It is currently Tue Mar 19, 2019 12:59 pm

All times are UTC - 6 hours




Post new topic Reply to topic  [ 60 posts ]  Go to page Previous  1, 2, 3, 4  Next
Author Message
 
PostPosted: Tue Jan 08, 2019 7:08 am 
Offline

Joined: Sun Dec 05, 2010 2:59 pm
Posts: 896
Location: Southwestern Ontario
Terry McGee wrote:
Woah, very interesting indeed. As I said, "we can't assume Boehm got it perfectly right". But equally, we can't assume WID got it right either, given it's pretty early days yet in its application to transverse flutes. But we owe it to everybody to find out, and, whatever we find, to progress it from there.

...

It's sobering to remember that if WID found that with the fiddling-the-head-bore approach, it might come up with some equally interesting suggestions for our older fiddling-the-body-bore approach.
Agreed, on all counts.

Now, Boehm was dealing with two things that I didn't include: extra toneholes to give all the accidentals, and notes up through the third octave. Either of these could change the outcome.

Terry McGee wrote:
Tunborough, are you planning to do anything physical or should we detail somebody else to have a shot at it? Geoffrey? Anybody?

The kind of shapes Tunborough has mentioned are a bit out of our usual comfort zone, but I bet we could find ways to achieve or at least approach them, even if it requires inserting some separately turned and bored sections.
Even if I was able to physically realize these flute designs, I wouldn't have the playing skills to judge the outcome. I am tempted to try something of the sort on a whistle.

The complex, expanding and contracting, headjoint would be a huge challenge, but with the right flute, the contracting-only headjoint would be as simple as sliding in a cylinder of 0.2 mm mylar down the headjoint for 60 to 80 mm, ... WIDesigner predicts there would be no perceptible difference in the tuning of the two headjoints. (Tone colour is another matter, but I'd have more confidence in the sonic consistency of the simple headjoint over the complicated headjoint.)


Top
 Profile  
Reply with quote  
 
PostPosted: Tue Jan 08, 2019 10:15 am 
Offline
User avatar

Joined: Thu Sep 13, 2012 1:15 pm
Posts: 287
One interesting aspect of the optimized xiao that I made to Yan Lang's specifications was that in order for it to work properly, many other factors apart from the bore had to be changed.

On a traditional xiao, the bore is cylindrical, the finger holes are uniformly sized and fairly uniformly spaced in two groups (not including the thumb hole). Wall thickness is the same throughout. The holes are slightly ovaled, which seems to contribute to better intonation, helping to bring the second octave to pitch a bit better. As I've commented before, it is a very sophisticated and functional design. It has some small weaknesses (like most wind instruments) but overall it's very solid.

BUT...the "optimized" version, in order to complete it's optimization, had not just the undulating inner bore, but the outside had to undulate as well! The size and placement of the holes was no longer as uniform, and the wall thickness had some variation. Not a ton, but in order for the wall thickness to remain reasonably consistent, the outer diameter had to mirror the inner diameter. This makes the manufacture of it (from wood with a cast bore) very tricky. It was dangerously thin in places, causing whip if put on the lathe and a host of other difficulties.

And the part that makes me curious about doing some similar redesign of the Boehm headjoint is: will the supposed "gains" be on par with those of this optimized xiao? It was undoubtedly done with a deep understanding of acoustics, using advanced computer modeling and a highly sophisticated prototyping process (see link below), but my point is that the result of the optimization wasn't really that impressive compared to the traditional design. I'm being redundant for those who have read my comments on this in another thread, but the result of the optimization was a xiao that no longer sounded like a xiao. It did seem to have somewhat more uniform tone (in terms of strength from note to note), the intonation was a bit more on target without having to do much tweaking, and it did improve the C natural note (which is weak on a xiao, just like on the Irish flute). But in the process, the timbre of the instrument changed and it actually became more difficult to play. Less free blowing, much less stable on the root note. And one serious xiao player (who trained on Boehm flutes and plays Irish flute as well) said "It sounds really badass, but it's too much like an Irish flute now".

So it makes me wonder (especially when it comes to world flutes from different cultures) whether acoustic optimization erases some character in exchange for these improvements. And as such, are they really improvements? Will all optimized wind instruments end up sounding the same? I'm curious. The Boehm flute seems to work remarkably well in it's present form, and will a different headjoint design make it better, or simply different? Same applies to conical bore flutes, of course. For the people who love them, how much improvement would be desirable if the cost is an alteration of the instrument's character?

