My Cart


Edition Comparison - Rose: 32 Études

Disclosure and disclaimer

Cameron Hewes, the author of this article, is also the publisher of one of the Rose 32 editions (Rivernote Press, edited by Larry Guy).  He contributed editing suggestions to Guy and was responsible for the music engraving preparation.  Hewes has made considerable effort to present this edition comparison in an objective manner whenever possible.  Rivernote Press is also the publisher of Daniel Bonade’s 16 Phrasing Studies, as part of the compilation The Complete Daniel Bonade

Although great pains were taken by the author to ensure accuracy, readers should bear in mind that the sheer amount of data collected for this edition comparison guarantees at least a small possibility of error.  If you wish to submit a correction or request a clarification, please contact the author at CAMco.

The Rose Report

CAMco offers a thorough examination and comparison of the Rose 32 in The Rose Report, which includes many visual excerpts from the publications; the edition comparison on this page is a much shorter summary of that report.


For any reader whose edition lacks bar numbers, they will find it helpful to download and refer to CAMco’s The Naked Rose, which contains bar numbers for absolutely all measures.

Comparing the following editions:

Rivernote Press
(originally Leblanc)
r. 2021
Ben Andrew

Besides the twelve complete editions in print today, this edition comparison includes two other publications: (1) the earliest known edition of 1893 from Evette & Schaeffer, a printing of which is housed at the Bibliothèque nationale de France, and (2) Daniel Bonade’s 16 Phrasing Studies, which comprises only the odd-numbered Rose études (available in Rivernote Press’ The Complete Daniel Bonade).  Note that the 16 Phrasing Studies received several corrections which aren't present in any printings prior to the Complete Daniel Bonade's Second Edition (December 2021).


Of the complete editions, four are notated by hand (DVR, IMC, LED, and STX), and eight are notated by computer (BIL, CF, CW, EUF, IMD, JI, RIV, and ZEN).

Prior to its new edition in 2002, Carl Fischer produced a different version of the Rose 32 which was first published in 1913.  DVR is a reprint of this old Carl Fischer edition with edits made, and has many unclear markings as a result of image scan degradation; experienced readers may be able to overcome these legibility issues, but they could present considerable difficulties for newer readers.  IMC is also a reproduction of this 1913 Carl Fischer edition, with superior legibility over DVR and an average three edits per étude.

LED is a reproduction of E&S with edits made, an average of about 16 alterations per étude. Although not as egregious as DVR, LED also suffers from some unclear markings as a result of image scan degradation.

Six editions have off-white paper (DVR, EUF, IMC, RIV, STX, and ZEN), and the other editions have bright white paper (BIL, BON, CF, CW, IMD, JI, and LED).  For any readers purchasing at CAMco who wish to avoid bright-white paper for legibility reasons, please contact CAMco to verify a particular edition’s paper material appearance, as publishers may occasionally change materials over time.  All editions have an acceptable level of paper thickness except DVR, whose paper is so thin that the notation shows through the other side of the page, and the paper won’t stand up as well to pencil writing/erasing and other wear.

Three editions, no doubt in an effort to save on paper costs, fit three études onto two pages for étude Nos. 15-17: DVR, IMC, and STX.  These études are still readable, but are uncomfortably cramped as a result of this material-saving measure.

For single-stave publications like the Rose 32, most musicians prefer at least 7 mm wide staves because they are simply easier to read.  By this standard, three editions have staves which are too small: IMD, CF, and JI.  While JI’s slightly undersized stave is less noticeable, the readability of IMD and CF definitely suffers due to their narrower staves.


Readers are warned that STX changed the numbering (and ordering) of 4 études: Rose No. 14 = Hite No. 17, R. 15 = H. 14, R. 16 = H. 15, and R. 17 = H. 16 (this edition comparison consistently refers to études using Rose’s original numbering scheme).

CAMco conducted a Notation Count-up by simply counting the dynamics, articulation, and other markings to literally answer the question of how much notation is in each version of the Rose 32.  For those who consider the earliest-known edition of Evette & Shaeffer from 1893 to be a “benchmark” of sorts, this tallying process can also answer the question of “How much notation is there compared to the amount which Rose marked?”  While the use of too much “ink” guarantees an uncomfortable read, a comparatively sparse amount of markings can still impair an edition if the markings are not placed legibly, so the Notation Count-Up should only serve as a starting point of comparison.

Notation Count-Up , all categories

CAMco Rose 32 Etudes Notation count up chart

*RIV has a total of 93 fingerings mentioned in the études’ accompanying commentary text. 
(Individual charts for each Notation Count-Up category with additional discussion are included in Section 3 of The Rose Report.)

