Why does this tweak work?

I am quite excited (probably overly much so) about a tweak I have just made to my Dixon three-piece low D.

The low end noted (esp D & E) were not as strong as some like, so I copied Copelands idea of placing a wall around the window (see interview with copeland).

I think it strengthens the low notes, and this was corroborated last night by a whistler with more experience than I. I forgot to ask his permission to quote him, so I wont name him. His opinion was that it strengthened the low notes and reduced the volume of the top end.

My prototype is poorly made from wood (my favourite material for craftwork) and held on with rubber bands and blue tack (aka poster putty?). It needs to be sealed better to the whistle, so I will try again and, rather than mess up the first version, I will start afresh. This gives me an opportunity to try some variations. Now, I do not understand at all the physics of the wind wall (well any acoustic physics really), and was hoping maybe that some readers who do understand such things might have opinions/advice on the relative dimensions and such.

For example what should the wind wall height be? How far past the end of the window should the windwall go? I have done my best to have the windwall fit exactly round the top and sides of the window, but should it be set back a bit? Should the walls be ‘vertical’?

Also, are there any ways of making objective measurements of my results? So that I do not fool myself into believing that I am achieving more than I am?

Thanks in advance for any help.

Hey DrPhil,

It’s a damn good question!

Mostly you see this kind of construction on pipe-organs. So there has to be some history to it - possibly some darkly secret math as well!

Maybe even, some of our resident organ-pipe-makers might divulge a little teasingly invocative hints??

From first blush, I’d say that the linear chanelling reduces the energy loss around the interface between the oscillating air column and the ambient air conditions. But I’d hazard that the frequency dispersion will be a lot more directional.

Essentially, sound is spherical - the waves radiate outwards from the source much as is seen when a stone is cast into a still pond. However, the sound source created by a whistle is definitely not as simple as that - the blade will tend to project the shock-waves outward acording to the blade topology. If you further focus that with walls, I would guess that certain frequencies would be accentuated and others diminished depending on where the listener was. If you have a wave anyliser - say an oscilliscope or the digital equivalent, you will easily see how the harmonics are differently perceived by turning the whistle to different angles from the mic.

I’d only ask this:
Does the sound seem a little less breathy when using the walls?

Thanks for the quick response - and some things I hadn’t thought of but need to…
I wonder if there is a free sound analyser on the web?

Definitely.

Hmm - grab G-Tune = google it.

It’s a bit buggy but mostly functional - good osciliscope and frequency analyser.

I guess I can weigh in on this one, as an organ-builder in residence - maybe there are others, too?

According to correspondence I’ve had with a handful of whistle makers in the past Michael Copeland being one of them, some whistle makers have studied, in various ways, organ pipes and applied their observations with whistle construction - as is mentioned in the interview with Mr. Copeland that DrPhill noted.

The terminology is a bit different – blade or ramp = upper lip, block = languid, etc., but other than that, a whistle is a the same as an organ pipe, it just has to make play more than one pitch - which in my career voicing pipes makes it quite mysterious indeed. I’ve no need to try it - at least efforts that would be made public!

Anyway, the ‘walls’ are called ears on an organ pipe, and much like if you hold your hands in that position on either side of your mouth and sing or make a sustained tone, you can hear a focusing, of sorts, of the sound - as you would do when you are calling for someone. Ears aren’t a requirement, but really a matter of tonal preference of the voicer or tonal designer of the instrument. They do become a requirement when they hold a “beard” or “bridge” of some sort, of which there are many types and for various purposes.

The type of beard that Copeland uses is commonly called a box beard. Its most common use - and again there are always exceptions - is on organ pipes that are capped or closed at the top - commonly called Quintaton or Quintadena. Normally, the upper lip (blade) of a capped pipe is cut-up high enough to make a flutey sound, which is what capping the pipes does, harmonically. However on the pipes that use a box beard, the cut-up is lower than usual, and the pipe wants to overblow to the next harmonic. Positioning a box beard - for whatever reason based in physics - allows the pipe to speak, not overblow to the second (octave) harmonic completely, and helps emphasize the 3rd harmonic, which is an octave and a fifth - hence the name quintaton. They are very fussy beasts to voice, believe me!