Terry, Tunborough and anyone else interested in flute acoustics and in checking out a very impressive bit of research should see this:

https://open.library.ubc.ca/cIRcle/collections/ubctheses/24/items/1.0167080

_________________
Geoffrey Ellis Flutes


Top
 Profile  
Reply with quote  
 
PostPosted: Tue Jan 08, 2019 12:49 pm 
Offline

Joined: Sun Dec 05, 2010 2:59 pm
Posts: 896
Location: Southwestern Ontario
Geoffrey Ellis wrote:
Terry, Tunborough and anyone else interested in flute acoustics and in checking out a very impressive bit of research should see this:

https://open.library.ubc.ca/cIRcle/collections/ubctheses/24/items/1.0167080
Impressive, indeed. Thanks for that, Geoffrey. I hadn't seen it before. Lan's methodology was very much like what we've done in WIDesigner: a transfer matrix model using the tonehole model from Lefebvre and Scavone, with an empirical mouthpiece model. One difference stands out after a quick skim: his optimization runs took hours or days on a large cluster computer; the optimization runs for the headjoint took WIDesigner maybe 15 seconds each on my laptop. At least part of the difference is being more selective about what we're asking the optimization to do.


Top
 Profile  
Reply with quote  
 
PostPosted: Tue Jan 08, 2019 5:25 pm 
Offline
User avatar

Joined: Sun Dec 12, 2004 4:12 pm
Posts: 1981
Location: Malua Bay, on the NSW Nature Coast
Tunborough wrote:
Even if I was able to physically realize these flute designs, I wouldn't have the playing skills to judge the outcome.

OK, then one of us should, and unless somebody else puts their hand up, it might as well be me. It doesn't sound hard.

Can you tell us the type and the ID and OD of the 2 samples of tubing you've been basing it on, or at least the one you would prefer to start with. First job is to see if I can procure some around here. Failing that, can I come back to you with an ID and OD of a tube I can get, and get you to recalculate?

Alternatively, I notice I have some bright orange* electrical conduit here 19.65 ID and 24.82 OD. A bit thin for an ideal embouchure, but I could slit and glue an offcut to thicken it. Then it would be a little thick at around 30mm, but I could turn it down to whatever.

Then what's the plan? Do I make two flutes - one a plain cylinder and one with the modified bore? And both to dimensions provided by WID?

Do you want the flute(s) tunable (to set A to 440 and then see where the other notes lie), or non-tunable (to see what we get)? Or we could start with non-tunable, then cut the tube and insert a slide if that has attractions. Or I could start with a piece a little too long, then snip bits off until I get a well pitched low D and we work from there?

*The Auld Orange Flute? https://www.youtube.com/watch?v=gSatdW0EfPU


Top
 Profile  
Reply with quote  
 
PostPosted: Tue Jan 08, 2019 7:49 pm 
Offline

Joined: Sun Mar 25, 2018 5:40 am
Posts: 376
this is so interesting. thank you for the discussion


Top
 Profile  
Reply with quote  
 
PostPosted: Tue Jan 08, 2019 9:06 pm 
Offline

Joined: Sun Dec 05, 2010 2:59 pm
Posts: 896
Location: Southwestern Ontario
That orange conduit would do the job. It's got an ID and OD in the neighbourhood I was looking at. Do you have something you could use to make a cylindrical insert to reduce the ID to 18.9 mm or so for the top 80-100 mm? Maybe a flexible plastic sheet about 0.4 mm thick?

The place to start would be to cut off, say, 570 mm. Push a stopper in 15 mm from one end, and cut an embouchure hole centred about 10 mm beyond that. I'll leave it to you to decide the length and width--and shape and depth--of the embouchure hole; just let me know what they are. If we say the centre of the embouchure hole is 0, then the stopper is at -10, and the far end of the tube is at 545. We'll probably lop off another 4-6 cm from that end.

Let me know what range of pitches you can get with that tube. For example, by varying your breath, you might be able to play D4 varying between 281 - 291 Hz, D5 between 566 and 586 Hz, A5 between 855 and 880 Hz, D6 between 1150 and 1180 Hz, before the flute switches into the neighbouring register.

For drilling the toneholes, are your drill bits metric or imperial? What's the largest size of tonehole you are comfortable with? How big a finger stretch can you accommodate between T2 and T3, and between B2 and B3?


Top
 Profile  
Reply with quote  
 
PostPosted: Tue Jan 08, 2019 11:49 pm 
Offline
User avatar

Joined: Sun Dec 12, 2004 4:12 pm
Posts: 1981
Location: Malua Bay, on the NSW Nature Coast
Tunborough wrote:
That orange conduit would do the job. It's got an ID and OD in the neighbourhood I was looking at. Do you have something you could use to make a cylindrical insert to reduce the ID to 18.9 mm or so for the top 80-100 mm? Maybe a flexible plastic sheet about 0.4 mm thick?

Not sure, but if not I could probably turn something. Ah, a plastic folder cover that is 0.38? Also in orange!

So that 80-100mm counted from the top of the tube and so includes the stopper? And the embouchure hole goes through it? Might need to glue it around the embouchure hole to prevent it moving?

