What gives flutes/whistles/recorders their distinctive signature sound?

[GreenWood]

...it takes time, but is only as demanding as anyone is towards it all. I think there is a bit of zen to it, because when approached in the right way everything seems to make sense, but to try to oblige an instrument to be a certain way and soon frustration will set in. So it is more an "enjoy the journey" sort of activity, and if the destination ends up good, so much the better.

[David Cooper]

It's the inner bisel that's the problem - I found a video with someone drawing a diagrams of how to shape that, and I'm now getting the lower note, but need to work on improving the quality. I've drilled the lowest three finger holes and got them as close together as the ones on my D tin whistle (which is a fifth higher in pitch), so it's really comfortable. The middle hole is bigger than on a normal quena, and the lowest one is smaller than normal (still a bit flat as I haven't tuned them carefully yet), but the volumes don't seem to be greatly different, so they're fine, and that means my main objective should be achievable, provided that I can get the sound quality I'm after as well. I need to buy a round file - using sandpaper round a pen is a poor substitute.

[GreenWood]

That's great. Personally I don't get too caught up on tonehole size...it makes a difference (response for small vs flexibility like slides for large, very small and you lose some tone and loudness) but not too much difference in my opinion. Still, some tend to focus on toneholes as an answer to how a flute plays, but I find the underlying sound is there whatever adjustment (within reason) is made to them, and that the sound is based more on bore, flute embouchure and player's choice of technique. So if you have found an arrangement that works for tuning and spacing, finding how to get a good sound from the flute itself is a supplement to that and should not need change in layout. Once a design is found that is good, then tweaking on subsequent ones of that if built, and learning about related small detail is possible, which can also be fun

[David Cooper]

I put the upper three holes too far up, so they aren't as big as I'd planned, but it doesn't matter on this experimental instrument anyway as it's now set a tone too high due to all the repeated attempts to get the notch right, and it still isn't sounding the notes well enough: I can't get it to play the highest two notes of the second octave at all. That's what I need to fix most.

By the way, I've found a way of measuring where the drill tip is within the wood to within perhaps a fifth of a millimetre while boring it. The key to this is to use a magnet. By luck, the speedbor has three cutting edges, and that allows you to measure the distance from each of them to the magnet by having the magnet lift towards them when they get close enough to it. If the magnet can't get close enough to lift itself straight up, that's no problem as it can be set to move sideways instead so that it moves more easily - I've tested this with a weak fridge magnet which can be set up at an angle such that it just swings over when the drill bit tip gets to a very specific distance from it, while wood is of course completely transparent to the magnet. This means I should be able to buy 600x25mm birch rods (the most suitable ones that I can get from Amazon) and drill them right down the middle with an 18mmm speedbor with the side walls averaging 3.5mm wide and only varying by perhaps 0.5mm. It might even be possible to do it with a 20mm bit with the sidewalls averaging 2.5mm and going down to 2mm in places, though I don't know if that's considered thick enough for a flute when it comes to sound quality, and it could limit the options too much when shaping the holes. The 18mm bit may be easier to steer than the 16mm one as the shaft looks as if it may be narrower relative to the cutting part, so it's likely the best one to use for making quenas from the readily available 25mm diameter rods. I doubt they knew it when they designed these bits, but I think they've made the perfect tool for boring out a rod with high precision while using a hand-held drill, and that makes this process very inexpensive. That said, I haven't actually done it yet for real, but the physics of it can't be wrong so I'm now going to risk it and buy some of those rods to try it. If it goes way off course within the first 20cm, I can abandon that attempt and try from the other end, so I'll get two goes at it for each rod.

[GreenWood]

That is a good idea

At 50cm,
Rod thickness 1cm
Bore 2cm
Steering room of rod at start is 2cm - 1cm / 2 = 0.5cm = roughly 5° change in direction.