If I had to guess at what the box beard is doing on a whistle, I would say that it is in order to get a strong bell note - I have a Copeland silver C here with ears and a beard. That would allow whistle maker to make the whistle a little “quicker”, that is the languid or block is a bit lower (or the upper lip is rolled out a bit more) which allows the top notes to speak more easily, but the result is that the lower notes also want to overblow easily. So, the ears and beard are there to keep the whistle from overblowing on the bottom end, and the effects of the ears and beard aren’t nearly so prominent in the top end - likely to the point of not really affecting the speech.

DrPhill, as for proportions of the ears and beard - there are no cut and dried rules. Box beards can be added after the ears as a separate piece that bridges the ears, it may or may not contact or conform to the lower lip, and it (and the ears) may be wide or narrow. Over the years I’ve worked with pipes from European and domestic builders and supply houses and everyone has their own set of proportions. I would say to experiment and do what works, but don’t get too caught up in what may or may not be ideal. If it works for you, then there you are!

Did that make any sense? There’s certainly a wealth of information on the web about this, though some of it I’ve read obviously wasn’t authored by a voicer, so it tends not to make much sense, or is perhaps too theoretical to be practical. As any instrument maker knows, and if there’s one thing I’ve learned after voicing thousands of pipes, there are always exceptions to the rules.

If I’m ever working in anyone’s area, you are more than welcome to drop by and I’ll be glad to give a lesson or two! If you can get your hands on some metal organ pipes, you will see quickly - and I will be delighted to assist - what I mean. That said, my hat is off to the whistle maker for taking what I would call a single-note pipe and working his magic and poking it full of holes so that it becomes a whistle!

Reg

i’m trying to follow… do you have a pic of the tweak to your dixon?

be well,

jim

rhulsey: Thanks - that is a wealth of information to digest, and some ‘correct’ terminology to search upon. I am currently working on v0.02… higher, longer ears, tighter fit to the window, better fit to the whistle body.

I was a little confused - the ears are the flaps to either side of the noise hole ( :laughing: ) but the ‘beard’ is the cross-wise member above the hole (closer to the blower)? I certainly followed the pictures of the Copeland as he seemed to be solving a similar problem to the one I was tackling, but my ears are longer than his. (Not that the Dixon is flawed btw, it just needs gentle blowing at the bottom of its range).

Jiminos: I can take a picture, but have not got round to selecting a picture-hosting site, so cant yet post one here. Tell you what - I’ll take a picture and stick it on mySpace later this evening, if that is good enough. Who knows I might even be able to blag the url and post the pic here.

Crossed posts with DrPhill, but I have a similar question: If the ears are the walls around the window, where is the beard? I agree that a picture might help, if one could be found. And btw, thanks for the detailed and interesting description of organ lore. Very cool.

Hi -

Sorry i’m not at home, or I’d provide photos a-plenty!

The ears are the walls that parallel the mouth (window in whistle-speak). The beard connects the two nearer the mouthpiece. Copeland seems to use a solid piece with two 90 degree bends, which would be easier I’d think than 2 ears and a separate beard.

Photos this evening, should you like.

Reggie

Lets see if I have this right - apologies if it does not work…

The tweak is a single block of hardwood, lovingly (read ‘somewhat inaccurately’) hand carved to shape.

Hope the pictures help.

Re: “beard”

Consider that an organ pipe is positioned with the mouthpiece towards the floor and the tube pointing up. In that position, the walls on the sides will be in the same position relative to the voicing window as ears are relative to a face. And the wall at the end of the voicing window will be below, in the same position relative to the window as a beard would be relative to a face.

Does that make sense?

Best wishes,
Jerry

Works for me - I had a suspicion that would be the answer, but I dont like to assume (assume = ass + u + me).

I think the tweak strengthened the low notes - what might I have lost? There must be a downside somewhere (apart from having an ugly bit of wood stuck to the whistle that is).

Jerry is correct, and all parts in the organ world are indeed anatomical… toe, foot, lower lip, languid (derived from French for ‘tongue’), mouth, upper lip, ears, beard, body… the windway is often referred to as the ‘flue’, which is the type of organ pipe we are talking about, the other being reed pipes.

Reg

One easy way to experiment with air dams, ears, etc. is to build them with a bit of blue tack putty or similar.

Another is to use an appropriately sized O-ring as an air dam positioned next to voicing window opposite the blade (the “beard” position). A ring in this position can be very effective in strengthening the lower end of the whistle, and can easily be left on with no permanent modification to the instrument.