Quote:
For drilling the toneholes, are your drill bits metric or imperial?

Metric, and the best ones are at 0.5mm intervals. EG 5, 5.5, 6 ......12, 12.5. But I have plenty of other drills if needed.

Quote:
What's the largest size of tonehole you are comfortable with?

Say 11mm? 10.5 better...

Quote:
How big a finger stretch can you accommodate between T2 and T3, and between B2 and B3?

Let's try 35mm between T2 & T3, and 37.5 between B2 and B3. Could push them a little more, but nice if not necessary. Those figures taken from an original Pratten. My Moon cylindrical pushes the B2-3 to 41mm but I find it a stretch.

Now, I'd normally aim for an oval embouchure hole of say 12mm long and 10.5mm across. To make it easier to repeat and calculate, if I use a 10mm bit and mill a slot 12mm long? So it will end up with the area of a 10mm circle, plus a rectangle of 2 x 10mm? No undercutting.


Top
 Profile  
Reply with quote  
 
PostPosted: Wed Jan 09, 2019 8:10 am 
Offline

Joined: Sun Dec 05, 2010 2:59 pm
Posts: 896
Location: Southwestern Ontario
Terry McGee wrote:
Ah, a plastic folder cover that is 0.38? Also in orange!

So that 80-100mm counted from the top of the tube and so includes the stopper? And the embouchure hole goes through it? Might need to glue it around the embouchure hole to prevent it moving?
Sounds excellent. The colour coordination is particularly important.

The length is counted from the stopper face, and yes, the embouchure hole goes through it. We might be able to get away with an oversize gap in it for the embouchure hole, so there's no risk of it encroaching on the embouchure hole in the conduit. Determining the length and installing it will be the last step in the construction, so don't glue anything yet.

From here on, I'm going to ignore anything above the stopper face, since it is acoustically invisible. As long as you have enough conduit there to move the stopper up and down and remain secure, all is good.

Terry McGee wrote:
Now, I'd normally aim for an oval embouchure hole of say 12mm long and 10.5mm across. To make it easier to repeat and calculate, if I use a 10mm bit and mill a slot 12mm long? So it will end up with the area of a 10mm circle, plus a rectangle of 2 x 10mm? No undercutting.
Whatever works to give a reasonable approximation of your usual cut. Undercut as you think necessary for playability.

Tunborough wrote:
Let me know what range of pitches you can get with that tube. For example, by varying your breath, you might be able to play D4 varying between 281 - 291 Hz, D5 between 566 and 586 Hz, A5 between 855 and 880 Hz, D6 between 1150 and 1180 Hz, before the flute switches into the neighbouring register.
Correction to these instructions ... For each note, vary only your breath speed to determine the range of frequencies for that note. Going between notes (harmonics), you'll want and need to adjust your mouth to take a good run at the new note. I could ask for only one nominal playing frequency for each harmonic, but since there are so few notes at this stage, I'd like the two extremes for each one to get more data.


Top
 Profile  
Reply with quote  
 
PostPosted: Fri Jan 11, 2019 3:45 pm 
Offline
User avatar

Joined: Sun Dec 12, 2004 4:12 pm
Posts: 1981
Location: Malua Bay, on the NSW Nature Coast
Thanks for that Tunborough, sounds very doable. I hope to get to get on to that today.

Sorry to have gone quiet on this, I've been a bit tied up with big carillon developments and plumbing. Installing a fire pump and fire hose reels around the house and workshop. Climate change is making bushfires an increasing risk in this part of the world.

And you thought the Fire Hose Reel was the name of a tune?

I quipped to my plumbing mate that we're in the same line of business. Controlling the way transparent fluids pass up and down inside hollow vessels...


Top
 Profile  
Reply with quote  
 
PostPosted: Fri Jan 11, 2019 5:58 pm 
Offline
User avatar

Joined: Sun Dec 12, 2004 4:12 pm
Posts: 1981
Location: Malua Bay, on the NSW Nature Coast
OK, I'm back!

Have made the start, but wanting to proceed with caution to get this right. It struck me that we might want a bit more room up at the head end if we find we need to increase stopper distance or whatever. And while the stopper is moveable, the hole is fixed, so let's use it as the point of reference in future.

I've set the embouchure hole 50mm down from the end of the tube and glued on a 50mm length of offcut split lengthwise to deepen the hole. The hole is a simple slot formed by milling with a circular cutter, and is 10.19 across, and 12.1mm long.

I left the tube a little longer for a start given I'm messing with the instructions. The far end of the tube is now 560mm from centre of embouchure.

The stopper is a 10mm thick cylinder of Delrin, with a shallow trench and thread grooves around it. I used Teflon String (rather than tape), a newly-fangled idea introduced to me by my plumber mate last week, to create an easily adjustable seal. It also worked great on the fire pump and fire hose reels!

