Well, looks like the 101st post in this long thread will be likewise long.
Anyone who has ever done any mathematical modeling in a real-world environment where the results have to be compared to experimental data knows it ain't perfect.
Engineers and scientists apply mathematical modeling for many reasons but almost never (outside a classroom) without careful comparison to experimental data. That would be the case here. I won't speak for Steve but the study I'd like to see is one that he actually has within is capabilities. It would combine results of FEA applied to a wire-frame mathematical model of a reed vibrating in a chamber under different loading conditions with data derived from high speed camera footage of an actual reed/chamber "wind tunnel" experiment under those same loading conditions (air pressure and velocity).
I thought it might be useful at this point to give a little background on what FEA is, how it is typically used in an aerospace example, and how we might use it to study harmonica reed dynamics.
In order to estimate the forces acting on an entire object, Finite Element Analysis (FEA) calculates the forces impacting each "pane" of the thousands that make up a wire frame model of the object. An example would be engineers designing a wing for a new aircraft. Long before they start welding, they create a wire frame mathematical wing and apply FEA software to simulate the wind impacting it.
To calculate the total lift of the wing at different air speeds, the software calculates the angle of the wind at each individual pane and then estimates the force impacting it. Combining this analysis from all the panes making up the top and bottom wing surfaces, they're able to approximate the lift of the complete wing. Then they can match the lift of both wings to the airplane's weight. They'll tweak the model until they get a design they think will work and then they test a physical clay model in the wind tunnel.
Generally there will be disagreement between the results of the mathematical and real world experiments. That disagreement is due to the inadequacies of the mathematical model. A good engineer will study those differences to improve his understanding of the system and create a better model. Only when the engineer achieves agreement on both theoretical and experimental bases are they ready to make an actual wing.
How might we apply this technology?
Steve will correct me if I'm wrong but it looks to me like the reed images above show the stress on the metal along the reed, showing regions with the highest stresses in red. Each image represent a higher mode of vibration exciting a higher sound frequency. You can see in the 3rd image where the stresses are distributed when the reed goes into a torsional vibration mode -- the kind that causes squealing on overblowing some harps.
It might be easier to relate to the motion of a drum head or circular membrane as a player hits it with increasing force. Just as Steve's reed, the drum head enters increasingly complex vibration modes.
In my little potential experiment, we would use FEA on a series of beam models of different initial arc or shape. We'd incorporate as many forces in our FEA model as we understand based on our collective real-world reed tweaking experience. We'll take reed shape measurements from the high-speed images and compare them with our FEA results. We'll tweak the FEA model until those result begin to agree with the camera measurements begin to agree with those from the FEA model. Then we'll know we are on to something.
What might we hope to gain from Steve doing all this work?
Will we magically jump ahead tens of thousands of reed tweak-equivalents in experience in our understanding of how to set up an initial shape of a reed? Prolly not.
Will we increase our understanding of the physics of the way air flows inside a reed chamber. Prolly so.
Will we gain a better understanding of how to set the initial arc of a reed in order to enhance its vibration efficiency so that the molecular level forces making up the reed are working with us instead of against us? I think most definitely so!
Will we all become "Rice" Spiers and suddenly open up Harpwrench II shops? Laughable! But will we advance our individual understandings of how harmonicas work in ways that enhance our harmonica technical skills? I think that's a definite yes. I feel that I've already increased my own understanding just by virtue of this thread.
Of course this is all blowing smoke unless we can convince STME58 to link arms and help us understand how we might help him access computer and high-speed camera time and the willingness to act as our engineering proxy to run the software and set up and perform the experiments.
At least that's the way I see it.
Steve? ;^)
Michelle
---------- SilverWing Leather - Custom leather creations for musicians and other eccentrics.
STME58 - Yes indeed. I think it's worth spelling it out like you have there so people see how scientists and engineers use these models in context.
There's one feature of your modes analysis I particularly like, which indicates that there is just one mode that is used in note production, the others would be squeal or some other noise. I have read the idea put forward that the harmonics of the note are created by the partials of the reed - as in a guitar string. I guess that it's so common in other musical instruments that I can see why people would assume that.
Re Helmholtz resonance - yes I bet it's in the mix somewhere. What struck me about the straw effect is that people use straws in experiments to create laminar flow. So if it were me, I'd be looking at that aspect.
If someone did do a scientific study of reed shape - I think the sensible first step would be to buy some of Joe's harps and measure and document the reed shapes he uses.
In my early days of harp I tried exploring reed shape, as I read people writing about it a lot. I couldn't find any principles that were any use to me - the optimum seemed to be a bit like the shape they came from the factory but perhaps a tad flatter, and even that is the simply the natural result of having a smaller gap. All seemed very subtle and nebulous - as opposed to gapping which was predictable and definite. So I've nothing to add on that score. I did consider (as now suggested by Joe here) a way forward would be to buy a custom harp - not to play but as a reference standard. I could keep it in a bell jar in a controlled atmosphere :)
---------- mtg - Blaring like a mad farm goose.
Last Edited by MindTheGap on Jan 23, 2016 2:26 AM
Michelle - trouble is, you go to all that effort and might find the best reed shape, with a solid theoretical base, then you've got the problem of actually applying it to a real world harmonica. How is that done?
I think you have to consider the possible outcome that there is no best reed shape - it might depend on the specifics of the player and their style. Correct me if I've misinterpretted, but that does seem to be what Joe is saying
If you put so much store by a player's resonance as dominant factor - and why not, it seems likely - then it's also likely to be different from one player to another, like their voices differ. Although I don't know why you've singled out Helmholtz resonance for special mention, there's all kinds of wobblings available in the body.
Myself, if anyone wants to apply some science I think it would be good to find an alternative material to brass reeds - one that you can play as hard as you like and doesn't fail. But then people will hanker for the sound of a real brass reed.
Last Edited by MindTheGap on Jan 23, 2016 2:19 AM
...overblow with a straw: the first thing I thought of when I heard about this was that people often use straws to create laminar flow in experiments. Indeed there could be multiple mechanisms but I'd look there first.
to the question of whether that which is good for one player is necessarily good for another... i'm pretty sure i dont know 'optimum', 'best', 'ideal'. i set reeds up using the widely - repeated wisdom of shutting the door at the same time, as best i can ascertain what that is. i seem to be able to make them play ok, and it seems those i do for other people are acceptable. i just set harps up to be 'good' according to what i think is good. i'd try and cater for a hard player, and if the person has no interest in overblows i dont bother much about that, but generally i'm paying the same attention to reed shape for everyone as i do for me... and when i think i've improved the harp, the customer always seems to agree
regarding reed material...ever since i put a brass reed in a steel harp i've wondered about the 'sound' of reed material. frankly i'm a bit dubious about it. i notice Greg Jones advocates repairing brass seydel harps with steel reeds, and seydel sell some harps with a mix of brass and steel reeds...and in the case of the steel harp i repaired with a brass reed (because i had it on hand) i could not pick the difference, once i'd adjusted it. ive had no feedback from the owner that he found it sounded wrong, and he has sent me more work since, so i guess he didnt hate it... ive used lee oskar reeds in hohner harps and couldn't notice a difference there either, except that it was in tune after i repaired it, it felt consistent to play and there was no tonal difference i could detect... all of which makes me wonder about the reputed tonal differences of various brass recipes..maybe i'm not listening carefully enough. i'll accept that
Superbee - Re the 'sound of the reeds', since this is a physics-based thread, I'm going to express the view that I can't see why the reed material would be a dominant factor. Because the sound of the harmonica is not a reed twinging about, it's the air being chopped about, then amplified by picking up resonances in the body. Isn't it?