In fig 1 below I try to draw the limit to this, that is to say to just set the bore parallel lengthwise (but still off center) in that example, the rod might need to be outside the bore (purple line) and in that circumstance you would end up with a bore that curves away from the edge but still moving towards the outside, but at a lesser angle (i.e. further down the flute, hopefully after the end of it).



However I haven't done a full calculation, and in theory, with rod 1cm diameter and bore 2cm, the leeway is therefore 0.5cm for the tip of the powerbore where you can still set it straight. At 0.5cm it would then continue parallel but the bore remain off center.

Remembering that the point will act like pivot and the blades will lift a tad towards edge at first when steered (fig 2)


So these are quite small tolerances to measure down the bore, and not forgetting that sometimes the bit will walk slightly by itself, and also there will be some mm of flex in the rod as it gets longer, something drillbits like to take advantage of.


So, at my guess you will have to keep the bit near center the whole time (or use a thickness of wood that allows for whatever error appears) for this to work, and I hope it does work.

The only idea I had with regard, is that if tilting the rod to steer the bit is not enough, then a half pipe of say plastic (e.g. 0.7 cm thick plastic if the gap between rod and wood is 0.5cm when centered) could be pushed down the side where the drillbit is too close to edge, so making the drillbit move back towards center by leverage. Powerbore have lead screws I think, which might not allow this to occur, but you could always load a simple spade bit for the maneuver ?

If you are not short of blanks of wood and drilling is fast, then even only getting one in every few right might be acceptable, but I hope you manage to consistently drill them as wanted.

2mm is maybe thick enough if the wood is good etc. , but in my opinion it is about at limit even then. I have made "bores" for lining bores about 1mm thick, and they were solid enough to handle normally (but not to apply great pressure to), but the problem is that if they were the actual bore, moisture changes etc. would probably weaken, warp and split them eventually. One flute ended up with a paper thin (literally) area (about 1cm wide by 3cm long) that I could flex by finger pressure.... it lasted for a few months before a very fine split appeared, unnoticeable by eye but enough to make the flute jump octaves randomly...patched it and all good again. For tuning I think there are both advantages and disadvantages to a thin walled bore, but I leave that there.

Top notes tend to be the more difficult of them to tune the flute to play, plus the player has to have confidence that he or she is well able to play those normally.

[David Cooper]

So, at my guess you will have to keep the bit near center the whole time (or use a thickness of wood that allows for whatever error appears) for this to work, and I hope it does work.

The drill gets hot and needs lots of opportunities to cool down, plus you can't drill far without bunging the thing up with dust which makes it hard to extract the bit, so it's best to drill just a couple of centimetres each time, then clear the tube, then put the bit back in and measure with a magnet. That should prevent it from wandering a long way off course as the course corrections can be made soon enough to ensure that there's always plenty of leeway to make stronger corrections if it keeps going off track due to density differences through the rod. If the head was to reach 2mm away from the centre, at that point there would be no way to steer it back other than relying on luck, or maybe trying to bend the rod, but it should never be allowed to get anywhere near that far out: the magnet can detect much smaller movements to the side and can likely keep the wandering down to under half a millimetre off centre.

Once half way through the drilling and with the head right in the centre, it may also be possible to switch to an auger and let it steer itself the rest of the way as it likely can't go significantly off course after that.

The only idea I had with regard, is that if tilting the rod to steer the bit is not enough, then a half pipe of say plastic (e.g. 0.7 cm thick plastic if the gap between rod and wood is 0.5cm when centered) could be pushed down the side where the drillbit is too close to edge, so making the drillbit move back towards center by leverage.

That might help in an emergency where it keeps wandering strongly towards the edge, though it could be hard to keep it in the right place with the shaft trying to move it round. Grease on the inner surface of the plastic might fix that and reduce the added friction. The friction does get high, but it may be that after every couple of centimetres bored you can pull the bit back a little (to get the screw-tip out of the pilot hole that it makes for itself), and then run the drill while moving the shaft in circles at the opening of the tube to make the head widen the bore a fraction.