Good suggestions, MTGuru, but

Yes, I thought of that, but only had small amounts of blue tack (robbed from another use!); I thought I could carve a chunk of wood quicker than I could walk to the nearest shop selling blue tack (ok, so I thought wrong, but what the hell, I enjoy carving wood). I also considered beeswax, but decided that a more robust solution would be best - if it worked. I was also worried that the acoustic properties of blue tack and beeswax would be inferior to a harder, denser, polished surface.

I tried that too, but none of my O-rings made a significant difference.

Version 0.02 is built. The ears and beard are twice as tall, and the effect is less marked. This is good, as I can slowly lower the ears and beard until I have gone just too far…

Thanks, Reg! Your whole post is most interesting, but I fail to understand the quoted paragraph above. What do you mean by having the languid or block “lower”? And what by “the upper lip rolled out a bit more”? The upper lip is the labium edge? Making the block lower meaning increasing the distance from the exit of the windway to the blade (labium edge)?

Edit: Okay, I think I understand it now, after reading this article:
http://www.pykett.org.uk/how_the_flue_pipe_speaks.htm

[edited to add image]

Thanks,
Hans

I suppose it’s stating the obvious that organ flue pipes sound the fundamental or bell note only. So it’s more straightforward to carefully adjust the voicing for that one note. Whereas on whistle, the voicing is always a compromise across the range of fingered pitches.

In general, and unscientifically, I find that an air dam at the windway end (beard) of the window strengthens and stabilizes the lower notes in the 1st register, and a dam at the blade end (like my Susato O-ring tweak) stabilizes the upper notes in the 2nd register. Dams along the side (ears) strengthen the stability across the range, but may overly suppress other desirable playing characteristics such as chirpy chiff and overblow response.

rhulsey, you might also want to mention something about notching and the effect on chiff. I’d be interested in knowing more about that!

In addition to outer ears, inner ears also work.
They do flatten the pitch slightly.

Mack Hoover had shared the “inner ear wall” with me some time ago, works great!

The outer ears were first introduced in French Organ pipes, (have a book on Organ pipes somewhere here at home), it focuses the air better. I’ve also seen back walls over the edge of the window opening and some of the really loud panic whistles have a cover way up above the window that protects it from being blanked out by the wind and increases the volume even more.

Made a three fipple whistle head once, loud it was but sure needed a lot of air.

Hans - I haven’t read the link you posted, but at a glance looks like a lot of information. Basically, the languid is the same function as the block. Its position determines how “quick” the pipe speaks. Too high and it won’t make a sound, too low and it “gulps” or overblows to the next harmonic - its octave. The languid is soldered around its edge during construction, then the body is soldered on top of that. That said, you are moving the languid so little as to not break the solder seam. Visually, you often do not see the languid move, the amount is so small between speaking and not… The same “quickening” effect can be achieved by positioning the upper lip, however, it is better, and yields a more consistent result, to begin with well-aligned pipework and use the languid for its intended purpose.

Obviously, whistles don’t have movable blocks once the whistle maker voices the instrument, so he relies on placement before he pins the block, if that is his practice, and/or the position of the ‘blade’ - including its thickness (having been thinned or skived from inside or out).

MTGuru, as for “notching” - or nicking - these are grooves that are placed in the leading edge of the languid, in varying depths and number depending on what you are after. This smooths out the speech of the pipe, removing what most refer to as “chiff” and extraneous noise or “sizzle” as some call it. Depending on the sound you want, the nicks can be nearly invisible as to be applied with an abrasive of some kind, a knife as small as an exacto, or something as large as a ‘nicker’ made especially for the job that might even nick the lower lip and languid at the same time - resulting is a tone as void of “sizzle” and chiff as possible. They are just another “tool” used in the voicing process, more likely seen in what you might call a “romantic” instrument, and less likely seen in baroque instruments.

Reg

I came up with my “Parks Tone Ring” while I was investigating walls on the window of some prototypes. I wanted to compare with and without configurations easily and made the ring. I had a “beard” in place already because of the mouthpiece standing proud of the whistle body. The ring can be slipped down the body to remove the walls. The walls do improve the low notes but do so at the expense of the highest notes. They sound more “dirty” or complex with the ring in place. Notes in the middle are less effected. The harmonics are effected across the range, producing a whistle of slightly purer tone without the ring in place. And of course, if one rotates the ring and closes off the window more and more, the sound get more and more breathy until it’s just a whisper. Given the design of my Every whistles the “beard” is always in place, so at times I have a beard without any ears.