However, if an experienced customiser or player tells me it does make a big difference in reality, I'd defer to that. I wouldn't know how to explain why though. The timbral qualities of other vibrating things (guitar strings, violin bodies) are affected by the thing's partials/harmonics. I think the reed's partials aren't relevant. But happy to be proved wrong.
These are all just thoughts but I can see why comb material might make a difference - lots of people say it does. Not because it vibrates differently so much as because the surface may reflect or absorb sound in a different way. I could imagine that might make a metal comb feel/sound lively and the opposite for a wood comb.
Last Edited by MindTheGap on Jan 23, 2016 3:46 AM
"There's one feature of your modes analysis I particularly like, which indicates that there is just one mode that is used in note production, the others would be squeal or some other noise."
When I put the back of a plate to my lips to test the sound and feel of a note, I generally play the note by drawing air at varying levels of flow to "feel around". (They are all draw notes because I am playing the plate off the comb.) I will bend the note to make is squeal. I'll also reverse flow and play the "overbend" and consider how that note feels and sounds.
I adjust the reed shape to make it as hard - or impossible - to make the reed squeal under those conditions. This is an important part of my process.
I think it would be just as important to model what makes a reed squeal as it is to model what makes the reed play well. Both things may not be the same and perhaps the "best" condition is when you can get the most of one and the least of the other.
As for different reed materials affecting timbre, I've tried replacing a single brass reed with a stainless one and I took it off - it stuck out like a sore thumb. I also obtained some older (yellow) brass reeds from Hohner and they definitely have a different sound (same pitch). I know Carlos Del Junco bought a lifetime supply of these old brass reeds in every size years ago because he prefers that sound so much.
SuperBee the method you describe has been used to create the gold standard customs for a couple decades. As long as you nail it in the tweak-to-fit-the-player department, then it's going to make them happy. Which makes it super practical for those tweaking their own. ---------- www.spiersharmonicas.com
Andrew, do you have any thoughts as to why the different reeds sound different? Of course it doesn't need to be explained for it to happen!
It's easy to imagine that different reeds with different materials, so different stiffnesses, but the same pitch might need different amount of air to get them going for instance. How about that? I'm keep to explore the idea that it's more about the air than the reed.
Last Edited by MindTheGap on Jan 23, 2016 5:06 AM
"Killa I'm certainly the mentoring type. Just about any serious tech will confirm that. Some of the ideas discussed on here and elsewhere originally came from me caring and taking the time to help others understand."
I can certainly testify to that. Pretty much all of what I've learnt about setting up my own harps over the years, has been thanks to Joe Spiers. His guidance, advice and more importantly,nudging me in the right direction, has been of immense help. Without Joe's input I'd have struggled for a long time to realise certain connections in setting up reeds. Of course there has been help and advice over the years from others too. The late Chris Michalek, Jason Ricci and BBQ Bob to name a few. The lions share of what I've learnt though has come from Joe. He has my utmost respect and admiration for his craft. He is the first person that comes to my mind when anyone asks about custom harmonicas. He is the direction I always point them in, when they want to buy them. Thank you for sharing some of your knowledge with us Joe. I for one certainly appreciate it, more than I can say.
Last Edited by Kingley on Jan 23, 2016 5:15 AM
Appreciate that Kingley. Andrew, sounds like an nice effective process to me. I think the different timbres of materials comes from the profiles associated with them and how they chop air. ---------- www.spiersharmonicas.com
"Andrew, do you have any thoughts as to why the different reeds sound different?"
"I think the different timbres of materials comes from the profiles associated with them and how they chop air. "
Joe, you may be right that is is the profile (as in mensur, not arc) that creates the difference. But if we assume that the manufacturers have optimized the profile to suit the material you can't separate the two. So the same pitch reed made of different materials will need a different profile to best optimize all of the desired characteristics (sound, response, durability, etc...)
In the end they are different and it wouldn't be practical to tease out whether it is indeed profile or material that changes the sound. You couldn't get a stainless reed with the profile of a brass one *and* get it to play at the correct pitch.
As for why the sound is different, I mentioned earlier when we discussed moments of laminar flow. I believe that some flow currents switch back and forth between laminar and turbulent flow as the reed goes through the cycle of passing through the slot. Although the frequency of the cycle is the same among reeds of different materials, perhaps the rate of rate of change is different and this causes a greater or smaller moment of laminar flow. The ratio of laminar to turbulent flow is greater or smaller because of this and consequently affects the timbre.
It's just a hunch. In general I believe it's more precise to describe breath force as flow rather than pressure although both flow and pressure gradient are coupled.
Is the purpose of this thread to identify equations that accurately mathematically describe the nonlinear dynamic processes involved in breath vibrating a harmonica reed to make sound? Is the goal to mathematically determine the optimal shapes and behaviors of harmonica reeds, slots, and reed chambers to be used in the wildly variable human created air flows and resonances involved in playing harmonica? Is the intent of this thread to find the equations that will improve harmonica playability?
Would arriving at some equations serve a purpose that has not been achieved by techs using the well established approach of successive approximation? It seems that many talented people have spent decades refining ways to improve performance by continually modifying reeds and reed slots, checking results, and resuming modifications. You don't have to have equations to improve playability.
Is this an intellectual exercise for those who enjoy trying to capture real phenomena with mathematical computer models, or a discussion on making practical modifications? So many assumptions must be made and factors discounted in creating mathematical models that I highly doubt useful results.
I believe there is still room for human artistry in harp tech, as well as in playing. It is fine to explore modeling these phenomena, just don't count on useful results.
Andrew I don't know about your laminar/turbulent flow thing (where's that come from?) but I can see that if the mass distribution is along the reed is different for different materials - including having a big blob of metal at the end of those low reeds - there is at least a mechanism for the different reeds to bend differently in the same conditions. I've no idea if that is what's responsible - nor does anyone else. But it is at least something a person could devise and experiment to test if anyone wanted to.
How about a very light but stiff reed with a big blob of solder on the end to make it vibrate slowly? I.e. all the mass up at one end. Compare with a thin-but- unstiff reed, thinned out at the end, at the same pitch. With a mix of materials and weight distributions - someone might actually find a 'better' reed :)
Just thought experiments though.
Dougharps - just speaking personally this an intellectual exercise for those who enjoy trying to capture real phenomena with mathematical computer models. I enjoy doing that. Re harmonicas, I just want to buy them and play them, not very well unfortunately. Currently I buy Lee Oskars because I don't need to adjust them, well I've not had to yet.
When I started playing I couldn't help but want to know the physics behind bending. I assumed someone would know, but I've not read anything that even attempts to explain it, beyond showing what the conditions are to make it happen, describing what happens and reproducing it with some apparatus.
Last Edited by MindTheGap on Jan 23, 2016 10:49 AM
Here's a completely inappropriate analogy: steel making. For centuries, people made tiny batches of fantastic steel using skills built from generations of experience, trial and error. Even into the industrial age, steel making relied on highly experienced people looking at colour of the mix and poking it with a stick, and throwing ingredients in to taste. Then at some point the science, understanding and modelling was good enough to do it instead - resulting in steels of undreamed of properties and consistency. I don't think it's going to happen with harmonicas.