Editing Issues

For older French publishers, a single accidental sign commonly applied to all octaves within a given measure, whereas today’s engraving standards typically recommend a new accidental for each additional octave in this context.  BIL, DVR, LED, IMC, and IMD occasionally apply the older practice of applying accidentals in only one octave within a measure.

Beams are meant to group notes together to simplify the reading of beats, according to the music’s meter, but E&S sometimes “breaks” beams when they should stay together, and in other situations keeps notes beamed together when they should be broken.  Unfortunately, many editions copy the beaming seen in E&S, even in the cases where E&S isn’t even consistent with itself within a single given etude. 

Étude No. 3 m. 9 is the earliest instance in E&S of a pair of markings which have caused confusion for some following editors, since the markings on beats 1 and 3 could be interpreted as dim. hairpins or accents:

The accents-versus-hairpins issue arises in several spots throughout the Rose 32, including etude Nos. 4, 9, 14, and 15.  Subsequent editors must somehow reconcile these differences in marking appearance.  Some strictly follow E&S' markings, while others seek to unify and clarify the markings according to the expectations of today's readers, and still others occasionally change or remove the markings altogether.  

Editors must decide how to manage the many grace note slurring inconsistencies present in E&S; if they are not slurred according to a consistent rule, this can cause confusion for readers.  Today’s engraving standards call for grace notes to be slurred to the following measured value note if they are meant to be slurred.  BIL, IMD, and LED are the least consistent because they most closely follow E&S—however, ultimately all editions besides RIV include significant inconsistencies.

Dynamics are perhaps the most contentious aspect of editing for the Rose 32, in how precisely they are physically placed, and how specifically/numerously they are applied.  E&S has the least amount of dynamics, and most editions generally follow E&S' amounts: BIL, CF, DVR, EUF, JI, IMC, IMD, LED, and ZEN.  The remaining three editions (CW, RIV, and STX) have noticeably more dynamics, and this disparity is greater in the even-numbered études; while E&S rarely marks more than a single starting dynamic letter in the technical even études, CW, RIV, and STX instead endeavor to instruct the shaping of phrases by adding numerous dynamics. Section 3 of The Rose Report covers dynamics in more detail.

The earliest editions of the Rose 32 have some comma markings that are potentially misleading for the uninitiated reader, since these markings aren’t always intended to instruct intakes of breath. For instance, BIL & LED mark a comma within the first three measures of several études, when it is far too early to need a breath, so readers must understand that these markings are likely intended as breaks or “lifts,” and not intakes of air.  A few editions differentiate between breaths and brief spaces without breathing by using different markings; for a break, EUF uses a “V” mark, and BON uses a vertical slash.  While this does help to avoid some potential confusion, readers should still plan to evaluate every breath/break marking carefully to best support the phrasing framework.

CF is the only edition which notates absolutely no breath marks.  Of the remaining editions, four closely match the amount seen in E&S (DVR, IMC, IMD, and JI), and the other seven have more.  STX notates the most breath marks by a large margin, and CW is relatively close to STX.  Although ZEN also has many breaths marked compared to E&S (still less than CW and STX), it has the added consideration of marking roughly a quarter of them as optional by enclosing them in parenthesis.  CW differentiates its breath marks by using black for "standard breaths" and grey for where “...breathing isn’t as easily interpreted such as a phrase ending.” 

Readers are warned that BIL, EUF, LED, and ZEN choose to replace some notes with rests in order to take a breath, without signaling that an originally sounding note has been expunged.  These omissions occur in étude Nos. 4, 12, 25, 26, 30, and 32.

BIL has the most fingerings notated in the actual music (usually one or two per étude), and uses the numbering system seen in fingering chart of Eugène Gay’s Clarinet Method.  LED notates about half as many fingerings as BIL in the music, and uses the same numbering system.  STX marks roughly the same amount of fingerings in the music as LED, although it focuses more on left- and right-hand pinky cluster directions and slides over forked/trill fingerings compared to BIL/LED.  CW has the least amount of fingerings marked in the music, about one-third of BIL’s total, and uses the same nomenclature as STX.  RIV technically offers the most fingerings of any edition, roughly three times more than BIL, but keeps them outside of the actual music notation by only mentioning them in the accompanying commentary text for each étude. 

Readers are warned that due to error or disagreement, there are bar numbering differences in many études (Nos. 3, 15, 16, 18, 19, 20, 21, and 30).  CAMco's reporting carefully refers to certain spots using referential position language (ex: “10 bars before the end)—as opposed to bar numbers—when necessary.

