The Great Flat-Foot Experiment

 

Introduction

English flutes from the first half of the 19th century usually exhibit an unusual tuning characteristic - the foot notes (C, C#, D and Eb) are remarkably flat compared to the rest of the instrument.  

There have been a number of theories advanced to explain the phenomenon.  They include:

  • that these notes were made purposefully flat to assist certain notes in the third octave.  The certain notes are usually unspecified, although F#6 is sometimes implicated
  • that an intonation other than equal temperament was intended.  Popular suggestions are "meantone" and "just".
  • that the low notes were made purposely flat to encourage the player to put more effort into them.  

An opportunity came up to look more closely at the issue ....


The case study

The case study was Clementi C. Nicholson's Improved No 315, sent to me for some crack repairs and for attention to the intonation.  As will be seen later, the intonation problem of concern was the classic flat foot problem.  Remembering that it was Nicholson who started the rush for better intonation, this flute is an interesting one to examine.

Rather than hack into the existing foot, rendering it no longer of historical interest and probably of compromised intonation, the owner and I agreed that a new foot would be constructed.  The new foot would be designed to complement the remainder of the flute, not pretend to be the original.  Just to ensure no future historians are deceived, we also agreed to leave copious visual clues that the foot is a replacement.  Chief among these will be my maker's stamp.

Because the owner was primarily interested in playing Irish music, the new foot need only manage Eb and D satisfactorily; I chose however to make a foot that, had it been fully keyed, would also do the right thing by C and C#.  That enabled a fair analysis of the effects of a non-flat foot on the remainder of the instrument.

Design and construction of the foot presented no real challenges, and as will be seen, certainly sorted out the flat foot problem.  The effect on the flute is substantial, considerably improving its low end response as well as the intonation.

The flute with its new foot


Effects of the new foot

The graph below enables us to see the effects of the new foot. The response of the old foot is shown in navy blue.  The red trace illustrates the response with the new foot.  The yellow trace shows the effect of some different fingerings using the new foot.

A-natural notes on this flute tend sharp (a common feature of many flutes of this time), so the bottom octave G was selected instead as our reference point. The vertical units are cents (100 cents = 1 semitone).  It is difficult to ensure accuracy better than about 10 cents in making these tests, so ignore any variations below this limit.  We're out to fry the big fish.


Low register

The most significant effects can be noted in the low register, where the foot notes (C4 to Eb4) were about 30 cents flat and are now fine.  There appear to be no other significant effects in the low register.

We can see however a clear argument against the suggestion that the low notes were flattened in order to encourage the player to put extra effort into them.  If that were the case, we would expect to see a gradual and progressive flattening from the middle of the octave.  Yet E, the first note above the foot, is one of the sharpest notes in the register.  Indeed the step between E and Eb is an extraordinary 1.5 semitones!

These results were obtained using the standard fingerings of the time, which included holding the Eb key open for most notes, including E.  In this case, a more accurate E will be had by closing Eb, although that gives a more veiled note.

Incidentally, it should be remembered here that these tests are being conducted around modern pitch (440Hz).  To achieve that pitch, this player needed to extend the head slide to around 23mm.  Choosing to use the flute at 440Hz already reduced the degree of flat-footedness from around 55 cents to around 30 cents.


Middle Register

Again the only significant effects are to the first octaves of the foot notes - D5 and Eb5.  These had been substantially flat, particularly in comparison to their surrounding notes C#5 and E5.  They are now quite conveniently placed.


Top Register

The graph illustrates that three notes in the third octave appear to be affected by the new foot - Eb6 sharpened by 15 cents, E6 by 10 cents and G#6 by 25 cents.  Even so, they remain within the deviations noticeable elsewhere in the instrument.  So, not exactly earth-shattering.  Still, worthy of some further examination.

Eb6 had previously responded best to xxx G# oxx Eb.  With Eb4 now at correct pitch, the extra sharpening was no longer necessary.  The best fingering is now xxx xxx Eb, either with or without the G# open.  Note that the new fingering is essentially the bottom octave Eb fingering.

E6 still responds well to the original fingering xxo oxx eb, indeed just a little better, no doubt because of the sharpening of Eb4.

G#6 is the most striking change, with the original fingering oox oxo Eb driving distinctly sharp.  The fingering ooxG# oox now gives a better note.  Interestingly this is  closer to the fingering recommended later in the 19th century.


Conclusions

It certainly seems that any "effect on the third octave" can be discounted as a reason why flat foot was so prominent in flutes from the first half of the 19th century.  Although changes to three notes in the third octave could be detected, one of these was an improvement and the other two were not overwhelming and easily dealt with minor fingering changes.

The "put some effort into the low notes" theory also fails to impress.  It might have held water if the degree of flatness was an achievable 10 to 20 cents and had been progressive in distribution.

The "temperament other than equal theory" also doesn't look relevant in the light of the measurements.  A temperament is a way of breaking up the octave into its individual notes.  One expects the octaves still to be octaves however.  Compare the four footnotes to their octaves and double octaves (eg C4 to C5 and C6, etc).  Ghastly.  The issue is not within the octave, it's between octaves.


My own theory

My own theory is that flat foot was an inheritance from baroque flute times, and that the subsequent development of the flute was a multistage process, going something like this:

  • baroque or one-key flute.  Small and relatively uniform holes give the highly compromised tuning needed to permit chromaticism without keys.  Flat foot evident, but not problematic, perhaps even important to the techniques needed to secure chromaticism.
  • first generation multi-key flutes essentially the same as baroque flutes but with keys added, mostly used for trills.  Flat foot still evident, not desirable but not too problematic
  • Nicholson the elder enlarges the fingerholes giving far better performance at large, and improved intonation over most of the range.  For some reason, attention was not given at that time to flat foot syndrome, which now stood out in far greater relief.
  • Flat foot syndrome chipped away at over the subsequent years, vanishing around mid century, probably with Pratten.

If I'm right, we may be looking in the wrong place for the source of flat foot syndrome - we need to focus on baroque flutes.  A number of questions still arise:

  • why did flat foot exist in baroque flutes - was it a feature or a fault at that time?  (It would be interesting to construct a non-flat foot for a baroque original to test the theory)
  • why didn't Nicholson or his contemporaries do something about it in the second generation 8-key period?

Your comments and suggestions welcomed!


Footnote

No attempt was made to experiment with stopper position during these tests (the stopper was set at the nominal one-bore diameter distance from the centre of the embouchure).  It is probable that the general minor flatness of the third octave could be ameliorated by this means.


Achnowledgement

Thank you to US fluteplayer, Terry Briley, for enabling and permitting this study.

 

 

Back to McGee-Flutes home page