But what if a nicely constructed model suggested, I don't know, a slightly different shape of channel in the comb? Maybe that would be worth it. Doesn't the Rocket have unusual shaped chambers? Sometimes these things can be found - think of golf balls with their dimples.
Last Edited by MindTheGap on Jan 23, 2016 11:10 AM
MindTheGap" "When I started playing I couldn't help but want to know the physics behind bending. I assumed someone would know, but I've not read anything that even attempts to explain it, beyond showing what the conditions are to make it happen, describing what happens and reproducing it with some apparatus."
Yes, that and the fact that it occurs regardless of any original or intentional design for harps to actually bend. It's truly unbelievable.
note: MindTheGap is a funny usename for someone who buys Lee Oskars because they don't need to adjust them.
Last Edited by Littoral on Jan 23, 2016 11:45 AM
Littoral - yes indeed. It's a very odd thing - one of those happy accidents. I remember reading manufacturer's blurb saying things like 'extra thick reed plates for easy bending'. Is that actually even a design thing?
Ah yes good point... MindTheGap. I chose it when I was into gapping and tweaking - I found all my SP20s needed some just to make them respond evenly across the board. I should change it to Don'tMindTheGap :)
I was amazed when I put that brass reed in the big six that I could not pick the difference. I really expected it was not going to work but I tried it because it was the only reed on hand that was right size and pitch. At first I thought it was in fact a bad idea, as the bend in that chamber was weird, but I adjusted the reeds as I normally would and to my surprise it normalised. I decided I could not tell there was an odd reed and my primary concern after that was whether the repair would hold up. Maybe I'm not really listening. I was trying to, but I suspect I don't have the 'best ear' around...like many harp players apparently! I have a couple of G marine bands here. One built by Joe,one by Mark P. They are both made in the same reed era afaik. Both have corian combs. I can instantly tell the difference in timbre between them. I'm not sure why that is. Maybe embossing. Maybe that is a more obvious difference? I can definitely notice a difference when changes are made to slots, or when a reed plate is thicker. I've also got a prewar marine band in G. I can see the reeds are milled differently to a modern reed. I guess it sounds different but it's still a work in progress so yet to be fully tuned. I wonder if players of the Seydel Solist Pro 12 feel there is a discernible timbal change when moving between the adjacent steel and brass chambers. My sample of experiences is very small. I'm not running a lab here, just doing repairs and learning on the job. I've often read about timbral qualities of different reeds but I've not specifically set out to verify it. Just a couple of observations which have led me wonder. Wrong thread I guess
Actually that's decided my first job for the day. I have a solist pro here which is in for a reed replacement. I'll try the steel reed and see what I think. Maybe my ear will be more acute today. Rats! It's actually session steel reed plates on a solist pro comb and covers...oh well..
Last Edited by SuperBee on Jan 23, 2016 1:53 PM
Superbee, the biggest single difference I've heard in a harmonica was when I swapped a Piedmont cover onto a Sp20. It dulled the sound something terrible, which was a disappointing result for me, since, once the terrible taste has worn off, I really like the feel of the Piedmonts on my lips- really easy to practice a lot with without ripping and shredding your lips to pieces. If you think about the sound waves coming out of the harmonica, the back end of the reed chamber is blocked, so it's going to go out through you, or through the gap in the back between the reed plates and the cover plates, and the first significant surface most of that sound is going to hit is the cover plate (and then bounce down off the reed plate). I even wrote Hohner and suggested a hybrid, a metal comb with the lip/cover plate boundary covered in some sort of plastic.
Mind the Gap, on the topic of golf balls- I've wondered some about the embossing on cover plates. It seems to me that they might work a little like golf ball dimples, at least as long as the embossing doesn't have any sharp edges. Your lip slides along, and then for that brief moment it's not actually in contact. It might even allow some air to slip in between your lip and the cover plate to create a buffer, but that's an entirely different set of equations!
Not to complicate things, but I'd also be interested in knowing a bit of how magnets affect reeds. In my head, if I'm picturing a steel reed swinging I'd imagine as it swung towards the magnet it would accelerate, and as it started to swing away it would do so less quickly until it got away from the magnets pull, sort of a stutter step. Apparently it slows it down enough to drop it a half step the way they are placed. I assume that's the entire effect although it's probably a little more complex. Moving through a magnetic field would, I would think, magnetize the reed a bit which might make it more complicated. I had an idea for an unbending reed once that would use a pivot point instead, like a teeter totter, with a magnet on the base end of the reed that swung towards two other magnets, faced top and bottom to repel it. The idea was a reed that wouldn't get fatigued, and that by slightly adjusting the height of the outside magnets you could tune the reed. That's the sort of crazy thing you think up when you play with magnets and don't actually have to produce workable results. (and it wouldn't even have to be a reed shape, really, just a valve of sorts).
On the topic of laminar flow and such, I have a toy my brother bought me. It's a whistle, complete with it's little ball that whistles have to modulate their pitch (I think), but then, farther down it has a harmonica like reed with a little plastic bird on it that swings through a slot. It's too leaky to be of any real testing value, and only a single reed system, but it makes me think that perhaps you could, fairly easily, rig up something to disrupt the airflow and see what sorts of results it has on the reeds. My first thought would be to make a variation of the straw test, but a ball with a bulge in it, kind of, I guess, like a tiny bong, and put some small balls of paper or whatever in the bulge that will jump around and interrupt the airflow and chop it up before it gets to the harp and see if the straw overblow trick still works.
And again, going off on a tangent, for reed durability, I still would like to see a model that shows how much flex a steel reed goes through to see if it's under it's fatigue threshold (at different volumes) to see if my anecdotal evidence soft players can play steel reeds 'forever' but hard players will blow them out really quickly (that's just the results I've tabulated informally in my head about complaints I've seen here and on FB about how long Seydels last.)
Very interesting discussion. As Superbee says, it's all about achieving the goal of "shutting the door at the same time" - ie. getting the reed to close evenly along its length without leaving a gap at the root end.
But there is a big assumption underlying the discussion that no one seems to question: that the only way to do this is to bend the reed lower in the slot at the base (lowering the zero point it's sometimes called), and then re-shape.
As Joe Spiers says, this is how every name customiser achieves the desired result, and has been shown in detail in Richard Sleigh's videos, as well as others online. There are variations in tools and methods, but the process is basically the same. It's widespread and accepted because it works.
However, there are some negatives to this approach. It lowers the pitch, it's time-consuming, and bending the reed so abruptly at the root must necessarily weaken it, as it would with any spring (which is what a reed is).
There is an alternative that avoids the negatives and can achieve the same even reed closing: build up the reedplate at the sides of the slot around the first quarter of its length from the base.
There are two main ways to do it: raising the reedplate brass itself, or adding extra material in the critical area. With most brands of harmonica, using a microscope and a suitable tool I find I can micro-chisel brass shavings that rest up against the reed sides, effectively raising the slot to the reed rather than lowering the reed to the slot. I do about 3 chisel marks per mm, so they produce a virtually continuous line of shavings that close the gap at the root of the reed. The reed makes its own path through them when plinked, and closes evenly when looked at from the side.
However it's not easy to control on all reedplates, and the reedplate brass of Hohner harps is particularly tricky: the shavings tend to break off easily. For them I use micro-applications of a foreign material along the sides of the reed to build up the perceive height of the plate and close the gap at the reed base. It's also applied under a microscope.