2mm is maybe thick enough if the wood is good etc. , but in my opinion it is about at limit even then. I have made "bores" for lining bores about 1mm thick, and they were solid enough to handle normally (but not to apply great pressure to), but the problem is that if they were the actual bore, moisture changes etc. would probably weaken, warp and split them eventually. One flute ended up with a paper thin (literally) area (about 1cm wide by 3cm long) that I could flex by finger pressure.... it lasted for a few months before a very fine split appeared, unnoticeable by eye but enough to make the flute jump octaves randomly...patched it and all good again. For tuning I think there are both advantages and disadvantages to a thin walled bore, but I leave that there.

I'll use the 16mm bit again the first time, getting 4.5mm thick walls on average, and if that goes well I'll give an 18mm bit a go: that'll give me 3.5mm walls which may still be enough.If it wanders off by half a millimetre, that should still be acceptable unless the mismatch between the thicknesses of the walls on opposite sides affects the tone badly. If that doesn't spoil the sound, it might be possible to put the bore slightly off centre on purpose in order to have slightly thicker wood where the holes are, though the easier solution if I'm going to make more than a few of these would be to buy a lathe and solve all the problems that way while drilling straight holes with augers and not needing to care if they go off in the wrong direction by a few degrees - I see now that high quality lathes that can handle the lengths of instrument I'm making start at about £300, which is a fraction of what I was expecting, and it would give me a better choice of wood to use. I just want to be sure I can make instruments of adequate quality before I start putting that much money into it.

[GreenWood]

Being able to keep to a mm of center is pretty good.

Right now I'm making a miniature forge to shape a one meter 16mm bit, but on a 12mm diameter shank. That will hardly flex at all compared to 10mm shank, and hopefully that will be accurate enough for the wood sizes I use. I will include all that on my thread (if it all works ... the forge does, I needed the extra heat compared to a torch, to be able to bend thicker metal, not just move it by hammering).

Drill bits heat up, wood insulates the heat. Usually it is friction with swarf and/or sidewall, which is normal. Unless you find a bit that works well and with little friction, then resting the work is the answer, and anyway that is done for clearing swarf. Bits fed with compressed air are used sometimes but that is going on a luxury in my opinion. To spend an hour drilling a bore instead of quarter of an hour is not such a big deal when only making the occasional flute.


Sure you will be able to build good instruments , and if not yet then a little learning goes a long way. The trouble I find is that there are many details that count, so most of my learning has been about being able to figure out which needs attention and how, though none are particularly difficult to resolve once understood. Also, with a new instrument, either a new own sort of design or a copy that I am not familiar with playing at all, it takes a couple of months of playing it to start to understand properly what that instrument is capable of, and what not, and so what possibly to adjust beyond standard. That depends on makers playing experience I suppose, for example someone who already knows how to play a Rudall flute will have some kind of starting point of embouchure and flute behaviour to expect of one similar.

A lathe... and why not (except to be sure that it will be used) . Including the price of that and even quality woods it could not be considered an expensive pursuit, and if others are willing to pay for a few flutes then costs are covered. I like just playing around making tools and that sort of thing, if I wanted to streamline it all a bit I would also buy a proper lathe. That would make drilling the bore and finding more ideal drill bits easier also.

I don't think mismatch of wall thickness matters much, if at all. Various flutes were built (probably still are) with the bore purposefully off center (for thicker wall at embouchure for example). Even a slightly wavy bore is ok I think , but if bore diameter changes size much in places it will have an effect that might or might not be acceptable.

[David Cooper]

Being able to keep to a mm of center is pretty good.

Five flutes needing to be cut out of wooden rods arrived today. The pilot hole went in the right place, but the first contact of the 16mm bit's cutting edges cut a circle into the end that's 0.5mm out. That's not unexpected, so the challenge now is to try to ensure that that's as far out as it gets. It's too late to drill any further tonight, but I've done enough to check that it can cope with this type of wood (birch). I'll probably take several hours to drill it, maybe over a few days with lots of careful magnetic measurements - I don't want to make any mistakes, or get any more smoke out of my drill (which doesn't seem to have lost any power, so hopefully it didn't burn too far through the insulation on the wires in the coils). By weighing it, it appears to be at least twice as dense as the wood I used for the first experiment.

