WIDesigner release 2.4.0 now available

Oxymorons "R" Us
Forum rules
Please see forum announcement for description and rules!
tstermitz
Posts: 527
Joined: Tue May 26, 2015 10:18 pm
antispam: No
Please enter the next number in sequence: 8

Re: WIDesigner release 2.4.0 now available

Post by tstermitz »

I successfully got it to run on Mac with Mojave.

I don't recall if I had to jump through any hoops. Maybe run it in the terminal window?
User avatar
MadmanWithaWhistle
Posts: 697
Joined: Fri Jan 27, 2012 10:15 pm
antispam: No
Please enter the next number in sequence: 8
Tell us something.: Instrument maker and researcher exploring new methods of creating traditional instruments with longevity aforethought. Player of the whistle, flute, and continental European border pipes.

Re: WIDesigner release 2.4.0 now available

Post by MadmanWithaWhistle »

I'm trying to use this software to help keep the voicing similar across my instruments, but I can't find any documentation beyond the videos for NAF. How does one use this to model whistles?
Tunborough
Posts: 1419
Joined: Sun Dec 05, 2010 2:59 pm
antispam: No
Please enter the next number in sequence: 10
Location: Southwestern Ontario

Re: WIDesigner release 2.4.0 now available

Post by Tunborough »

The wiki pages provide more extensive written documentation. Using WIDesigner gives a brief introduction, with hints for starting out and links to other documentation pages. Working With the Whistle Study Model explains how to calibrate a whistle model, and other pages discuss the tools for evaluating an instrument design, and optimizing a whistle design. Let me know if you need documentation for other aspects.
MadmanWithaWhistle wrote:I'm trying to use this software to help keep the voicing similar across my instruments
At the moment, WIDesigner deals only with tuning, and can't say much about tone colour or playability. To me, "voicing" is about shaping the windway and window specifically for tone colour and playability, so WIDesigner may be of limited help with that.
User avatar
MadmanWithaWhistle
Posts: 697
Joined: Fri Jan 27, 2012 10:15 pm
antispam: No
Please enter the next number in sequence: 8
Tell us something.: Instrument maker and researcher exploring new methods of creating traditional instruments with longevity aforethought. Player of the whistle, flute, and continental European border pipes.

Re: WIDesigner release 2.4.0 now available

Post by MadmanWithaWhistle »

Hi Tunborough, thanks for getting back to me - I'm unable to find anything beyond the brief wiki articles, which don't describe things like 'fipple factor.' In the Feadog Mark 1 study, it's blank, and it keeps returning the error, "fipple factor, if specified, must be positive." Is there lengthier documentation anywhere?

Ah, by voicing I meant only what this tool can asses - the tonehole optimization. I was hoping to create and validate through experimentation a preferred design (i.e. larger holes, or flatter Cnat, etc) and from that derive a correction factor I can apply to optimization schemes in other keys.

At any rate, Bracker's calculator (which has worked well for me in the past) seems to hate this new headjoint design, and I'm having a hell of a time trying to produce a well-tuned body tube.
Tunborough
Posts: 1419
Joined: Sun Dec 05, 2010 2:59 pm
antispam: No
Please enter the next number in sequence: 10
Location: Southwestern Ontario

Re: WIDesigner release 2.4.0 now available

Post by Tunborough »

Mention of "fipple factor" tells me your WIDesigner options are still set to use the NAF study model, when you want to be using the whistle study model. From the menu bar, select Edit -> Options... to bring up the Options dialog. On the General Study Options tab, select Whistle Study instead of NAF Study. While you're there, you may want to check the values for other options. For whistles, I generally use a temperature of 27 C and 100% relative humidity; on the Whistle/Flute Study Options tab, I use 40,000 ppm CO2 (appropriate for exhaled air), and an absolute air pressure suitable for my altitude.

Close any files you have open, then re-open the Feadog file. You should no longer see a mention of fipple factor. What you will see instead are "beta factor" and "window height". Like the NAF study model's fipple factor, these are magic numbers we use to help the mathematics in WIDesigner model a real whistle. We start out with them set to values that worked for a similar whistle, and use the whistle calibrator to adjust them to match the tuning of the whistle under construction.