I'm always experimenting with materials. Some are liquids that wick into the gap and harden, the bond later to be 'cracked' when the reed is plinked. This gives a really close fit naturally; nail polish works well as long as the reedplate is really clean (otherwise it will flake off).
Another way is laying a narrow strip of see-through tape over the reed bases and then carefully cutting it under the scope, using the reed sides as the guide. This can give a very neat and tight-fitting gap-closer too. It's similar to Jim Antaki's Turboliner idea, but specifically for the reed bases only, and the tape gets much closer to the reed so it actually does something (the Turboliner doesn't, in my opinion, because it is too far from the reed).
If you get one of these methods right and emboss the rest of the reedslot you achieve a very crisp-sounding responsive reed, because you are achieving the same end as via reed arcing/shaping: even closing of the door. But it's 'non-destructive', as the reed is not bent out of its original shape at any time.
Does anyone else use this alternative approach to solving the problem?
Last Edited by Brendan Power on Jan 24, 2016 5:35 PM
No worries. I thought it was funny to be told I was 'full of shit'. I don't hear that very often. :-)
Have a good day. Mark ---------- I'm out of the Biz for a while till I get over my burnout. You can try HarveyHarp or arzajac, or just look the page nacoran put together under Forum Search. .
Michelle, Im still learning set up etc I was really into working on harps a few years back--Ive mellowed since then--if it works don't fix mentality or lack thereof--you probably have ruperts video ,where he gently pushes the rivot end down==If I see a lot of light at rivot end [more than the other reeds] and it is airy or won't bend and or ob ,or close so I can overblow, then I push it down -but definitely not way down into reed slot. so that reed has to come back up to be level with reed plate again--as I said if u look at a manji or even crossover the reed is very close to reed plate ,with very little gap at rivot end--more so on manji===what I don't get is how one harmonica if u look at reeds, say hole 5 ---from one harp to another 1] the gaps on the sides looks tight, ] ==2] the height looks the same on the 2 harps==========So lets say u have 2 melody makers both key of C---the gaps look exactly the same ,but on one ==the blow reed closes and the draw pops ob easily--the other doesn't--I mean even putting your finger on the blow reed, so its closed and seeing how easy the bottom will open for ob-they both seem the same ==then even using the same comb on both harps and trying ob -- one goes easily one doesn't--its maddening-----these days as long as the harp plays well---will easily ob 6 sometimes ob5 occasionally 4ob and plays without jamming up im happy==some will od 7 for b5 purely hit and misss-I gave up trying to set up 7 for od-------------------------------------------[I LEARNED THE HARD WAY IF U CAN'T GET 1] THE CORRECT EMBOSS, NO BURRS----2] CORRECT HEIGHT FROM REEDPLATE--3]-REEDS PASSING THROUGH SLOT EVENLY AND CLOSING IT EVENLY ETC---------IF CANT GET IT SOON B4 OVERWORKING THE REED YOUR BETTER OFF PAYING A POLISHED PRO-----ZAJAC SPEIRS ETC-------------------------BECAUSE OVERWORKING A REED IS ALMOST THE SAME AS BENDING A THIN PIECE OF METAL ---BY-WORKING AND BENDING THE METAL BACK N FORTH TILL IT TEARS FROM WEAKNESS AT THE BEND JOINT --MY GOODNESS WHY DID I MAKE A SHORT BORING STORY INTO A LONG ONE---IM NOT MUCH BUT IM ALL I THINK ABOUT PEACE
MP ... Lol. Good. It's just how i talk, to everyone. I have to dial it down in here tho, but i didn't mean nothing by it.
But atleast i say what's on my mind. Instead of making smart chicken shit comments behind my computer like some folks. Peace. ---------- "Trust Those Who Seek The Truth. Doubt Those Who Say They Have Found It."
Last Edited by Killa_Hertz on Jan 24, 2016 8:12 PM
I have been reading over all of the comments and I am struck by the wealth of experience possessed by the contributors to the thread. It would be really cool to find some new insight in to how a harp works, but it would require great effort and/or great luck. It is very easy to get bogged down with fancy equipment and analysis tools and never make any progress. Both MindTheGap and Dougharps have pointed to the importance of clearly defining what you are attempting to do before dragging out and firing up the tools. This is good advice.
I have wondered if the goal of the reed all entering the gap at the same time is really what is desired. If I understand it correctly, the reed acts like a valve and it's opening and closing creates the pressure waves we hear. It never closes all the way, but it seems that a lot of customizing work involves trying to get it to close as far as possible. Brendan's post really emphasizes that. I have been wondering if the area of the opening from the slot edge to the bottom of the reed as it closes from the top would relate closely to the flow through the reed and therefore the pressure. As Andrew has mentioned, flow and pressure are closely related. If this is the case, the area of the reed opening with respect to time should very closely match the time pressure curve as measured with a microphone.
I will have to think about this a bit more but I thought I would put it out there to see what others think of the idea. Is anyone aware of any research in this direction.
Last Edited by STME58 on Jan 24, 2016 11:55 PM
The other modes do indeed contribute to the overall timbre of a particular reed pair and chamber. In fact, I think the other modes excite overtone frequencies. Different reed materials (and their associated different mensurs) result in different overtone frequencies, which account for the different timbres of the different brands and reed materials. If the reed vibrates at a different frequency it chops air at that different frequency. Here is a figure from one of the physics of free reed scientific abstracts on dougharp's link. It shows the amplitude peaks corresponding to the frequencies excited by the different modes of vibration of an accordion reed (same thing occurs with a harmonica reed or any other free reed.
Your analogy of a blob of solder on the tip of a reed is the same thing as changing the effective reed profile
As Joe and Andrew mention, there's no doubt that reed mensur (profile) also influences the natural frequencies of reed vibration and the resulting overtones. Which is the primary and which is the secondary effect I don't know. FEA could tell us, though. I don't believe flow cross-section (laminar vs. turbulent) has a significant impact on timbre, but that's just an opinion. I think it more likely to be a matter of reed makeup and profile.
Helmoltz resonance is afoot but it occurs inside the chambers so it seems that it would therefore predominantly affect only draw reeds.
Since Joe Spier's reeds exhibit the current pinnacle of reed shape, that would be an ideal starting point. And, a rigorous experiment would likely end up with a shape that would be very similar if not identical. Wouldn't it be great if our experiment verified that Joe's reed shape is indeed ideal Wouldn't it be even better in the grand scheme if it didn't? As I've said before, that's when we would learn something new that advances the art. Everybody, especially Joe, please take note that I'm not knocking Spiers reeds in any way, shape, or form. I'm trying to design a valid scientific experiment without systematic bias.
As to applying the results of such an experiment, I'm not proposing to derive a reed shape de novo with FEA, call it good and shape reeds like it henceforth. It is often the case in designing complex systems that engineers use theoretical computer modeling to seed real-world experiments. What I'm proposing is to first use FEA to test hypothetical reed shape(s) that we think will enter the slot all at once. When we're satisfied that we have a good theoretical shape we'll create a real reed to mimic it so we can test it in an actual, not virtual way in a wind tunnel. That is how we'd tie the experiment to reality. Procedurally we'd take high speed camera images of the reed to see if the initial shape we calculated results in the reed indeed all entering the slot at once. If not, we'd measure the amount of deviation from perfectly flat along the reed at the time of entry into the slot from the images. We'd then change the shape of the resting test reed by a fraction of that difference and repeat the wind tunnel experiment. If the reed still isn't flat when it enters the slot we repeat the process of measuring how much it differs and changing the shape of the resting reed. This is a common method called iterative solution of equations using what's called numerical analysis. I didn't invent it and it's used to solve far more complicated problems than this. Quite honestly this brute-force method of solving equations could start with a straight reed. Starting with a good guess, whether it be computer or human-derived just reduces the number of iterations required to converge on a solution. Eventually the deviations diminish to the point where there's no practical point in continuing to iterate.