Although E&S marks repeat barlines in only one étude (No. 14, repeating mm. 25 through 47), nearly all subsequent editions add repeat barlines to five other études (Nos. 6, 8, 20, 26, and 32); only RIV exactly follows E&S’ usage of repeat bars throughout the entire publication.

Section 5 of The Rose Report lists a handful of notable differences between the editions, proceeding étude-by-étude.  They point out various errors that should be fixed, and spots where editors endeavored to make Rose’s études more similar to the original Ferling studies. A particular editorial approach may also bring to light a new phrasing or expressive opportunity to consider.  The following list selects a single interesting point of differentiation for each etude:

  • No. 1: editions use between four to seven slurs for the first line of music, from mm. 1 to 5.
  • No. 2: while E&S has contrasting articulation for the thirty-second notes in mm. 27 and 29, some editions unify the articulation in these two spots. 
  • No. 3: in m. 24, E&S notates the last 2 notes' rhythm as [dotted-thirty-second + sixty-fourth] notes, but some editions change the rhythm to two thirty-second-notes.
  • No. 4: in m. 21, some editions notate the fourteenth note's pitch as clarion C5 instead of the throat tone A4 seen in E&S. 
  • No. 5: slurring is either consistent or varied in several analogous spots, including mm. 2 and 35, mm. 3 and 36, or mm. 51 and 53.
  • No. 6: editors disagree about the pitch of m. 59's second note; some mark F-sharp4, and others mark F(natural)4.
  • No. 7: in m. 39, E&S places a natural sign one note too early (presumably)—the natural sign should effect the throat tone G4, not the A4—and some following editions copy this error.
  • No. 8: some editions notate arguably excessive courtesy natural signs, like those often seen in mm. 62, 63, and 65.
  • No. 9: some editions mark a slightly slower tempo in m. 19, an indication not seen in E&S.
  • No. 10: in m. 8 of E&S, beat 4 is erroneously missing; there is roughly an even split of subsequent editors either adding a quarter-rest on beat four, or changing the note value on beat three to a half-note.
  • No. 11: editors disagree about whether m. 7's second note should have a drop in dynamic level; E&S marks a forte on the second note, while some following editions mark either piano or mezzo forte.
  • No. 12: m. 27's last two pitches are clarion G5 and E5 in E&S (as seen in the original Ferling étude), and all later editions except RIV change them to clarion E5 and D5. 
  • No. 13: editors disagree about whether mm. 5 and 6, which are sequentially related, should have uniform or contrasting articulation.
  • No. 14: there are 3 spots of pitch disagreement amongst editors: m. 10's first grace note (D5 or D#5), m. 23's seventh note (A4 or A#4), and m. 43's fifth note (C#5 or A#4).
  • No. 15: dynamics in mm. 7–8 differ widely amongst editions; some establish a dynamic peak in m. 7 or m. 8, while others solely notate a diminuendo in this area (and others have no dynamics here)
  • No. 16: editors disagree about how consistently the articulation patterns [slur 2 + tongue 2] or [slur 2 + slur 2] should be employed; the most common spots of variance occur in mm. 26, 33, and 39.
  • No. 17: editors may omit some of the five total possible fermatas: m. 8 (downbeat), m. 21 (last eighth-note and following eighth-rest), and m. 29 (beat 3's eight-note and following eighth-rest).
  • No. 18: following the “meno mosso” in m. 17, editors place the “Tempo I” text in 1 of 3 different spots; if the new tempo is intended to include the pickup note, then the text must be positioned far enough to the left such that it includes the pickup note.
  • No. 19: E&S' tiny hairpin swell on m. 42's downbeat compelled some later editors to clarify and/or tweak the dynamics according to today's engraving standards.
  • No. 20: there is disagreement about the pitch of m. 30's last note; E&S notates a throat tone A4, and some later editions notated clarion C5 or throat tone F4. 
  • No. 21: some editions notate the turn in the recapitulation (6 mm. before the end) without a sharp sign, thereby notating the ornament's bottom pitch as B-natural4 (clashing with the turn's first appearance in m. 3); some other editions fail to cancel said sharp when a B-natural4 is called for later in the measure.
  • No. 22: a few editions do not mark staccato dots on every eighth-note, omitting them from seemingly random notes partway through the etude.
  • No. 23: editors disagree about the pitch of m. 13's last note (F5 or F-sharp5) and m. 26's last note (G5 or G-flat5).
  • No. 24: while E&S contrasts the articulation at the recapitulation in mm. 33 & 34 as compared to mm. 1 and 2, some later editions maintain the slurred pairs articulation at the recapitulation.
  • No. 25: the articulation in m. 4 varies widely, as the slurred groups could be (4+6), (5+5), (5+4), or (4+1+4).
  • No. 26: in mm. 9–10's sequence of descending chords, editors disagree about whether or not each fragment is entirely encompassed by a single slur; similar slurring disagreement arises with the sequential chromatic scales in mm. 25–27.
  • No. 27: m. 8 has two pitch disagreements across editions, on the first note (G-sharp4, B4, or E4) and third note (E4 or G-sharp4).
  • No. 28: while E&S independently articulates the downbeats of m. 5 and analogous m. 51, some following editors instead slur these downbeats into the following notes. 
  • No. 29: although not marked in E&S, all following editions add a "poco meno" tempo instruction for mm. 27 to 34.
  • No. 30: a few editions disagree about the pitch of m. 44's fifth note: most notate E5, but others notate A-sharp4 or C-sharp5.
  • No. 31: most editions mark a tie over the barline of m. 3, despite E&S articulating m. 3's downbeat (the original Ferling etude also articulates the downbeat).
  • No. 32: minor differences in articulation occur across editions in mm. 16, 27, and 48.