Right now I'm making a miniature forge to shape a one meter 16mm bit, but on a 12mm diameter shank.

I look forward to seeing pictures of that.

Sure you will be able to build good instruments

I hope so, but I'm having no luck with getting the notch right, so that's the big potential project wrecker. I've had a go with blu-tack to try out a host of different shapes without having to cut anything, and I just can't find any pattern to what works and what doesn't. However, when I try the same on bought quenas (not high quality ones, but two of them produce attractive notes), when I form blu-tack notches over the existing notches and modify the shapes, I get similarly high quality notes out of them, which leads me to think the real problem may be the type of wood I used: its low density may lead to it sapping too much energy out of the sound waves in the tube, making high quality notes impossible. I don't know if birch is good enough either, though I've seen information that suggests it can be used for flutes, so it should be okay for the next few experiments. I considered getting beech instead as I could have got twice as much for the same money, and it may be more dense, but the dust is carcinogenic and I don't need that added complication. I've got a holly tree with a fat trunk (30cm) in the garden that shouldn't be there and which is damaging other trees, so I could maybe use that in the future.

I don't think mismatch of wall thickness matters much, if at all. Various flutes were built (probably still are) with the bore purposefully off center (for thicker wall at embouchure for example). Even a slightly wavy bore is ok I think , but if bore diameter changes size much in places it will have an effect that might or might not be acceptable.

Judging by the shapes of tubes used in the french horn, I suspect a very wavy bore could be okay too. I've decided against changing the bore diameter as I can achieve everything I need to just by varying hole sizes and positions, and it seems to work fine for two octaves with prototype 1. I can't get any notes out of it in the third octave, so I can't check how badly that might be out of tune. One possible issue is the shape of the leading edge of the drill bit as it produces a convex bump at the end of the tube that it bores. It may be better to follow it up with an auger to flatten that. One thing I want to try though is drilling a hole in from underneath at the far end of the instrument instead of having a hole in the middle of the circular end: that should prevent drips of condensation getting stuck there and rotting the wood in that location. I may also be able to make it big enough to get in there to modify the shape of the interior end of the bored tube to remove the convex bump that way. Having the end hole underneath may also improve the sound for anyone it's pointed at as you can hear strange and unpleasant effects when a flute is aimed straight at you - I noticed that at a James Galway concert. Perhaps all flutes could benefit from having the end redesigned in the same way.

Edit (on 26th Mar):-

The drilling's getting there slowly with the cutting edge of the bit now at 17cm down inside the rod and almost perfectly central. I drill for one minute at a time and do this every half hour so as not to get anything too hot, with that leading to the bore extending by only a centimetre or so each time. The dust comes out as a fine powder instead of the small chips that I got out of the original type of wood that I used, while the bit no longer pulls itself through the wood but actively needs to be pushed instead. The measuring with a magnet works beautifully: I've used a series of ball-bearing-like magnets sitting on a bigger magnet, tuning this to just the right weight for the magnetic strength for it to lift straight up and hold itself against the wood when the bit is almost exactly central. At times it only sticks in two places round the tube instead of three, so I steer back towards the side the magnet failed to stick to. The deeper the bore gets, the easier it is to steer smoothly. Rapid steering tears strips out of the side which may be hard to sand out later, but that's only happened in the first two inches where I was making the biggest steering corrections, and that part of the tube will be cut off later. From there on down the bore looks very smooth. I was planning to move the shaft in circles at the drill end to use it to widen the bore sideways a little, but that's now cancelled as I realise that it would scar the wood. I'm confident now that I could do this with an 18mm bit too in a 25mm rod, though it may be better to get to that diameter just by sanding. For sanding, I'm thinking of attaching sandpaper to a rod with a layer of sponge in between and running the drill anticlockwise so that the rotation speed is guaranteed to be low.

