Historical Flute Syndromes




If you've come here from the series on Reel Time Tuning Analysis, you may well be looking at a plot of your own flute and wondering what it means.  I thought a look at some historical examples might be handy way of exploring what you might be looking at.

Or if you've just wandered in here off the street, you might like to know what sort of systemic failings we find in the tunings of old flutes, and what therefore to watch out for in new ones.

What do we want from a flute?

Now, before we get to examine flute tuning, we need to consider what our musical context is. Probably our most stringent requirement is to be able to play with fixed instruments - ones that cannot tune to us and therefore require us to tune to them.  Accordions, concertinas and keyboards are probably the best examples.  Stringed instruments like guitars, banjos and fiddles do have some capacity to retune, although you'd get a strange look from the guitarist if you asked them to tune oddly just to suit you.  So we can reasonably toss them into the same category as the accordions, etc.  Coincidentally, these are also all very authoritative-sounding instruments.  If it came down to the audience judging who is the one that's out of tune, I'd say they'd be looking at you.

So, let's look at a RTTA analysis of an Irish box player:

Note that the tune being played didn't include notes C5 and G5, hence the tall thin error boxes on those notes.  Let's disregard those.  But we do see just how well this box is tuned in equal temperament.  Note everything is a little sharp, no doubt it's tuned to the fashionable A442, rather than the international standard A440.  As you can see, it also suggests Tartini-R is working very well - detecting the underlying tuning of the instrument while playing tunes.

So, like it or not, if you're going to a session with a box player, them's the notes you're gunna hafta play.

A Level Playing Field?

But what happens if you are just playing alone?  There is some evidence that, in the absence of other instruments, we actually prefer our scales slightly stretched.  That would make it permissible for the right hand end of our graph to tilt upwards a little, and the left down.  It would certainly make it unsatisfying to trend the other way.  And the key word is slightly.  A much stronger instinct is to avoid beating between notes - we have to be able to lip back into tune with fixed pitch instruments on demand.

How far out is out?

Not an easy question to answer, but perhaps one we will be better equipped to answer after a bit of RTTA experience. 

It's easier to say what isn't a problem.  If two instruments play A, but one is 440 and the other 442, we'll get a two-per-second beat rate between them.  Wah, wah, wah, wah, - no big deal, especially as it's rare that you play notes long enough to hear a half-second beat.  So an 8 cents error is not a problem.  Play 2nd octave A though (880Hz and 884Hz), and you'll get twice the beat rate, so expect to be more critical up there.

A 4 Hz difference, 16 cents, will give us 4 beats per second, or DRRRRRRRR.  But I think you'll find you can lip that distance without too much trouble, so I'd expect few problems in practice.

But when we come to deviations much over 20 cents, I think most of us will have trouble lipping reliably, especially at speed.  But, that's something for us to find out!

Meet the patients...

Here's a very typical first half 19th century flute, this one by Metzler.  It's similar to other large-hole, London-made flutes by Nicholson, Rudall and Rose and other makers of the time.  I've chosen it for the example as it displays many of the common tuning anomalies of the period.

And here's the tuning of the Metzler, as determined by RTTA from my playing.  I tried not to influence the tuning, just playing a few relaxed tunes.

Hmmm, rather different story, eh?  You can see that if I wanted to play this flute with a fixed pitch instrument, I'd need to put quite a bit of extra work in, pulling down the E5, G5, B5, and lipping up the D4 and F#4 at least. 

Could anyone play this flute accurately in tune?  I frankly don't know - if you're passing this way, do feel free to drop in and try!  It reminds me of the words of Mr Nicholson, who ushered in the era this flute was made in and whose own flutes tuning set the pattern we see above:

"...and it was generally asserted that I was the only person who could play in tune on a flute with large holes". 

Perhaps they were right!


Now let's compare that with a flute designed after the middle of the 19th century, R.S.Pratten's Perfected:

By this time (early 1850's), Boehm had brought out his cylindrical flute, based on scientific principles.  His flute proved that flutes could be (and therefore should be) in tune, so anyone wanting to introduce a new flute now had a certain standard to aim at.  Pratten also had the advantage of the work done by Siccama, who was clearly (from both his flutes and his writing) a tuning fetishist.  OK, OK, like someone else you know.

Note, with the exception of C5 and C#5, the notes all cluster within about +/-16 cents.  And in regard to C5, I need to confess I habitually used oxx ooo, whereas I should have remembered Prattens recommended oxo xxx (if you weren't using the c-key).  Indeed I prefer oxo ooo on the Pratten's original.

(Historical Aside.  We have to be aware that I'm doing something a little anachronistic here.  I'm comparing flutes made in the 19th century to a pitch standard set in the 20th century and still in use now in the 21st.  The Metzler was made in a period when domestic pianos were tuned to 430Hz, and this partially explains the tilted results.  The Prattens' was made in a period when professional gigs were conducted at 453Hz, although we don't see the medians sloping down to the right.  So it's not fair to criticise George or Sid because their flutes weren't made for A440 - it hadn't been invented by then.  But our results do indicate what happens when we try to play their flutes in today's musical climate.)