In the flute study, we use airstream length and beta factor for the same purpose; in the reed study, we use parameters called alpha and beta.

That should get you one step farther in the process.
User avatar
MadmanWithaWhistle
Posts: 697
Joined: Fri Jan 27, 2012 10:15 pm
antispam: No
Please enter the next number in sequence: 8
Tell us something.: Instrument maker and researcher exploring new methods of creating traditional instruments with longevity aforethought. Player of the whistle, flute, and continental European border pipes.

Re: WIDesigner release 2.4.0 now available

Post by MadmanWithaWhistle »

Awesome, thank you! I will be working on bagpipes next, so I'm excited to dive into that functionality soon...
User avatar
MadmanWithaWhistle
Posts: 697
Joined: Fri Jan 27, 2012 10:15 pm
antispam: No
Please enter the next number in sequence: 8
Tell us something.: Instrument maker and researcher exploring new methods of creating traditional instruments with longevity aforethought. Player of the whistle, flute, and continental European border pipes.

Re: WIDesigner release 2.4.0 now available

Post by MadmanWithaWhistle »

Edit: I got it to work, but the documentation doesn't really specify how to load a tuning, nor that this is done through "open constraints" and going to a completely separate file than the study model. Once I have this working well I'm going to write some better steps for this; it's really a shame such a powerful tool is missing the documentation for new users to get started. It just says stuff like "create a new instrument file." There is no way to do this in the GUI, and as far as I can tell one has to load an existing file and perform some operation like an optimization on it, and then save the optimization as a new instrument file. What happens if you lose the original lol?!
Ok, I've successfully loaded the Whistle study, but all the optimization tools are grayed out except for "sketch instrument". I have the FeadogMk1 loaded, and I've selected "1. Whistle Calibrator" in the optimizer tree.

https://imgur.com/a/gRjGvVj
Tunborough
Posts: 1419
Joined: Sun Dec 05, 2010 2:59 pm
antispam: No
Please enter the next number in sequence: 10
Location: Southwestern Ontario

Re: WIDesigner release 2.4.0 now available

Post by Tunborough »

Good point. I've updated the Starting Out and Data Files sections on Using WIDesigner to clarify how to open files. (Although Open constraints... will inadvertently work for opening tuning files, it isn't the intended way. File-->Open... should be sufficient.) Does that look better?

You are right that there is no way to create a new instrument from scratch. It's generally easier to start with an existing instrument and change it than starting from scratch, so that wasn't a priority. You can make changes directly on the GUI and use File-->Save As.... You don't need to use an optimizer to create a new instance of the instrument.

I realize that wiki page needs a new section explaining the basic parts of the GUI. That will take a little longer.
User avatar
MadmanWithaWhistle
Posts: 697
Joined: Fri Jan 27, 2012 10:15 pm
antispam: No
Please enter the next number in sequence: 8
Tell us something.: Instrument maker and researcher exploring new methods of creating traditional instruments with longevity aforethought. Player of the whistle, flute, and continental European border pipes.

Re: WIDesigner release 2.4.0 now available

Post by MadmanWithaWhistle »

Tunborough wrote:Good point. I've updated the Starting Out and Data Files sections on Using WIDesigner to clarify how to open files. (Although Open constraints... will inadvertently work for opening tuning files, it isn't the intended way. File-->Open... should be sufficient.) Does that look better?

You are right that there is no way to create a new instrument from scratch. It's generally easier to start with an existing instrument and change it than starting from scratch, so that wasn't a priority. You can make changes directly on the GUI and use File-->Save As.... You don't need to use an optimizer to create a new instance of the instrument.