[GreenWood]

That all sounds fun ... density of Birch is given at around 700 kg/m3 , which is around fruit wood often used for flutes. Density isn't hardness and there are various ways it is measured (i.e. how dry when measured etc.) so not always comparable right down to a detail between charts , but it gives an idea of how suitable a wood is...hardness mostly follows density. Definitely the harder the wood the more difficult it is to drill

https://www.engineeringtoolbox.com/wood ... -d_40.html

I don't know what could be the answer to the tone/notch. Wood quality, roughness of bore are two examples, air leak... if not sorted, then try the question on the whistle forum maybe. There are quite experienced makers or players who might have suggestions.


Gentle sanding on a sponge tends to give an even finish (at least as long as there is some length to the sandpaper) .

[David Cooper]

density of Birch is given at around 700 kg/m3 , which is around fruit wood often used for flutes.

The wood I used before wasn't only a bad type, but I was using branches which are much less dense than the wood of the trunk. The wood I'm using now feels radically different.

I don't know what could be the answer to the tone/notch. Wood quality, roughness of bore are two examples, air leak...

Well, I think it may just have been the result of the bore being very rough, and the wood being much too full of microscopic air cavities: it was likely absorbing too much energy to allow a good sound to be generated. I've made a notch at the end of the rod I'm currently boring out and it plays a good note already, even though there's no hole at the other end yet and the bore only extends in by 23cm (while the magnet tells me it's exactly in the middle).

I'm now drilling about 1.5cm each time, once every 20 minutes. The top of the tube is slightly bent, and that may be because the extra heat generated by all the sharp steering I had to do early on had warped it there. I'll be cutting that bit off anyway as it's scarred inside, but it again suggests it's best to take things slow and avoid allowing too much heat to build up in the wood, so it isn't just the drill that needs to be protected from overheating.

Gentle sanding on a sponge tends to give an even finish (at least as long as there is some length to the sandpaper) .

It occurs to me that I could, if I need to, get a conical bore by sanding in this way by gradually inserting a rotating rod with a short stretch with sandpaper at the end (and sponge underneath) and then drawing it back out, all done at a constant rate, then repeat this with it going a bit less far down the tube, and each repeat would go less far than the one before. It may be difficult to measure the resulting conicity, but I'd only need to note the timings and the effect the result has on the tuning of the third octave, building up statistics over time that lead to better instruments with each one I make.

[GreenWood]

"Well, I think it may just have been the result of the bore being very rough, and the wood being much too full of microscopic air cavities: it was likely absorbing too much energy to allow a good sound to be generated. I've made a notch at the end of the rod I'm currently boring out and it plays a good note already..."

Terry McGee did an experiment with pinewood and basically it was porous, that's on his site. Oiling adds the finishing touch to any flute also.

For warping it much depends how long, and how, the wood has been dried or seasoned, as well as what part of the tree it is from. It is a large topic, but usually most wood only warps slightly if at all. It is usually still playable if it does warp.

It sounds like a good idea to make a conical bore. I take it you read the way I use to make one, the difficulty with your idea might be that it is not easy to get a good idea of the shape of the bore, or to be able to measure it properly. Even with gauges, an undulating bore will give false readings unless a lot of different angles are tried. Still, even the best flutes undulate (intentionally or not) and are still praised. The accuracy is maybe more to be consistent with the result.


Here is a snapshot of my mini forge, I will write it up later. A £5 usb fan, some sheet metal, an empty blow torch cartridge, some clay and perlite for lining. From being lit it heats a 12mm rod to orange in five to ten minutes, then a couple of minutes for each reheat. A couple of handfuls of charcoal is all for a small project. I tried a mini gas forge but it didn't get this hot, plus it uses a lot of gas... some small furnaces use better torches that give results though.



Here it is in daylight... it actually heats to orange going yellow, a smaller rod to yellow. It allows me to easily put in a 90° bend on 5mm thick iron with a pair of pliers (or grips).



