New mouth piece for our Artist Series Irish Whistle – part II

In Part I we learned that creating or renewing a part of an instrument is quite tricky: the new design needs to solve a problem, it must be functional, it needs to add or improve something and it needs to look good! We are of the opinion that you can always improve things, because we (techniques, computers, our thinking, experience) are evolving, so our solutions must also evolve…, right? This is true for ‘the’ saxophone, and it is more than likely true for any windinstrument, since we are always dealing with a compromise between the laws of physics on one hand, and human perception on the other hand. The closer we can get the physics to the ‘ideal’ human perception, the more we are improving our instruments. Do we sound like tree-huggers? Read on, because we’ll get you wrapped around our tree.

In part II (what you are reading) we are looking at different parts of the Artist mouthpiece, which areas are improved (theoretically) and why. Part III is about the simulations we have done on the mouthpiece and its results. Part IV is all about 3D printing. Part V is about finishing a 3D part and testing it in practise. Part VI is all about the artists, and we’ll have some sound samples of musicians playing on our whistles (with sound…). Can’t wait to hear it? Here’s a sample on YouTube.

Part I – Introduction | Deel II – Building the Design

Written by Ruud Roelofsen (design, brand and 3D)

Building the new design

In part I we explained that by drawing a circle and extruding it, you can build a complete model (part). In the image below, you can see two cylinders on top of each other. The model is represented as a cross-section.

In this cross-section, two cylinders with different diameters are attached. The reason one cylinder has a different diameter than the other is because we do not want our body to be pushed all the way into the mouthpiece, so it acts as a stop. The distance between the fipple (the end of the blade) and the tone holes determine the intonation (along with the diameter and the shape of the tone hole), so we need a consistent build. This end stop will help us be consistent in our build.

And now, for the hardest part of the design; how do you go from a perfect circle (the beginning of the chamber) to an ovale shape (the beginning od the windway)? The answer: software. We are drawing a circle on the tube that we showed. Next, we are drawing a new plane, perpendicular and at a certain distance from the tube end. On that plane, we draw an ellipse. In addition, guide curves are drawn, to let the software know how we want to achieve the transition. Next, we are creating a lofted extrude from the circle to the ellipse, with the guide curves as the…yes, you guessed it; the guide.

In above picture you can see the guide curves (purple) and the initial circle (sketch 10) and the ellipse at the other end. On e guide curve is a straight line, the other is a spline (based on a number of variable curves).

Let’s see how it looks like when we extrude a solid from the circle to the ellips, using the guide curves:

By applying ‘zebra-stripes’ to the preview, we can check for consistency and symmetry. By turning the model, we can check every angle.

View from above. The model is completely symmetrical. In this model, you can also see that the actual mouthpiece is slightly smaller than the rest of the chamber.

By applying and confirming the preview, we create a solid. Next step is to create the cuts like the windway, the chamber, and the window, leaving room for the blade. However before we do this, some corrections are made to the outside of the model.

Next is the windway. By drawing a rectangle at the beginning of the mouth piece and extruding it over a length of 45.5mm, we create a cut that will mark the end of the windway.

Camber and blade

Next is shaping the chamber: In order to do this right, we need to take into account the thickness of the material, as we do not want to deal with anything less than 1mm thickness; we need to preserve the strength of the 3D printed model. In order to achieve this, we make use of our trusty guide lines again, slightly tilting the chamber downwards. Okay, we are aware of the fact that this may sound plain and stupid. However, we are convinced about the theory behind this. In the picture below, you can see the guide lines that determine the cut of the chamber from the windway to the end of the window.

When we turn the model, we can shape the window. We are going for an unconvetional trapezoid shape; the top is slightly wider than the bottom. The reasoning is two-fold here; we give both room and direction to the air stream. The result: a nicely round and breathy sound, while improving playability. You don’t have to believe us here, just read the reviews…

By now, you have probably noticed the odd shaped window; it has a trapezoid shape. This way, more air contributes to the tonal trigger, which, (in theory) makes the mouth piece easy to blow, creating a steady, nicely projecting sound. In the drawing below, you can see the actual dimensions of the window of a Low Whistle mouth piece. Downside is, because of this abnormal width, we could loose resistance in blowing, which could make it hard for musicians to produce long notes as they loose their breath. To compensate, we’ve added a new feature to the windway which we will explain later on in part III.

Next is the position and the shape of the blade.

Next is another innovation that we’ve incorporated; if you look at the way flute players are blowing into their flutes (slightly downward, as blowing into a bottle opening). it makes sense to apply this into a mouth piece design. This is way we have added a windway insert into our design.

This way, more air is blown into the mouth piece, rather than over the blade, resulting in the air contributing more to the tone.

In the next picture, we remove the ‘bump’ (light blue) to let the air freely move over the blade.

So here it is, our newly designed mouthpiece!


To really test the model, we need to adhere to our design principles; it must be thick enough to print (remember we need to finish it as well). Let’s do a test and set your reference thickness at 2mm. In the picture below, you can see anything that is of 2mm thickness is purple. Anything more than 2mm is light blue. The blade is obviously thinner, so more orange.

So this is Part II. Initially written in Dutch, I hope this translation makes sense. In part III, we take a close look at analysis and simulation. We’re also adding a body and tone holes to get close to reality (picture below).

Back to Part I | Part II – Design | Part III – Simulations (in Dutch, please be patient while we are translating

This post is also available in: Dutch