Rockstro & Ellis analyse a Rudall & Rose flute

Introduction

Richard Shepherd Rockstro published his enormous "Treatise on the Construction, the History and the Practice of the Flute" in 1890.  In it, he provides us with an analysis of the tuning of a Rudall & Rose 8-key flute, at the time only about 60 years old.   Does his analysis tell us any more about flutes of the time than we can determine today?

Here's Rockstro talking about it.  Comments from me in [square brackets].


548. The Flute in 1826-7. The flute was never more popular than at this period of its history. In England, France and Germany, especially, there were not only excellent manufacturers, but many fine performers, both professors and amateurs. The music for the flute was perhaps no better and no worse than that for other instruments ; all the greatest composers were making free use of the flute in orchestral and chamber-music, and if there were, for that instrument, numbers of airs with variations which were not of great merit, there were, on the other hand, many fine classical compositions, in which a prominent part was assigned to it, by Bach, Beethoven, Gabrielski, Gluck, Handel, Haydn, Hoffmeister, Hummel, Kuhlau, Kummer, Mozart, Reicha, Schneider, Spohr, Tulou, Walckiers, Weber, and a host of stars of lesser magnitude.

Amongst the distinguished professors of the day were Ashe, Birch, Carte, then only nineteen years of age, Lindsay, Nicholson, Roe and Rudall; Berbiguier, Camus, Drouet, Farrenc, Guillou, Tulou and Walckiers; Bayr, Dressler, Fuerstenau, Gabrielski, Karl Grenser, Kuhlau, Kummer, Saust, Soussmann and Weiss; Boucha, afterwards known as Bucher, Gianella, Monzani, Negri, Rabboni and Sola.

Noted amongst manufacturers were Clementi and Co.; Gerock, Astor and Co.; Monzani and Hill; Potter, and Rudall and Rose; Buffet, Godfroy, Laurent, and Nonon; Grenser and Wiesner; Griesling and Schott; Koch, and Liebel.

549. Though neither Potter nor Monzani made flutes with the large holes of the Nicholsons, they were obliged, in common with the other English flute-makers, to follow to some extent the prevailing fashion of the time, and the small holes of 1807 circa went completely out of use in this country.

It has been stated that the large-holed flutes of Rudall and Rose were unrivalled in tone, and even if their intonation was not as perfect as might have been wished, still they were better in that respect than those of any other makers.


INTONATION OF AN EIGHT-KEYED FLUTE.

The tuning of the lowest octave of an excellent specimen of these flutes was tested by Mr. A. J. Ellis and myself, and the results are given in his latest edition of the Tonempfindungen of Professor Helmholtz, (1885).

[A.J. Ellis is an early acoustician, perhaps best known for his translation into English of Helmholtz's "On the Sensations of Tone".]

550. The table below is compiled from the MS. kindly sent to me by Mr. Ellis.
 

Notes.

Vibrations observed. 
Given
to the nearest whole number.

Theoretical vibration numbers for equal temperament.

Theoretical vibration numbers for meantone temperament

c"#

560

571.4

564.6

c"

543

539.2

540.4

b'

518

509.0

505.0

b'b

478

480.5

483.3

a'

461

453.3

451.7

g'#

426

428.0

422.1

g'

404

404.0

404.0

f'#

381

381.3

377.6

f'

36o

359.9

361.4

e'

341

339.7

337.7

d'#

314

320.7

323.2

d'

299

302.7

302.1

c'#

280

285.7

282.3

c'

268

269.6

270.2


551. As the object of the experiment was to prove the real intonation of the instrument, I was of course extremely careful to avoid correction [confirming that he, Rockstro, was blowing the flute, with Ellis taking the measurements.]. To that end, before and after each observation, the sounding of the flute was tested by comparing the pitch of its g' with that of a resonating jar previously tuned to 404 vibrations.

[Tuning to a resonating jar!  Flute researchers sure had it tough in those days!] 

[Rockstro's choice of G=404 becomes clear when we see in the centre column above that its equivalent is A=453.3, in the range A=452-455 given for British High Pitch.  I've highlighted the G row in the table above to remind us that this was his reference note.  He explains why in the lines below.]

