Very interesting! I didn’t know that and I’ve not seen/heard Donald Lindsay’s whistles either. I’d like to know more about his whistle head solution for second octane tuning.
I should mention that a bunch of other whistle-makers do something similar to correct octave tuning. Goldie/Overton, Kerry, Lofgren, Mazur, Karavaev, and various other Overton-style whistles all have a noticeable “pinch” in the head that creates a slight outward taper right after the air hits the blade. But all these whistles have straight windways. The taper is naturally integrated into their design.
The reason I mentioned Donald Lindsay is that he makes curved-windway whistles (similar in shape to most wooden whistles), but he intentionally adds a similar taper on the inside of the windway to correct octave tuning. Unlike on Overton-style whistles, you can’t see the taper from the outside; you have to look down the tube to see it.
I could be mistaken, but I believe MK does something similar with his low whistles.
This “outward taper in the head” strategy is exactly how Boehm-system silver flutes have corrected octave tuning despite having a cylindrical bore. It’s also how Doug Tipple makes his cylindrical-bore Irish flutes with corrected octave tuning.
OK, back to the “recalibrated” Clarke. It came while I was out of town and I picked it up from the post office today. It’s a nice whistle. Definitely less breathy, a bit more of a solid tone…it sounds nicer to me than any tweaked Clarke I’ve played (they always sound a bit tamped down to me). Is it worth the cost? IDK…that’s a personal call. I’m definitely going to play the heck out of it for a month or two and see what I think. I’ve had several whistles I loved initially but ended up going back to my O’Brien Rover and my Water Weasel (I love those two whistles). This one has a chance because I’ve always liked Clarkes and truly appreciate the as close to perfect as possible tuning of a conical whistle.
Eric
1 Like
“This one has a chance because I’ve always liked Clarkes and truly appreciate the as close to perfect as possible tuning of a conical whistle.”
Interesting Eric. I’ve played the older style of Clarkes whistles, in D and C, and didn’t find the tuning all that good. I’m aware that there are very different styles of tapered whistles, eg:
- bold, full length cones like the Clarkes,
- whistles with shallow cones, and
- whistles more like conical flutes, with cylindrical upper sections, leading to bold downward taper in the finger hole area, and finally ending with a flare.
There may well be other entomologies!
Of these, I have found groups 2 and 3 to have better tuning than group 1. But maybe it’s a blowing thing?
Hi Terry - I ran through a bunch of tunes this morning with my tuner out, and it’s really, really in tune in both octaves. One thing I had forgotten, though, is that you really can’t half-hole well on a Clarke, so it’s truly a diatonic instrument. It’s so easy to half hole on the O’Brien and the Water Weasel. I tend to play tunes that utilize a lot of sharps and flats on my flute anyway, so it’s not an issue for me but definitely could be an issue for others.
Eric
That’s a shame. Are its holes too small for half-holing? And how are the cross-fingered C-naturals?
Will you consider recording sound samples for it eventually? I’m sure we’d all love to hear how it sounds if you do!
Yep, the holes are too small…like any Clarke, it’s a pretty extreme conical form. The Cnat at oxx ooo is a tad muffled as is the Cnat at oxo xxx. I don’t use any other Cnat fingerings.
I can put up a sound sample later…I think. With the new board, I’ll have to figure that out and whether or not I have to have it hosted somewhere.
Eric
I don’t mean to be pedantic, but I’d say that Clarkes are acoustically more in group 3 than in group 1, although perhaps not exactly either the way you described it.
If you look closely at a Clarke whistle you’ll see that the top section has a squared-off profile. Squaring off a tube like this reduces the cross-sectional area compared to what it would have if the profile was circular. So there is a reduction in taper in the head of a Clarke whistle, and this will have some impact on tuning.
Where did you purchase it from? I noticed Clarke is encouraging people in the U.S. to purchase from their U.S. store, but I don’t see the recalibrated model available there.
Hmmm, not so sure about that. Wouldn’t squaring off the tube simply squish the circular cross section created when the cone is first rolled, into a square cross section of the same area?
And the squishing doesn’t extend very far, unlike the typical Group 3 whistles I have in mind, where the cylindrical bore extends from the window down to not far above the first finger hole.
Thinking again about the use of the term “recalibrate”. Fiddling the wind way and ramp to reduce air noise and perhaps wasted air sounds like good improvements. But I wouldn’t think of it as a “recalibration”. More like a “refinement” perhaps?
But given Jayhawk’s advice that the tuning is very good, I wonder if that’s what has been “recalibrated”? Different taper? Finger holes relocated, resized or both? Other, please specify?
I guess we need someone with access to both to compare them for us. Or take some measurements and post some numbers so we can compare with older forms.
The area of a circle is always larger than the area of a square or rectangle with the same length perimeter. A circle is the optimal shape for maximizing the area enclosed by a given length perimeter. We can get into the math if you like, or you can look it up if you don’t believe me.
Compressing a tube from a round cross sectional shape to a square one does not stretch the material, it compresses it. So the perimeter of the resulting square is certainly no longer than the perimeter of the circle it started from. Likely, it is a bit shorter, compounding the reduction in cross sectional area.