I had though of making something like this before, but wasn't sure, then recently saw this video

https://m.youtube.com/watch?v=eCc6SRxsisI

Which showed it possible for small work.

[David Cooper]

That forge looks the business: much better than the one in the video (though I like his anti-gravity machine). I'll have to find a reason to need to make one, but I now know who to ask for help with it if that need arises.

I've reached full distance with my bore, keeping it very close to the centre all the way along: the magnet-measuring method works perfectly. Fortunately I only need to go in by about 38cm, but if I ever do anything longer (like a quenacho) I'll have to switch to a longer auger to finish the job - it should be steered down the existing path with precision the rest of the way just by its own shape. The main advantage of the speedbor and magnet method is that you can make steering corrections early on which you can't do with an auger, but after about half way it would be a lot quicker to switch to an auger, so that's what I'll do in future. The speedbor is only just long enough, and I actually had to put something in the drill to push the bit 1.5cm further out to reach that full length of the bore and give me a bit of leeway to shorten the mouth end of the tube later, which isn't ideal, but anyone working on a really tight budget can achieve it with just the one tool.

SAFETY WARNING: This way of drilling the bore is potentially dangerous. I was doing it all by holding both drill and tube by hand with only the far end rammed into a place where it wouldn't move around. Anyone trying to do this should never hold the tube where the head of the speedbor is rotating inside it. That's a rule you must not break. I actually did do that once while pulling the bit back out with it running at speed to reduce friction, and I could feel the tube flexing in a weird way when the head of the drill passed my hand. With a defect in the wood, it could potentially shatter and put shards in your hand, so don't break the rule. Wearing a strong glove would also be wise.

I've sanded the bore a bit just by taping some sandpaper round a rod and running that up and down inside the tube, and it looks reasonably smooth. I'm not going to experiment with conical bores for now. Next up, I'm going to make a shallow hole at the far end, coming in from below just beyond where the bore ends, then I'll connect the two by drilling at 45°. I'll also make a decorative dent in the end where the hole would more conventionally be, just to make it look less weird.

By the way, is there any reason you know of for not using yacht varnish on a wooden flute to seal the interior?

Edit on 29th:-

Hole now drilled at end, and it worked well, apart from going through the bore so fast that it came out the other side. I'll have to stick a diamond in that one. Anyway, it doesn't matter for a prototype. The sound's okay this time: not quite as good a tone as my best bamboo 22mm bore quena, but not far off. The problem I had before with the notch was entirely down to the wood quality. Birch is orders of magnitude better, but I clearly need something more dense, unless the 16mm bore is just too narrow for good sound. Adrian Villanueva made some quenas out of papier maché and was getting top quality out of that material (and high prices too), so maybe that's a route to try - he didn't reveal what kind of glue he used for them, but whatever it was, once it's infused through the paper it could achieve higher density than normal wood. Anyway, I now have something I can work with to see if I can fix all the potential tuning issues, but I also need a slower drill.

[oleorezinator]

Everyone knows that it’s all done with mirrors.......

[GreenWood]

For a quena I would not know how bore diameter works, if it is the same as flute. Also when it comes to speaking about tone it gets very difficult. For flute narrow bore tends to favour higher notes, but there are just too many variables to generalise.

I think you could put whatever varnish you wanted, as long as it tasted ok. Seriously, I would go by only finishes you would be happy eating.

Polymerised linseed oil was the last I read of, there are also synthetic varnishes that are rated suitable for drinking water containers. You would have to research for yourself though because beyond natural oils, I don't think anyone will straight recommend any alternative.

I cannot imagine what glue he used for the maché flutes, some flutes are also made of ebonite

https://www.ellisflutes.com/blog/what-is-ebonite

[GreenWood]

Everyone knows that it’s all done with mirrors.......

Not for measurement, though I expect mirrors would give some idea, maybe enough for skilled crafstmen working freehand ? Otherwise calipers are the only way I know

https://www.historical-bassoon.ch/instr ... g-methods/