We could argue laminar vs/. Turbulent flow 'till the cows some home. It would be nice to know but not necessary to conduct a wind tunnel experiment.
Grey Owl, The Seydel/Sjoeberg video is but a teaser. It tells you little about how to know when you've done something right. That stroboscopy is great, though. It really shows what's going on. If we had views from the side we could calculate reed curvature from them. One of the things that strikes me from these images as well as Steve's and the ones from TurboDog's web site, and that is how asymmetric the displacement of the reed is depending on whether it is opening or closing. It makes sense that it would open further than it swings on the closing cycle since that's where the air pressure is coming from. But it wasn't intuitively obvious to me.
Andrew, your statement, “I adjust the reed shape to make it as hard - or impossible - to make the reed squeal under those conditions. This is an important part of my process.” sort of goes without saying. But since you said it I have to ask what you mean. Seems to me we know that torsional movement causes squealing. So anything that would limit torsional movement would also limit squeal. I think that's why nail polish works. It dampens torsional vibrations.
As I said above, I'm confident that the reason different reed material sound different is the overtones secondary to the higher order vibration modes.
Doug, I strenuously seek to avoid leaving the impression that any effort we could muster would make it possible to derive comprehensive equations for optimum reed shape. Only to seed the real experiments in the wind tunnel. However, FEA could be used to model different aspects of the physical problem as to why the experimentally derived reed shape works. But, that's for some PhD thesis on done the road.
MindTheGap, the Wright Brothers certainly didn't use a computer to design their wings. But don't think for a moment that we could have advanced wing building technology to the level of the Space Shuttle's wings without them. We'd be shooting arrows from long bows instead of recurve bows without computer modeling. Counterexamples to your claim abound.
Doug, I had never heard of Dr. James Cottingham's work. Interestingly, none of it has directly to do with harmonicas. He writes about organs, accordions and a variety of Asian free reed instruments, but not harmonicas. Also, almost all his work is I the form of scientific abstracts or slides given at scietific presentations. Those don't receive stringent peer review so that may be why he isn't quoted as much as the other researchers. None the less, as you can see from the figure I stole from him above that there is some interesting content there and I'm grateful for you sharing it.
Nate -
It makes sense that changing cover plates makes a big difference in how a harp sounds – to a player. I need to be convinced that a listener could tell the difference. Same thing with vented vs. non-vented cover plates. Interesting point about the aerodynamics of golf ball dimples, but not with repect to embossing. That creates sharp edges fer sure. I do wonder what the effect of dimpling on the surface of a reed might be, though.
The only reed material with a significant magnetic moment is steel. Brass and Phosphor Bronze don't have significant magnetic moments to be affected by external magnetic fields. You are probably aware of Dr. Antaki's Turboslide harmonica that works using magnets on a Seydel steel reed harmonica.
The straw thing. It doesn't matter what the flow cross-section is inside the straw. As soon as the air flow hits the harmonica reed channel, it goes turbulent. The effective aperture of even the open reed is too small for it to be any other way.
Interestingly, a siren also works by chopping a flowing column of air. It's a rotary aperture though, with a plate with a hole in it rotating over another plate with a hole.
Uh oh. Hit the character limit. Continued next reply...
Last Edited by mlefree on Jan 25, 2016 10:59 AM
Brendan if we can count on your for anything, it's to think outside the box! You probably change lightbulbs differently from everyone else, too. ;^)
Yes, let's raise the track to the locomotive! How about machining a depression in the reed plate the depth of the desired zero point, right at the reed pad. The reed would then enter the slot below the top surface of the reed plate without a dog leg bend at its root. What shape the reed takes from there would be up to the customizer. One could machine a channel along the locations of the the rivet holes in a bare reed plate and then mount the reeds with their pads down inside the channel. It would be difficult machining a variable depth channel if you want each reed to have its own zero point. It would be a manufacturing nightmare but it might work for the customizer with a milling machine.
I always try to keep brass shavings as far away from the reed slot, especially at the base of the reed. Applying some kind of sealer and plinking after it dries seems pretty foolproof as long as you use the right sealer. I'll try to think of something that will stick to the reed plate better than nail polish.
Steve, we're definitely talking the 90/10 rule here. It takes 90% of the effort to extract that last 10% of functionality.
To address your question about the areal change of the reed with time through its vibration cycle you might take a look at the Cottingham abstracts that Doug provide the link to. Otherwise I need to sit down when I get a couple hours and calculate just that. That's just the information we need to know to calculate Reynolds number to put the laminar vs. turbulent flow question to rest.
Again, Thanks to All for all the fine thinking and time invested behind these great replies! If nothing else, we're putting lots of reference material and fine thoughts on the question of reed shape in one place. This thread may help someone find some answers later, if not sooner.
Michelle
---------- SilverWing Leather - Custom leather creations for musicians and other eccentrics.
Last Edited by mlefree on Jan 25, 2016 1:20 AM
Michelle - Re reed modes, that image doesn't show for me I'm afraid. But if it says that the reed itself has harmonic modes then it means STME58's analysis is wrong. His shows one mode that makes the music and others that are something else, possible squeal.
If in fact there are harmonic modes in the reed itself, that's is interesting and would be easily enough to explain why different materials would have a different timbre.
However, I don't see why a reed would have harmonic modes. It's not a guitar string or a column of air in a trumpet - those systems have boundary conditions that give that. The resulting sound can have harmonic components, but that's a different thing. A drum skin doesn't have harmonic partials naturally. Neither does a bell - you have to build them in - they even have specific names: Hum, Prime, Tierce, Quint etc. And yes I did have to go and look it up to remind myself :)
Anyway it would be interesting to see. If the reed does a have all these harmonics then that opens a whole new box of tricks. That slow-mo strobe video doesn't seem to show it - it just looks like a reed bending to and fro in one arc. Harmonics would look like ripples - the 2nd harmonic (aka 1st overtone) would be have a node (a static point) somewhere in the middle, and would have to have precisely twice the frequency. You know - like when you play harmonics on your guitar. Ok the strobe may not capture it - but I'd expect to see it as a kink in the reed at least.
The air raid siren has harmonics as a feature of how it chops up the air, not because the plates with holes in have a some kind of harmonic vibration.
I've measured the 3rd harmonic as being particularly strong in the harp sound - as strong as the 1st harmonic if you go for a shrill note, and if it were the reed modes providing that, I'd expect to see it happen.
Guitarists and bassists love their choices of guitar string partly because of the different harmonics content. Imagine that with reeds.
Anyway, thoughts and words, if the picture shows harmonics in the reeds I can forget all that. :)
Last Edited by MindTheGap on Jan 25, 2016 2:22 AM
Re the straw-overblow thing. I'm sure the flow may either be laminar or turbulent at different places. But what effect does that have on anything? Does it even matter? I have no idea. My point was that it has been asserted that to overblow you have to have a resonant cavity in your mouth. Seems a reasonable assertion to me - that's certainly what it feels like. But then some say they feel they are bending from back in their throat - there seems to be actual evidence that it's done with the tongue. So feelings and reality may differ.