Recognisable Flute Syndromes

The charts above might look like random jumbles of pitches, but we can actually identify some common syndromes that you might look for in flutes of your own.  These are my own names for them, but feel free to use them!

Flat foot

We really need to see the other foot notes, C4, C#4, Eb4 as well as the D4, but trust me, the Metzler suffers from an impressive case of flat foot.  We can define it as that the foot notes are significantly flat compared to the average of the body notes.  Because the octaves of those foot notes are not flat, we can dismiss any suggestion that this is the product of an intended temperament.  Not really evident on the Prattens, although note that the D5 is actually sneaking a bit sharp.

Low Octave Tilt

Take a line averaging the notes D4 to B4 on the Metzler (we'll come back to the C's) and you'll see it's far from horizontal.  The scale of this flute (the distance between the holes) is too long for easy playing at modern pitch.  Get a note at the top of the flute in tune and the notes further down get progressively flatter.  No tilt evident on the Prattens.

(The Metzler has a C#-D# length of 253mm, which I would normally regard as suitable for playing at around 430Hz, the domestic pitch in the period it was made.  The Prattens C#-D# length is 245mm.)

2nd Octave Tilt

If the low octave is tilted, we can expect the second to be as well, usually not quite so much.  Draw the line from D5 to B5 on the Metzler.  Not disappointed, tilted.  Pratten pretty good.

2nd Octave Sharp

Compare the average pitch of notes D4 to B4 with those from D5 to B5.  The second octave set are sharper than the first octave in both flutes although less on the Pratten's.  So not only do we have tilt within the octaves on the Metzler, but also between them.

Wide Octaves

Apart from the general sharpness of the second octave, we note that some octaves, eg E4-E5 and F#4-F#5, are spectacularly wide on the Metzler.  This goes beyond being a tuning inconvenience; it starts to introduce instabilities and tonal problems as the harmonics struggle to reinforce each other harmoniously.  Tone and responsiveness suffers.

Top note droop

This is one of the few syndromes the Metzler does not show, but it happens elsewhere, so good to be on the lookout for it.  The top notes of the second octave, A5 and B5, sometimes tend flat of their first octave counterparts, with the B5 worse than the A5.  There's a touch of it in the top note of the Pratten's, even more noticeable when you investigate C6 (not on this chart).

A4 not centred

Again, not evident in the Metzler, but it is in the Prattens.  The A4, rather than being conveniently centred among the spread of the other notes, is close to one extreme.  So, when you tune to A4, you are actually tuning the rest of the flute sharp or flat.  Our analysis system permits us to pick a better note to tune to.  In the case of the Metzler, D5 would be a slightly better choice, although there isn't much in it.  G4 or B4 would be a definitely better choice on the Pratten's.

Other syndromes?

The syndromes above are those that occurred to me as I write, but there may be others.  If you see other recurrent patterns, let me know! 

Does it have to be like this?

No, these are all historical syndromes that we can deal with in modern flutes, providing we can detect them (not difficult in the case of the Metzler!).  And in many cases, these are also syndromes we can much reduce by retuning original era flutes or modern flutes that have copied their failings.  In the past, we've been limited by having only static measurement systems to go by.  If RTTA lives up to its early promise, we will be able to measure the flute's performance dynamically - as we play - and with greater precision than with the old static methods.  That should enable us to detect any remanent flaws and rid ourselves of remaining vestiges of these old syndromes.

Problem Notes

We do have to recognise that it is very hard to get all the notes on an Irish flute in perfect tune.  And for good reason - the maker can't put the hole where it needs to go, as our hands simply can't stretch that far.  The usual culprits are F#, cross-fingered C5 and C#5.  F# and C#5 are usually flat, and C5 often tends sharp, although in the case above, it's tending flat.  The keyed C note should be fine of course, but was not tested in these runs.

What about player syndromes?

We've talked above about flute syndromes - the kind of systemic errors built into flutes old and not so old.  But what about systemic player errors?  That's a bit trickier for me to comment on - I have a pile of flutes at my disposal, but not players.  It might be a task for someone else to carry out.  Take one well-tuned flute and one badly tuned flute, and a bunch of players of various skill levels and find out what we can.  Important work.  Speak to me.


I hope I've managed to stimulate some thought, and some suspicion towards your flute and your playing, and some resolve to delve further into it.  I'm always happy to talk flutes, so if something is bothering you, get in touch.

"Believe me, my young friend, there is nothing absolute nothing half so much worth doing as simply messing about in flutes." 

(Ratty, from The Wind in the Willows.  At least that's what I think he said ...)

Back to McGee-flutes Index page...

Created 18 April 2008, last edited 23 April 2008.