  • CW collects four of Rose’s étude books in a single volume: 26 Études, 32 Études, 20 Grand Studies, and 40 Études
  • DVR collects two of Roses étude books in a single volume: 32 Études and 40 Studies
  • STX collects three of Roses’s étude books in a single volume: 40 Studies, 32 Études, and 9 Caprices
  • BON includes two publications by Daniel Bonade besides Rose’s 16 Phrasing Studies: Clarinetist’s Compendium (a concise course of Bonade’s fundamental clarinet concepts) and Bonade Orchestral Studies (orchestral excerpts from 98 different works)


Two publications offer audio recordings of the études: IMD and CW.  IMD’s recordings are on a CD which is included with the book, containing recordings of the Rose 32 arranged for clarinet and piano, performed by Philippe Cuper (cl) and Caroline Esposito (pn).  CW’s recordings are freely available online at the publisher’s website, although at the time of writing, it provides audio recordings for only ten of the 32 études.

There are other audio recordings available online outside of sheet music publications, including complete sets by Alexey Gorokholinsky, Claire Grellier, Christopher Mothersole, Salvador Navarro Valero, and Sean Osborn.  A complete set of CF’s clarinet and piano arrangements are recorded by Christopher Hill (cl) and John Walker (pn).

Additional Adaptations

Two publishers offer arrangements of the complete Rose 32 for clarinet and piano: CF (by John Walker) and IMD (by Patrice Sciortino).  CF’s arrangements are available solely in digital (printable) PDF format along with the corresponding audio files, redeemed online through a download code included with the clarinet étude book (the piano accompaniment audio is also sold separately in CD format).  IMD’s arrangements are published separately in physical book format.  Additionally, STX separately publishes a single Ferling étude (No. 27, equating to Rose No. 11) arranged for clarinet and piano by Paul Jeanjean.

Clarinetists may learn from or otherwise enjoy reading Ferling’s original études which Rose adapted.  A popular choice is the version edited by saxophonist Marcel Mule; he expanded the Ferling 48 with 12 additional études, thus “completing” the cycle of keys.

Clarinetist Mark Wolbers adapted Ferling’s Op. 48 to exploit the full range of clarinets with low C extensions, including the modern bass clarinet, contrabass clarinet, A clarinet with low C extensions, and basset horn.

In his 16 Études, clarinetist André Vacellier sought to “finish what Rose started” by adapting the remaining Ferling études left untouched by Rose. 

Wesley Hall has arranged several Rose études into clarinet duets, taking form as three suites (Suite No. 1Suite No. 2, and Suite No. 3).

In Les Essentielles, Didier Yves shortens 30 études selected from Rose’s 32 and 40 Études with the intention of making them more approachable for lower skill levels.

CAMco offers a digital download (PDF) format of the Rose 32 called The Naked Rose, which is completely stripped of all dynamics, articulation, and other expression markings.  Readers can use this “blank canvas” version to explore their own phrasing devoid of outside editorial influence.

Summing Up

Both older and newer editions exhibit weak points in practical issues of spacing, line thickness, beaming, and adherence to other notation standards.  Readers are encouraged to check Appendix C and Section 5 of The Rose Report to catch errors and other blemishes in a given edition. 

Caution is advised with the use of DVR, IMC, and LED due to their legibility difficulties.  IMD, LED, and ZEN align most closely to E&S, particularly in terms of their comparatively sparse markings.  RIV is the only edition available today which includes an accompanying page of commentary for each etude, offering guidance on practice and interpretation.  BIL and CF align their editing most closely to the original Ferling studies, although readers should ensure that these changes are actually desirable in the context of the clarinet.  CW, JI, EUF, RIV, and ZEN are the most clear and comfortable to read.