There is evidence that you can overblow with a straw, so how does that fit with the assertion and feeling of a resonant cavity? STME58 suggest two different mechanisms. Something to explore.
Often it's odd effects like this that give people a toehold in understanding the thing overall.
Last Edited by MindTheGap on Jan 25, 2016 2:30 AM
"the Wright Brothers certainly didn't use a computer to design their wings. But don't think for a moment that we could have advanced wing building technology to the level of the Space Shuttle's wings without them. We'd be shooting arrows from long bows instead of recurve bows without computer modeling. Counterexamples to your claim abound."
I'm not sure what claim you are talking about, I gave the example of steel making where modern materials science took over from centuries of experience and now rules the roost. Money usually drives these things, as in aviation, and I don't think there's enough money in harmonicas!
As for longbows and recurves, well, fortunately you've happened upon something I really do know quite a lot about. Or maybe it was by design and if so, thank you! The concept of recurving to increase cast and reduce bow length has been around for centuries :) They reckon Odysseus' bow (the one no one else could string) was curly and so needed special skills to string. You should have gone for compound bows - but even the they started in the 60's.
But yes, even in archery, modern materials science has made a big impact. They put layers of bamboo in long bows these days - nature's carbon fibre!
Talking of which, if anything thinks harmonica players are obsessive about their kit and disputing endlessly and pointlessly about how it works - go and chat to some archers :) That will put it all in perspective.
I'm guessing Michelle that you make things for archers? - they love their leather goods.
Last Edited by MindTheGap on Jan 25, 2016 3:08 AM
It would not be too hard to verify the modes I posted. The reed I measured was out of a Seydel Session Steel in Low D. I believe it was 4 blow, which should have been the note D at 1175Hz in A = 440 tuning. The first mode from the FEA was 1110 or closer to C#. Making a recording and doing an FFT to get a graph like Michelle presented about would tell you if any response from the predicted modes can be detected in the spectrum of the reed. Most of the predicted modes are out of the range of human hearing, so a good mike will be required to capture them. I still have the harp. I will try to verifiy which reed I measured and make a recording.
I think your model is likely to be right. The way to test it is to pump pure sine waves at the reed the predicted frequencies and see if the reed picks them up and vibrates like in the pictures.
Michelle says she thinks "...the other modes excite overtone frequencies". As I remember, the mode frequencies were not in a harmonic series 1110Hz, 2220Hz etc. And I can't see why they should be. What are the boundary conditions to bring that about - it's a not guitar string?
If you measure the output from the harmonica, you will see harmonics 1110Hz, 2220Hz etc. My expectation is that comes from the way the air is disrupted (chopped). The signal looks like getting on for a square wave - so that's just what I'd expect to see.
A good experiment would be to rig up a driven system that chops the air in similar way. Even if the driven 'valve' emulating the reed is driven with a pure sinusoid I'd expect to hear harmonic content in the sound. Like the air-raid siren doesn't sound like flute.
I wonder if the confusion is around the observation that the reed seems to be vibrating like a simple harmonic oscillator. That's what the strobe pictures look like to me. But that doesn't mean it's disrupting the air in smooth sinusoid does it? All the time it's going through the plate, which is much thicker than the reed, it's a valve thats 'OFF', for instance.
Last Edited by MindTheGap on Jan 25, 2016 9:10 AM
Brendan, for scaling that up to industrial levels I wonder if 3D printing might work... I know there are more expensive models of 3D printers that will work with metals. The problem, of course, is that 3D printing is relatively slow compared to CNC and really slow compared to stamping. I don't know, however, if the process they use for 3D metal printing works to add something to a preexisting part. I'd imagine it would. There might be some issues with preparing the surface. (And maybe you wouldn't even need it to be metal?)
If you put a reed plate in a 3D printer and just printed a raised ridge near the base it might automate what you are talking about. Or perhaps Michelle is on the right track with lowering the mounting on the reed. (I'd wondered about that too). From the point of view of machining it that would seem to be an easier solution. It's easier to take something away than to add it (for now). You could, theoretically, do the raised ridges by subtraction as well, but you'd have to start with a thicker (and therefore more expensive) piece of brass and end up with more waste.
I wonder if it might be possible to stamp a raised ridge, after all, coins are stamped and have raised parts.
Mind the Gap, talking about the air-raid siren is making me think a bit about how the sound is actually being shaped by the reed- air is, if the reed is going through the gap evenly, starting at the same time, but there would be less air going through the gap near the base compared to the tip because of the relative amount of swing between the free and anchored ends. That might round the wave some?
P.S., Michelle, yes, I'm aware of the Turboslide. In fact, I've got one. That's what got me thinking about the magnets. Of course Brendan went a different way with his enabling reeds.
And another thought, in relation to Brendan's raised ridges, on the blow reeds, might it be possible to do something on the comb? It would be tricky, but since a lot of the combs are being CNC'd already, could you add the raised ridge on the comb instead? I know most of the combs are cut with a straight down thrust, but if you came in from the end you might be able to leave a narrower hole for the blow reeds. (Of course, I've wondered about mounting reeds directly to a finely cut comb. I know that would be a reed working nightmare unless the comb maybe opened up into top and bottom (more or less) halves, but it seems it would also solve issues with losing air between the comb and reed plates.
Nate - I have no idea, I expect it's pretty complicated. All I'm saying is there is a mechanism for making it choppy not smooth - smooth would be a special case and probably very difficult to attain. Just to add the to the mix - I think Andrew has said that embossing tend to make the timbre more bright? Bright = more high harmonics, wherever they come from, they're not coming from the changing the reed's characteristics in this case! Obviously I have no idea what's going on (I mean in with the reed, not in general, but that too if you like).
The stated aim (I read here) is to 'Shut the Door' meaning to have the reed closing as smartly as possible. Maybe that's why it sounds/feels good: the more on/off valve like the better, richer harmonics. Something that someone could investigate if minded.
What I do know is this is a long thread. What is the longest thread there has been? Aside from For Sale and Where are You From? Are we close??
Re your magnets idea - it doesn't need to be ferro-magnetic material. What about the effect of eddy currents in the brass? Probably nothing much, forget it! :) Just I know you want to get some magnets in there somehow.
Last Edited by MindTheGap on Jan 25, 2016 10:01 AM
I believe I have likely passed beyond the boundaries of my comprehension in these matters, but that has never stopped me from asking questions, sometimes irrelevant questions.
In pursuing this "vibrating reed in airflow" knowledge and creating non linear mathematical representations of these dynamic processes (including turbulence!), are we in an area in which chaos theory comes into play? Might some of the modes that have been mentioned, including those with torsional squeal or even choking a reed to allow overbending, be linked to chaos theory strange attractors?
Is this a situation in which there is sensitive dependence on initial conditions, where the smallest difference can have large consequences? ----------
Wow, what a vibrant and interesting thread this has become! Such a wide range of ideas and experiences. It's exactly the kind of thread I love to see on harmonica forums. And, it's neat that it's happening on Dirty South. Kudos and Kind Thank Yous to all of ya's.
MindTheGap -
I'll say this about harmonic modes of vibrating structures and then let it rest. It really isn't a subject open to debate.
