Chiff and Fipple Forums

The latest release of WIDesigner, wind instrument design software, has a number of improvements for modelling whistles and flutes, including:

• Simplified single-step calibrators that work with either min/max frequencies or measured playing frequencies.
• Optimizer for setting the stopper position on a flute.
• More powerful optimizers for the bore profile of flute head-joints and flute bodies.
• Enhanced tonehole model, with an allowance for the presence of a finger closing the hole.

This new release, available at https://github.com/edwardkort/WWIDesign ... ses/latest, should be installed as a new, full installation. Because of the changes to the tonehole model we recommend you re-calibrate instruments you already have on file.

Some of the changes arose in the course of C&F threads on PVC flute temperament, Fixing octave tuning for a Killarney whistle, and complex bore of R&R flutes. My sincere thanks to Terry, paddler, Geoffrey, Andro, Hans, and the other C&F members who contributed their ideas.

Great, Tunborough, I look forward to getting in amongst it when I can. Unfortunately, I'm very pinned down by the massive carillon upgrade I'm consulting on, which looks like keeping me out of mischief until about Christmas. Last big opportunity for others to get ahead of me here, nudge, nudge....

I had a look at the WID. Pretty interesting. Very complete. Odd interface, but I'm a tech-guy, so I eventually figured it out.

I had a thought regarding all the confusion about whether a flute is A440 or A435 or whatever.

Is it possible or easy to take measurements from a random flute, plug them into WID, and determine the optimum flute A-frequency?

Secondly, is it possible to use measurements from a (high-resolution) photograph to determine the flute tuning? Maybe you have to make some assumptions about things you can't observe, like bore profile.

Sounding length is often given, as a measure from the embouchure, but that doesn't account for tuning slide. I didn't notice if there is a tuning slide parameter in the WID, or if that is just a manual entry from the hole positions.

I also wonder about which parameters are more sensitive to mis-measurement, vs which ones are relatively insensitive. If you have to make assumptions (because maybe you don't have the flute in your workbench), or are imprecise with certain measurements, then which ones can you let slide a little. (I'm remembering partial differential equations from college, maybe.)

Sorry about the odd interface. WIDesigner uses a lot of parameters for it's calculations, and they maybe aren't organized as well as they could be.

At present, WIDesigner doesn't try to isolate a tuning slide. For now, it's best to set the tuning slide to an intermediate position when taking all the tuning and geometry measurements. We might try to add a tuning slide parameter, if we can do it without making the interface even more challenging.

If you work up a model of a flute and an A440 tuning file, and WIDesigner predicts most of the notes will be around 20 cents flat, then you probably have an A435 flute in front of you. The net error at the bottom of the Deviation column in the tuning table gives a rough indication in this case.

tstermitz wrote:Is it possible to use measurements from a (high-resolution) photograph to determine the flute tuning? Maybe you have to make some assumptions about things you can't observe, like bore profile. ... Which parameters are more sensitive to mis-measurement, vs which ones are relatively insensitive. If you have to make assumptions (because maybe you don't have the flute in your workbench), or are imprecise with certain measurements, then which ones can you let slide a little.

Good questions, and closely related. There are some parameters that are impossible to measure precisely, so you have to make your best guess. If you try it first on flutes for which you can measure geometry and tuning, you should get a sense of where best to set things like the blowing level, airstream length, airstream height, and beta factor. I'd suggest you start with blowing level 9, airstream height 1 to 2 mm, beta factor 0.4, and an airstream length a bit smaller than the width of the embouchure hole. I'm not sure there are any parameters that are absolutely crucial to measure precisely.