I realize that wiki page needs a new section explaining the basic parts of the GUI. That will take a little longer.
I realize that a lot of this is built within the limitations of the file formats and Java, so the process doesn't need to be perfect UI, it just has to be documented. The biggest issue for me was that I didn't understand the way the program "thinks" enough to reasonably infer what next steps I needed to take. I think a "quickstart guide" describing the following would be helpful (I'm happy to contribute once I feel more confident working with the tool, which I truly think has amazing potential):
  • What the different file types are, and what they functionally do (constraints, instruments, tunings/scales)
  • Where the different file types are located in the directory (I had to comb the whole damn thing several times to find what I needed
  • How to select the kind of instrument study (I would re-name this to "Analysis Mode" or something like that to differentiate from instrument files)
  • The necessity of duplicating relevant files, and a reminder not to overwrite the originals
  • An example of how to load an instrument file, tuning, and relevant constraints, and perform an analysis/optimization (the broadness of the language used in the documentation made me think this would be much more automatic)
There's clearly been a lot of time spent on explaining how each function performs acoustically, but it's all jumbled in with the steps that actually need to be performed in the program. If I am skimming documentation looking for an answer to a "where is this file" question, I'm not going to read a detailed description of NAF acoustics to find the one place a step is described. I'd separate these into a glossary of acoustic effects (I know you have a glossary of definitions, this could be expanded to describe how the measured acoustic effects influence one another from the main doc), and a user guide (i.e. "here's how you perform [task]").

I think a good example is tonehole optimization: the documentation describes this process as being quite automatic, but the default state makes all the holes as wide as possible. So then one must load a constraints file and set the min/max for each tonehole, but that's not really an "optimization" without the user having some rationale for the constraints they're setting. Providing guidelines for setting tonehole min/max, be it mathematical or "just measure instruments you like" (another very valid approach) would be helpful, since one seeking help plotting toneholes is likely doing so precisely because they don't have a good idea of what the size should be.

All this feedback aside, I am really excited about what this tool can offer pipe and flutemakers. I am currently collaborating with a number of makers, such as Donald Lindsay, on next-generation bagpipe and flute/whistle designs, and the analyses this tool can provide are really promising. The avenues opened by CAD and 3D printing certainly supercharge the usefulness of something like this.
Tunborough
Posts: 1419
Joined: Sun Dec 05, 2010 2:59 pm
antispam: No
Please enter the next number in sequence: 10
Location: Southwestern Ontario

Re: WIDesigner release 2.4.0 now available

Post by Tunborough »

Thanks for the suggestions. I'm pretty confident about writing documentation, but my weakness is knowing what novice users need documented. You've given me lots to work on, and I'll work away at it as time permits.
User avatar
MadmanWithaWhistle
Posts: 697
Joined: Fri Jan 27, 2012 10:15 pm
antispam: No
Please enter the next number in sequence: 8
Tell us something.: Instrument maker and researcher exploring new methods of creating traditional instruments with longevity aforethought. Player of the whistle, flute, and continental European border pipes.

Re: WIDesigner release 2.4.0 now available

Post by MadmanWithaWhistle »

Tunborough wrote:Thanks for the suggestions. I'm pretty confident about writing documentation, but my weakness is knowing what novice users need documented. You've given me lots to work on, and I'll work away at it as time permits.
Cheers, I hope I didn't come across as overly critical here. I'll report back with a "case study" when I'm done with this one.

Is mouthpiece position the height of the splitting edge in the airstream?

Ah, follow on: before I get into this, how valuable is the calibration data? Should I really test four registers for each fingering position? Would it be enough to just measure D, G, and C#? I do not want to change the dimensions of the head, so can I just pull over the beta value and not the window height?

See my chart, where F is the fingering (# of holes covered):
https://imgur.com/a/eo3iVDO

Documentation of my steps: https://docs.google.com/document/d/1zP4 ... sp=sharing
Last edited by MadmanWithaWhistle on Fri Aug 14, 2020 7:30 pm, edited 1 time in total.
Tunborough
Posts: 1419
Joined: Sun Dec 05, 2010 2:59 pm
antispam: No
Please enter the next number in sequence: 10
Location: Southwestern Ontario

Re: WIDesigner release 2.4.0 now available

Post by Tunborough »

MadmanWithaWhistle wrote:Cheers, I hope I didn't come across as overly critical here. I'll report back with a "case study" when I'm done with this one.
Not at all. Your contributions are welcome.
MadmanWithaWhistle wrote:Is mouthpiece position the height of the splitting edge in the airstream?
As noted in the glossary, "positions" are measurements in the longitudinal direction, along the body from the mouthpiece end to the far end. The mouthpiece position specifically is the position of the splitting edge. Positions are measured from whatever reference point on the length of the whistle that you find convenient. Often, that convenient reference point is the splitting edge, in which case the mouthpiece position is zero by definition.