Whether it is anchored on both ends, just one or none, a rigid object can be made to vibrate. When it does the physical characteristics of the object (density, rigidity, size, shape, etc.) and its method of attachment define the frequencies at which it “wants” to vibrate. The energy at other frequencies is attenuated and is mostly absorbed at the molecular level. (Important to our discussion of reed harmonics, the object still vibrates at those frequencies, just not as much, as shown in the above plot of the spectrum of reed vibration.)
Describing the vibration involves the well-known mathematics of linear algebra which enables the calculation of the mathematical entities Eigen vectors and Eigen values which define the “natural” frequencies at which the object will vibrate. These are these called harmonic frequencies. As you apply gradually increasing levels of exciting energy, the reed “jumps” to the next higher harmonic frequency (it's no accident that its called a harmonica. ;^) ). The different modes of a reed's vibration are visible to the casual observer of Steve's FEA simulation images. Further, those images show the regions along the reed where the stresses occur. Not surprisingly, the stresses are distributed where the reed bends the most.
Now, these different vibration modes give rise to harmonic overtones in the resulting chopped air stream. Even though the reed wants to vibrate at its natural frequencies, it's still vibrating at non-eigen frequencies. It's those non-eigen frequencies that give rise to the sound signature, if you will, of that reed. As to an experiment, you do it all the time when you plink reeds. No need to excite the reed at its natural frequency. A good plink is the physical equivalent of a mathematical impulse or Dirac delta function which excites all frequencies. You plink a reed an it will vibrate at its natural frequency.
You mention the Helmholtz effect. It is also incontrovertible that a player must adjust his/her oral cavity to make any particular reed sound. Those Cottingham scientific abstracts and several of the other technical references in this thread mention this.
You're absolutely correct about bow technology I misspoke when I said recurve bow. I meant compound bow.
As to my leather, I mostly make harmonica cases but I living where I do I also make modern and cowboy action pistol holsters, gun belts and knife sheaths. I haven't tapped into the archery market. Clearly I need to educate myself about it. Thanks for the tip!
Nate and Brendan, I have a little vertical milling machine I used to make x-ray test objects with. Maybe I'll try to make a nice little divot in an old reed plate for kicks one of these days.
Doug, I know little about Chaos Theory. Reading up a little it makes sense that it also applies inside a harmonica. If you know how Chaos Theory might apply here, I'd loved to know more!
Thanks,
Michelle
---------- SilverWing Leather - Custom leather creations for musicians and other eccentrics.
Last Edited by mlefree on Jan 25, 2016 12:18 PM
Sorry Michelle but I don't understand any of that. Steve's model clearly shows that the reed's modes are not a harmonic series (1110, 6714, 10935Hz) - Ok mode 2 could possibly fill in for the 6th harmonic. But the sound of the harmonica includes 2nd, 3rd, 4th... harmonics. How do they come about? You can force the reed to vibrate at the 2nd harmonic? That would mean you can make it play it's octave? I can't do that - I can do it easily on a guitar string or a flute.
Actually I was surprised, I could have imagined that mode 2 might have been close to the 2nd harmonic. But it isn't.
As for resonance. Yes resonance, but why specifically Helmholtz resonance? And you can overblow with a straw (sorry too bring that up again). Where's the tuned-resonance chamber in that setup? I think you can use the same length of straw on different holes.
Last Edited by MindTheGap on Jan 25, 2016 1:08 PM
Warning, the article and concepts may or may not apply. The article was tough going for me. My scientific education was in a different century, before chaos theory was taught at the undergrad level. ----------
Doug S.
Last Edited by dougharps on Jan 26, 2016 1:58 PM
MindTheGap, the problem of a vibrating harmonica reed is very different than that of a vibrating guitar string in many respects. The geometry of a harmonica reed is not nearly so as simple as that of a guitar string. It has non-uniform profile (mensur), rectangular cross section and is cantilevered being anchored at only one end. As a result, its harmonic frequencies aren't simple multiples of the their fundamental frequencies. Steve's results and Cottingham's are in agreement with each other and are consistent with theoretical expectations.
I'm curious, if not Helmholtz resonance, what other sort of acoustic resonance of (nearly) closed cavities might you be thinking of?
As to the straw and overblows, I don't know what all is afoot there. From the Bahnson article we know that “Overblows and overdraws are examples of things not being exactly what they seem on the harmonica... The pitch of both overblow and overdraw tones is outside the interval between natural pitches of the blow and draw notes of the hole, in contrast to bends which are almost completely within these intervals. Hence, there appear to be three modes of useful function of each reed: closing as in the natural modes of simple blow or draw, opening as in bending with the pitch within the interval of the two tones, and an overblow or overdraw as an opening reed with a pitch above that interval.”
It could be that the straw focuses and directs the player's air flow into the reed chamber in such a way as to override the tendency to go into normal Helmholtz resonance by choking the closing reed and causing the opening reed to go into the anomalous overblow mode. But that's a guess. If you figure it out, please let us know. 8^)
Thanks,
Michelle
---------- SilverWing Leather - Custom leather creations for musicians and other eccentrics.
"its harmonic frequencies aren't simple multiples of the their fundamental frequencies". Then they aren't harmonics, by definition. That's what I'm saying: it appears that the reed's modes are not harmonic. (or if you insist, it's eigenvalues).
But the sound from the harmonica does contain harmonics. I may be absolutely, totally, hopelessly wrong but I've suggested a mechanism that can create them - the reed chopping air. Like the air-raid siren. It's a simple idea - chop air flow with a valve and you'll probably get a sound wave that doesn't sound (or look on a scope) like a pure sine wave. This means that it contains harmonics. Someone can test that. Really, I don't claim this is some radical clever insight.
But if I am right, then the manufacturers and customisers could do things to change the nature of the chopping, like building up the walls of the reed slot, as Brendan is doing. Sounds like they are doing it already.
Re resonances, it's not my area, but from what I've read the resonances associated with speaking and singing are well explored and understood. I think any investigation - and this is what this thread is about - should at least look at these. There's all sorts of things going on, here's a start https://en.wikipedia.org/wiki/Vocal_resonation
Maybe all the resonators listed there are Helmholtz, in which case great. I always took a Helmholtz resonator to be have the special features of a cavity with a hard surface with a neck or small opening.
Just a quick quote from that article (for the purposes of naming some other resonances I might be thinking of)...
A conical shaped resonator, such as a megaphone, tends to amplify all pitches indiscriminately. A cylindrical shaped resonator is affected primarily by the length of the tube through which the sound wave travels. A spherical resonator will be affected by the amount of opening it has and by whether or not that opening has a lip
And as you will go on to read, the tissues in the body are not like the hard surfaces of an inlet manifold or exhaust pipe so affect the resonance. I'm sure you'll be aware of the effect of dampening in a resonant system in smearing out the resonant frequencies. As I say, lots of stuff going on. I don't know much about it - but I think other people do know a lot.
Re the straw thing, I've absolutely no idea. I've also no idea why one reed chokes and the other vibrates a minor 3rd above it's normal frequency. Weird isn't it?
Last Edited by MindTheGap on Jan 26, 2016 1:05 AM
Here is a LINK in PDF form to an interesting study on the 'Acoustical and physical dynamics of the diatonic harmonica'. I'm sure this must have appeared here before but I've just started reading it and it's very interesting. Edit:- JUST FOUND A LINK TO THIS STUDY IN 'Pat Missin's Scientific study ARTICLE MENTIONED IN AN EARLIER POST BY MICHELLE.
So many discoveries to be made on this simple 'Toy' of an instrument. It seems the more I learn the more I have to learn.