So, the cross sectional area of the bore is definitely reduced when a cylinder is compressed into a square shape. Several whistle makers, including Colin Goldie, do this, and they compress a short section of the bore around and below the window, consistent with what you see in Clarke whistles.
So, rather than sit here and wonder, I pulled out my scarcely used Clarke Sweetone (in D) and tried it into a tuner. Didn’t spend long on it, but the tuning seemed OK in the second octave but the top of the lower octave dropped back 25 cents or so. But that could be me not it! It’s very different to my favoured whistle which is of the Group 3 type.
Here is a little rudimentary math, crudely expressed in text, regarding the cross sectional area of cylindrical bores vs square bores. Please correct me if I got any of this wrong … its been a while.
Area of a circle = pi * r^2
Circumference of circle = pi * r * 2
A square with perimeter = pi * r * 2 has an area of ((pi * r * 2) / 4)^2
The area of this square, divided by the area of the circle with the same length perimeter is:
((pi * r * 2) / 4)^2 / (pi * r^2) = pi/4 = 0.7854
So, the cross sectional area of a cylinder which has been squashed into a square profile will be at most 78.5% of what it was previously, provided that the squashing doesn’t actually shorten the perimeter, which it likely will.
Or looking at the same thing the other way around, you can say that the cylindrical bore has a 27% larger cross sectional area than the square bore formed by squashing it.
(4/pi) = 1.273
My take away from this is that squashing a cylindrical bore into a square profile has a very large impact on its cross sectional area, which is what matters acoustically.
The Clarke Sweetone whistles do not have the squared off upper section. Only the classic Clarkes have it. But of course, I can’t easily see what is going on inside the plastic head on the Sweetone.
I will concede I think your maths is correct. Whether it’s enough of a reduction to make it acoustically equivalent to a cylinder in the region between window and a bit above the first finger hole is hard to guess, especially as the square fades to circular over about one third of the distance. Modelling might give us a clue, but maybe I’m not THAT interested!
I had a quick peek into mine. The conical body comes up and becomes roughly cylindrical for the last 13.5mm or so. This is so it will plug into the cylindrical socket in the plastic head.
The bore of the body tube is about 15.7mm (a bit hard to measure as it’s not perfectly circular). The bore of the plastic head beyond the socket area is slightly less, about 14.7mm. That reduction might be just to make sure the body can’t be pushed beyond the bottom of the socket. Or it might be prompted by acoustic considerations.
The flattened section of the Clarke whistles I have here (I have 2 in D) extends between 20 and 30 mm from the block. They are not the same as each other. On one it is 20 mm and on the other it is 30. This works out at roughly 10% of the overall bore length.
The squaring off does peter out as the whistle narrows, so it does seem like the cross sectional area might be more or less constant, and certainly more constant than it would be if the cone continued unmodified.
In any case, it does seem like exactly the kind of thing you would do when designing a bore to improve intonation, i.e., constant cross sectional area near the head, transitioning to a taper in the body. And while the length of the non-tapering (or reduced tapering) part is less, as a percentage of overall bore length, than it would be in a flute, that also seems to be the case for the difference between cylindrical bore whistles and flutes. The whistles that are corrected at the head have shorter modified areas than flutes do, as a percentage of their bore length.
I think we can get an approximate handle on this without resorting to modeling.
The fully squared off area on my Clarke is roughly 14.6 mm on each side OD.
The wall thickness is 0.33 mm.
So the square bore has a cross sectional area of (14.6 - 0.66)^2 = 194.3 mm sq
If we divide this by pi and take the square root we get the radius of the circle that would have the same area, and if we multiply that by two, we get the bore diameter for an equivalent cylindrical bore.
That bore diameter comes out at 15.7 mm.
In other words, if this square bore were a cylinder of equivalent acoustic size, its diameter would be 15.7 mm. This is at the top end of the bore at the block.
Now, when I measure the OD of the whistle at the point where the squared off section ends and the cross sectional shape becomes circular, I find the OD to be about 16.3 mm.
When I subtract the wall thickness (0.33 x 2) I get 15.64 for the bore diameter at the point it becomes circular.
These two points are about 30 mm apart on this whistle, one at the block and the other 30 mm down towards the foot.
I’d say that, given the margin for error in my measurements, the 15.64 mm measurement is close enough to the 15.7 mm measurement for us to conclude that the cross sectional area likely does remain more or less constant over that head section, as it would if it were a cylinder of bore diameter approximately 15.7 mm.
So, its quite interesting to see that your measurement of the bore of your Clarke Sweetone whistle is also 15.7 mm !
Clearly there was some design continuity going on there. Although it does seem that they shortened the cylindrical section by close to a factor of 2. I expect that would have some impact on the tuning differences between the two models.
But the difference in my Sweetone is that the 15.7mm bore is the top of the full-length body cone. There is about 13.5mm of cylinder there but that’s the male tuning slide which has to fit snugly in the plastic head socket.
Perhaps what we need is someone to do a YouTube video comparing the various models of Clarke’s and highlighting any signs of “recalibration”. We could then try to see what physical changes might have lead to them.
Jayhawk. Any initial impressions?