Is that somehow less authentic? I get that recording in a studio is different than a live performance.

I’d also like to see us get to a point where performers and record labels allow us to have a say in the outcomes: if each movement was recorded with 5 takes as an example, why not include them all, and let us, the consumer, select which combination we want? Or let our software mix them up so that we can get a “different” interpretation each time based on the set provided?


I like the tempo adopted here. The lightness in texture is maintained with using lute again for the only basso continuo instrument. I think Onfori’s ornaments used in the repeats are tasteful.


I can see Onofri playing this piece; he keeps it simple, and we could imagine, perhaps, it being sung. His style is such that things work up to the cadence points, where you can see in his face the relish he takes in giving us a little tickle with a trill or shake as the phrase concludes. The tone he’s adopted here is really beautiful, no doubt best captured in the last note of the movement.

This is an example where so much more could have been done to ornament the original line but he’s chosen to keep it simple.


Alongside the other movements, he maintains the somewhat “quiet” or intimate nature of the sonata as a whole. Something likely no other performer would do, when the line takes him to the lower range of his instrument, he changes the tone of his instrument (likely both with the bow and the left hand), and in so doing, he changes the color so dramatically that it excites the listener. I equate it with maybe turning on a colored light in the room for those few notes. It’s an example in the baroque sound world of ornamentation to notes, not in the sense of a trill or mordant, but through color or timbre. It’s one of the things I really appreciate in his playing overall.

Stefano Montanari and Accademia Bizantina


This recording uses cello and harpsichord and lute in the basso continuo. Immediately comparing it to Onofri’s recording, I prefer the more intimate single lute with violin. It’s simply better balance. I also like the drier acoustic in Onofri’s recording. I feel Montanari’s tempo is a little slow in the opening, but it allows him some room to add variation on the repeats with a more florid reading than what’s adopted by Onofri.


Montanari’s repeats go quite a bit further away from the original, which is nice. It almost sounds like a different piece, although I think his adopted tempo is a bit slow for my taste.


Again, Montanari uses the repeats of the dance to provide a new take on the original material. I can’t say for certain if it’s improvised or he’s playing the written-out ornamentation. Again, the tempo for me is a tad slow.


Montanari and ensemble take this piece to a far more dramatic and grandiose place, at least it sounds that way with the full sonority of the ensemble and the very “wet” acoustic. It’s a good example of how the continuo can “add” something, like weight, to a two-part construction.

If I had to choose just one, I’d adopt the recording by Onofri and Imaginarium for this sonata.


So perhaps more so than, say, the sonata for violin and keyboard by Bach, these 12 pieces by Corelli are a more complex proposition for the performers. There seems to be a lot of decision making involved and I think in every case cited here, performers aren’t taking a simple choice in how to approach them, but instead are making “compromises” in terms of the approach for the set. Onofri’s second volume in his reading is a good example: he makes different choices in terms of scoring (four continuo instruments versus one), in terms of ornamentation (guide by his ear, vs. guide by supposedly-captured improvisations), in terms of tempo (sometimes the dances feel leisurely and other times pushed), and in comparing him to others, he’s really no different in these approaches.

His choices however are uniquely his and what I discovered in listening to this new recording specifically (volume 2: sonatas 12, 6, 8, 4, 2, 11) was a very agreeable tone from the instruments (best described as being easy to listen to). Compare that with Montanari, where the acoustic treatment in the recording makes it fatiguing for me after one sonata) or with Manze, whose violin is somewhat abrasive in tone. Onfori has a roundness to his violin sound that I believe he tried to maintain throughout which was nice.

I had high expectations that Onfori’s set would become my clear new reference. I am not sure that is the case. And that says less about the quality of this recording than the nature of Corelli’s text: there are so many variations involved it’s hard to pick just one.

This recording I believe does especially well in the balance and sound quality, save for when the full compliment of Imaginarium is playing.

I’ve learned my own preference in comparing recordings to simple continuo, for instance, the solo lute used in the example cited above from Imaginarium, or the solo harpsichord used in Manze/Egarr. I appreciate Wallfisch is more the slave to the printed improvisation whereas Onofri is not.

All said and done, the desert island version of Corelli’s opus five may just be your own mix. And what’s ultimately sad, I think, is that any one of the these performers could have provided us that “mix” themselves if only the record companies would play along.

Probably more so than other performance I have on record, Onfori is willing to dabble into the obscure and subtle corners of the music to apply the baroque aesthetic. I made mention above of his adoption of a change in tone for when the line dips into the lower register of the violin. Another great example is in the last track of the recording, the Gavotta from sonata 11: like in the earlier volume, he takes this dance with such strict precision, but then on repeats, changes the length and volume of the same line. These are subtle things that you have to listen for, but they are purposefully made in accordance with what we read about in violin performance, and what we might already expect in vocal performance. It’s this type of attention to small details that sets this set apart.

So in the end, I highly recommend this set. And I do so with the admission that one performance or another might stand out as a favorite in contrast to this endorsement from another ensemble. But if I were confronted with the make-believe proposition that my “mixtape” or “custom playlist” wasn’t allowed on my journey to the desert island, I might just pick up volumes 1 and 2 as my overall favorite set thus far. I say that not only because of the very musical result, but also the thoughtful nuance applied to matters of performance and the performer’s duty not only to the text but the sound world and customs of performance.

J.S. Bach: Bach De occulta philosophia

Mondonville, op. 2