Which puts me in mind of one of my favourite quotes on discovery made by the great Sir Isaac Newton relating to science on a grander scale.
'I seem to have been only like a boy playing on the seashore and diverting myself in now and then finding a prettier shell or smoother pebble than ordinary while the great ocean of truth lay all undiscovered before me.'
Last Edited by Grey Owl on Jan 26, 2016 8:22 AM
"But there is a big assumption underlying the discussion that no one seems to question: that the only way to do this is to bend the reed lower in the slot at the base (lowering the zero point it's sometimes called), and then re-shape."
I don't assume this is the only way to do things. But most of the time, this is the most practical way I have to get a predictable and long-lasting result. I'm hesitant to rely on applying something to the reed plate that may eventually flake off.
"However, there are some negatives to this approach. It lowers the pitch, it's time-consuming, and bending the reed so abruptly at the root must necessarily weaken it, as it would with any spring (which is what a reed is)."
You are correct that there is a risk of work-hardening the reed which will lower the pitch and shorten reed life.
That being said, If the work I do causes the reeds to be weakened, I'd be seeing a lot more harps come back for warranty service than I do, so I believe the risk is low if the reed work is done carefully.
I think there is a correlation between the amount of lowering of the pitch and damage to the reed as it is worked. I'm constantly plinking or playing the reed as I work on it. Activating the reed helps it settle into it's resting shape. It also lets you hear the sound and feel the response of the reed which are important to making changes to the reed. There are some things that I just don't do because they instantly and significantly drop the pitch - like raising the base of a reed by shoving a shim under it for example.
If you shape a reed by applying pressure and counter-pressure is areas so as to not focus the stress in the same spot, you can do a significant change to the shape of the reed and only drop the pitch by a cent or two, sometimes not even that.
"Seems to me we know that torsional movement causes squealing. So anything that would limit torsional movement would also limit squeal. I think that's why nail polish works. It dampens torsional vibrations."
So does a blob a blu-tak in the middle of the reed. But the context of what I was refering to was in adjusting the shape of the reed, nothing more. I eliminate squeal by adjusting the curve of the reed. In general, if air has nowhere else to go but to make the reed move up and down, it won't squeal.
Nail polish can chip off with time. Wax is great, but I only apply it after the fact, once the reed shape is where I want it. It doesn't contribute very much but it's so quick and easy, it makes sense to use it. I use my hair dryer to warm the plate so that the wax sticks and apply a very tiny amount to the corner of the rivet base. I use so little that even if it melted, it would be cleared by the reed the first time you played it. I just thought I'd mention that before someone asked - again, my point was about adjusting reed shape.
"Now, these different vibration modes give rise to harmonic overtones in the resulting chopped air stream. Even though the reed wants to vibrate at its natural frequencies, it's still vibrating at non-eigen frequencies. It's those non-eigen frequencies that give rise to the sound signature, if you will, of that reed. As to an experiment, you do it all the time when you plink reeds. No need to excite the reed at its natural frequency. A good plink is the physical equivalent of a mathematical impulse or Dirac delta function which excites all frequencies. You plink a reed an it will vibrate at its natural frequency."
Does that explain the following: - As you emboss the slot, the sound becomes brighter, there is more emphasis on the higher overtones. - As you improve the reed shape, there is more sustain as you plink. - As you change the reed shape, the timbre changes. I would describe the sound of the best reed shape as woody or fat. Likewise an ill-shaped reed will sound dull.
Michelle, Im still learning set up etc I was really into working on harps a few years back--Ive mellowed since then--if it works don't fix it mentality-----==If I see a lot of light at rivot end [more than the other reeds] and it is airy or won't bend and or ob ,or close so I can overblow, then I push it down[rivot end] -but definitely not way down into reed slot. so that reed has to come back up to be level with reed plate again--as I said if u look at a manji or even crossover the reed is very close to reed plate ,with very little gap at rivot end-----I use an “Ice cream cup wood spoon”-sand and shape the wood fairly straight at one end—that same end is wider than reed, so that it picks up the reed plate as a [ stop or guide]---I keep the stick as close to parallel with the reedpalte as ppssible-then work it slowly towards rivot end-sometimes start at rivot end and go to tip,gently--- the wood helps with no scratches [affecting tuning] and works slower than metal---try not to overwork reeds—if u want to break a piece of thin metal in half, you bend it and work it till its weak, then break it or tear it in half-I personally think if u OVERWORK a reed its easier to accidently bend it when re tuning et—also order’’ on line [.001] [.002] [.003] metal shims------http://www.victornet.com/subdepartments/Feeler-Stock/1030.html------- ther about ½” wide or so a foot long---order 3-4 of each—cut them with sharp scissors to width and length u want and use to gently push [pry] a rivot end back up or put .001 about a 1/4” below rivot end then do popsickle stick push down –thusly help preventing reed from going to deep---also use to debur and some times help straighten a bent reed----and to test height of reeds at rivot end---the bent reed will last awhile and usually is ok but may fail down the line--- hope this helps
Timber is related directly ti the way the pressure varies over time. This can be plotted using a microphone and an oscilloscope. A microphone converts pressure to voltage. With little software you can use a PC instead of an oscilloscope. You can even plot a sound file in excel without too much trouble. Here is a chart of the pressure vs time graph for a few different sounds. I mentioned before the idea that the size of the opening between the slot and the reed of a harmonica may vary in a way related to the pressure vs time curve. My first pass approximation of a reed, a straight rigid infinitesimally thin reed passing though an infinitesimally thin reed plate, with the reed swinging sinusoidaly, yielded a sinusoidal are vs time curve. If some of the complexities of harmonica timber are to be understood by the area change over time, the will show up as the asymmetrical aspects of the interaction between the reed and the reed plate. AS the reed enters the reed plate, it closes off the flow and it stays closed for a while until it opens as the reed passes out of the bottom of the reed plate. This basic asymmetry may explain some of the timbre but I suspect that a lot of the complexity of the sound depends on exactly how the reed shuts off and re-opens the flow.
Michelle, you had mentioned attempting to model this in more detail. Just adding thickness to the model I described complicates the math quite a bit. If you add the fact that the stamped edges of both the reed and the reed plat are not flat and square, it gets complex fast. That is actually a good thing if one is hoping to see a complex area vs time graph that matches the complex timbre of the harp.
Last Edited by STME58 on Jan 26, 2016 9:42 PM
Yes, that's it exactly. Phew, thank goodness for that. You don't have to explain or model the complexities to understand that this mechanism can create the timbre.
It's a theory, and can be tested. But a very sensible one that is both physical and fits the observations. And doesn't require invoking some other mysterious effects.
And obvious consequence is that changing the thickness of the reed plate, the shape of the reed, the base height of it - anything that changes how the valve opens/closes ought to affect the timbre and response.
Great thread, interresting read! Another approach is to let the reed find it's optimal arch by the acctual airstream/pressure it self ... and not by tools which with untrained hands easily can overdo and overstress the reed, applying to much force at single points along the reed In short ... start with a curve/arch close to flat ... then (in the same way Arzajac described when he playes the reed directly on the single reedplate) quite forcefully playing the reeds OB/OD until you run out of air ... do the whole plate and repeat a couple of times. This will guide the reed to find the natural/optimal arch ... which acctually, like we all agree upon (?) is closer to flat ...
Read more about it: (It's not like it's my finding) http://www.harmonicaspace.com/harmonica-forum/maintenance-repair-customization/reed-arcing