I would dearly like to include the height of the splitting edge in the airstream as a parameter of the instrument geometry. I'm pretty certain it affects tuning, tone colour, and sound quality. Alas, I have no idea how to reliably measure that height with tools the average whistle person would have on hand. (Or indeed any tools that I know how to use.) I asked for suggestions here a few years back, and none came up.
MadmanWithaWhistle wrote:Ah, follow on: before I get into this, how valuable is the calibration data? Should I really test four registers for each fingering position? Would it be enough to just measure D, G, and C#?
If you've got all the holes drilled, two registers should be plenty, the full two-octave range of the whistle. I would say you need to test all the fingerings, though, not just those for D, G, and C#. If you don't have any holes drilled yet, that's when you need more registers, to get a few more notes for the calibration.
User avatar
MadmanWithaWhistle
Posts: 697
Joined: Fri Jan 27, 2012 10:15 pm
antispam: No
Please enter the next number in sequence: 8
Tell us something.: Instrument maker and researcher exploring new methods of creating traditional instruments with longevity aforethought. Player of the whistle, flute, and continental European border pipes.

Re: WIDesigner release 2.4.0 now available

Post by MadmanWithaWhistle »

Brilliant, thank you so much!

My approach to measuring airblade position is that zero is flush with the windway floor. With the Kalischeck's 'three tube' system of making a whistle (which most people follow, whether they're aware of it or not) the blade is inherently flush with the windway floor. I find that an offset factor (I apply it to the windway) that raises the blade slightly into the airstream is necessary for optimal tone and response. I call this factor "Windway offset" and it's usually around -0.02 inches. In recorder (gasp) literature, the rule of thumb is that it should be raised into the lower third of the airstream.

I don't know how well this would translate to a theoretical model, though, since it's not a direct measurement of the blade in relation to the air column. I also have the benefit of using CAD and SLA for development, but with a good eye and a small ruler or feeler gauge you could certainly measure this manually.

(btw, I've included the procedure I'm following in my previous message, and will keep adding to this)
Tunborough
Posts: 1419
Joined: Sun Dec 05, 2010 2:59 pm
antispam: No
Please enter the next number in sequence: 10
Location: Southwestern Ontario

Re: WIDesigner release 2.4.0 now available

Post by Tunborough »

RTTA isn't the best choice for measuring min and max frequencies. RTTA doesn't know what fingering you're using. B, C-nat, and C#, in particular, can easily stray into each other's neighbourhood at the extremes of breath pressure, and RTTA logs each measurement under the closest note, regardless of the fingering. I prefer to play each fingering in front of an electronic tuner or tuner app as I vary breath pressure up and down. If you want to use RTTA, you can enter its median measured frequencies in WIDesigner as actual playing frequencies, and calibrate your whistle using those rather than min and max frequencies. However, the whistle has to be reasonably well in tune, so RTTA logs XOOOOO as a B, and so forth.
User avatar
MadmanWithaWhistle
Posts: 697
Joined: Fri Jan 27, 2012 10:15 pm
antispam: No
Please enter the next number in sequence: 8
Tell us something.: Instrument maker and researcher exploring new methods of creating traditional instruments with longevity aforethought. Player of the whistle, flute, and continental European border pipes.

Re: WIDesigner release 2.4.0 now available

Post by MadmanWithaWhistle »

Ha, I was only using RTTA because I was convinced my other tuning app didn’t give a Hz value. I opened the wrong app as I was beginning and found that it DOES display Hz, so I just used that!
Post Reply