The note g' was selected as a datum because it was the nearest to the average pitch of the instrument. Both the f# and the g were fingered with the f key open, and the c" was fingered with the long key. The flute was sounded at its mean pitch, and was carefully kept at an even temperature. Mr. Ellis used a Scheibler's fork-tonometer tuned by himself. See 295.

[295 briefly describes Scheibler's fork-tonometer, a revolutionary device for measuring pitch, involving no less than 52 tuning forks!]

At the time of the experiment it was thought that the flute was about forty years old, but the probable date of its manufacture was afterwards found to be 1827.


So what do we learn?

Let's take a break from Rockstro's narrative to consider what we can learn.

Firstly, we'll probably find the following graph of deviation expressed in cents a bit easier to interpret than Rockstro's lists of vibrations in Hertz:

Secondly, there's our old friend, Flat Foot Syndrome, illustrated there, especially in the navy blue Equal Temperament trace.  With the usual odd fingerprint - C not bad, D flat, C# next and Eb worst. 

Thirdly, Meantone Tuning (in pink) doesn't offer the panacea many have hoped for; indeed it's slightly worse.  Rockstro doesn't mention why he included both Meantone and Equal Temperament in his table - perhaps he was hoping for a miraculous revelation too!

Fourthly, note the zigzag tuning of adjacent semitones in the LH notes G# to c#.  That suggests that the pads on the keys might be too thick, or the opening of the keys set too low.  Or you could argue that it implies that holes 2 and 3 are too big, making A and B sharper than desirable.  But consider the next point.

Fifth, remember that Rockstro tested this flute at the equivalent of A=453.3, High Pitch, whereas the flute dated from a period before High Pitch kicked in.  He couldn't really do much else - he'd argued back at 299 that:

the idea of [English pitch] having risen a semitone in the last twenty or thirty years is utterly without foundation.

This is another example of Rockstro's all-too-clever use of deceitful writing.  It may not have risen a semitone in the last 20 to 30 years, but it had risen about 90 cents since the time this flute had been made, as we showed in The Rise and Fall of English Pitch!  But Rockstro, one of the leading proponents for retaining High Pitch, couldn't afford to let readers dwell on that - it would weaken his arguments in 300 that a 70 cent reduction in pitch to the French standard was variously "impossible" and "utterly impractical".

Of course, had Rockstro chosen to test the flute against a lower pitch, he would have found it better in tune - the top notes flatter and the bottom notes sharper. 

You might be a bit surprised that F# hasn't ended up as flat as usual.  Rockstro tips us off in 551 above when he reveals that he opened the Fnat key for F# and G.  Interestingly, he doesn't open the c key for c#.  Neither workaround features in fingering charts from the period of the flute in question.

Now, back to Rockstro...


MEASUREMENTS OF A FLUTE BY RUDALL AND ROSE.

552.   I regret that I allowed this instrument to pass out of my keeping without taking its measurements, but I find the bore of a similar instrument, made by the same firm at about the same time, to be as follows:

 

Diameter.
Inch.

Diameter.
mm

Head-joint,

74

18.8
Upper end of the second joint,

74

18.8
Lower end of the second joint,

59

15.0
Upper end of the third joint,

61

15.5
Lower end of the third joint,

54

13.7
Upper end of the foot-joint,

52

13.2
Open end,

44

11.2

[I've added a second column for those who prefer to think in metric.  Figures are rounded to not exaggerate the precision.]

[Readers who have been following the flute strangulation series will be interested to note that the left hand section starts with the same bore diameter as the head.  Presumably, in the 60 or so years since the flute had been made, not enough compression had set in to reduce the top end bore diameter.  But there might be some signs of strangulation in the middle tenon.  Note that the bore at the bottom of the LH is smaller than the start of the RH section by about 0.5mm.]

At a distance of one inch from the open end [i.e. the end of the foot], the bore becomes cylindrical. In other respects the flute closely resembles that described in 540, excepting that the b, g and f# holes are .02 inch [0.5mm] smaller than the corresponding holes of the Nicholson flute, and that the distance between the e and f# holes is .16 inch [4mm] greater, the e hole being bored perpendicularly.

This instrument was kindly lent to me by Messrs. Keith, Prowse and Co.


Conclusions

Interesting stuff!  We learnt that Meantone is not an answer to the intonation problems we see in old flutes.  And Rockstro might have given us a few clues about early signs of strangulation.
 


Back to McGee-flutes Index page...

  Created 24 May 2011