Image

For a start, and to check the flute was workable, I've set the stopper at 19mm from centre of embouchure. After warming up, I got these "comfortable playing pressure" readings:

Note Samples Delta in Cents
C#4 1270 11
C#5 1111 14
G#5 1145 -1
C#6 1165 -4
F6 1014 -10

What would you have me do next?


Top
 Profile  
Reply with quote  
 
PostPosted: Fri Jan 11, 2019 9:18 pm 
Offline
User avatar

Joined: Sun Dec 12, 2004 4:12 pm
Posts: 1981
Location: Malua Bay, on the NSW Nature Coast
I noticed I had some leakage between the tube and the snapped on lip plate, so I've glued that up just in case. No leakage now!

And rerun the test, which gave slightly different results. Probably more to do with player variability than the leakage! Here's three sets of readings, post gluing.

Note Delta in Cents
C#4 +4 +7 +3
C#5 +6 +5 +5
G#5 -4 -7 -5
C#6 -4 -10 -3
F6 -8 -15 -14

(Anyone know how to neaten up tables in C&F? Is there a way to do tabs or columns?)


Top
 Profile  
Reply with quote  
 
PostPosted: Fri Jan 11, 2019 9:41 pm 
Offline
User avatar

Joined: Sun Dec 12, 2004 4:12 pm
Posts: 1981
Location: Malua Bay, on the NSW Nature Coast
And reducing the stopper face to centre of embouchure distance to 10mm produces:

Note Delta in Cents
C#4 11
C#5 11
G#5 3
C#6 12
F6 7


Top
 Profile  
Reply with quote  
 
PostPosted: Sat Jan 12, 2019 11:20 am 
Offline

Joined: Sun Dec 05, 2010 2:59 pm
Posts: 896
Location: Southwestern Ontario
You can use the "Code" button to get a block of monospaced text that doesn't throw away extra spaces. Example below.

By adjusting the calibration, I got WIDesigner to predict the frequencies in the tables below.
Code:
With stopper at -19 mm
       Measured    Predicted  Error
C#4      278       278.72   4.5
C#5      556       554.69   -4.1
G#5      828       830.5    5.2
C#6     1105      1105.3   0.5
F6      1387      1378.47   -10.7
         
With stopper at -10 mm
C#4       279   278.96   -0.25
C#5       558   556.88   -3.48
G#5       832   836.9    10.17
C#6      1116   1118.53   3.92
F6       1402   1400.96   -1.28

I'm a bit concerned that G#5 is sharp. That does suggest a weakness in the model that we'll need to address.


Top
 Profile  
Reply with quote  
 
PostPosted: Sat Jan 12, 2019 12:02 pm 
Offline

Joined: Sun Dec 05, 2010 2:59 pm
Posts: 896
Location: Southwestern Ontario
So, if we trust the WIDesigner flute model, and if I've got the right calibration, here is how I'd like you to drill the holes. All distances are from the centre of the embouchure hole. If you don't think you'll be able to play this geometry, particularly the T1-T2 spacing, let me know before you drill. I can adjust the constraints to give you something more playable.

Code:
Stopper at -13 mm
End of bore at 526 mm
Hole   Position  Spacing   Diameter
T1      236.6               8.5
T2      271.6     35.0      9.5
T3      306.6     35.0      8.0
B1      365.0     58.4      9.0
B2      389.1     24.1     10.5
B3      426.6     37.5      7.0


The resulting flute won't play in tune for three reasons:
- The headjoint insert isn't in yet.
- I haven't got the calibration factors right.
- The WIDesigner model needs work.

But with luck, the tuning you measure from this geometry will get us closer on all three of these.

If you want to hedge your bets, you're welcome to drill the holes undersize and cut the bore a bit long, and let me know what you measure from that geometry before we remove any more PVC.


Top
 Profile  
Reply with quote  
 
PostPosted: Sat Jan 12, 2019 6:17 pm 
Offline
User avatar

Joined: Sun Dec 12, 2004 4:12 pm
Posts: 1981
Location: Malua Bay, on the NSW Nature Coast
Tunborough wrote:
You can use the "Code" button to get a block of monospaced text that doesn't throw away extra spaces. Example below.

Great, thanks for that!

Quote:
I'm a bit concerned that G#5 is sharp. That does suggest a weakness in the model that we'll need to address.

Heh heh, or a weakness in my blowing that I need to address?

Either way, those predictions are very close. This is getting exciting....


Top
 Profile  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 60 posts ]  Go to page Previous  1, 2, 3, 4  Next

All times are UTC - 6 hours


Who is online

Users browsing this forum: Conical bore and 10 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group
[ Time : 0.329s | 12 Queries | GZIP : On ]
(dh)