[David I Am]

So having a *ahem* few melodicas, and not infrequently needing to tune them, which is tedious to say the least, and also being ridiculously handy and technological, I was considering that I could make a small bellows out of 3D printed parts, with a microprocessor, a pwm fan and a pressure sensor. So I’m planning on making one for myself – eventually – but if there is significant interest, maybe sooner.

Imagine, if you will, a device approximately the size of a paperback book. It has a little lcd screen, a couple buttons, a couple leds, and a flat top with a rectangular opening.

When powered up, there will be two things on the display – the requested pressure drop (which can be adjusted by the buttons up or down) and the tuning display.

To use it, you place the reed in question over the opening – when the pressure hits the requested level (probably in a fraction of a second) a green led turns on, and by this point I expect the reed will be sounding. You can adjust the reed as you need to by the usual filing techniques, seeing the note position feedback on the screen immediately and interactively.

This should make tuning go as quickly as at the factory. Which would be quite a relief. Would be usable for harmonicas too.

I suppose people would ask ‘how much would such a thing cost’… well, given the time I’ll put into proving the design (I’ll estimate 50 hours or so dev time to develop), *I* rate that at about $5000 value up front, plus another $35 in miscellaneous parts and materials per unit created, and another $65 in print and assembly. The market for these things can’t be very large – If I was to imagine a kickstarter, the ask could be $7500, you can get your melodica tuned 24 earlies for $35 and the rest for $50, the early-bird-early-dozen product reward would get a tuner bellows for $100, the next two dozen pay $150, the rest of the starters at $200 with a retail price of $250 if I ever sell another one. It’d be fully funded on 24 $35 tunings, 12 $100 bellows and 24 $150 bellows, and 12 $200 bellows.

First I’d have to make a prototype though, to show how awesome it is. 😀

Your opinion?

[David I Am]

Wow, I overestimated the interest in this.

[Alan Brinton]

There is interest in tuning, David, and some of us do a lot of it. I have tuned and retuned 30+ melodicas, current as well as vintage, and have a pretty good feel for it. What you envision is interesting, and it would be great if there were some simpler, more mechanical and predictable way of tuning. But in my experience, a critical aspect of tuning is getting the feel for sounding the reeds so as to get stable readings that are consonant with where you’re trying to get your instrument. And becoming sensitive to reeds not sounding quite right after initial efforts to tune them. I think much more, though I can’t explain it well, is involved in this than getting constant pressure at the right level. I am able to tune a melodica to my satisfaction, making adjustments to gapping as I go along, so that the notes all sound right, are comparably responsive, and sound right in combination in combination with other notes and occasionally in chords. I’m skeptical about a more mechanical and measurable approach. I’m fairly confident about being able to tune to my satisfaction, but also about being able to significantly improve the tuning of someone else’s melodica. I know there are more sophisticated musicians who have the ear to do this better than I do, and I envy their ability. They can take the same approach and tune better than I can. But, as I say, I’m pessimistic about a more mechanized approach, even though I find the idea attractive. I have a lot of melodicas, several of which I obsess about keeping finely tuned. Some models hold their tuning well, for example the Yamaha Pianicas. Some need more frequent attention, such as the vintage Suzuki MA-32 Super II and the Suzuki II 25, which have less stable reeds but produce compelling sound.

Another simplified approach to tuning is to tune by pinging the reeds and using a pickup so the melodica doesn’t have to be repeatedly opened and resealed. I have experimented with this a bit but don’t believe this does justice to the role of the ear in tuning. Maybe for some it could do so.

But I also don’t think a lot of people actually do much tuning, as a result of which there’s less interest in your proposal than it merits. I think every serious melodican should get into tuning. I enjoy tuning. It’s very satisfying. I strongly encourage you to experiment with your envisioned approach. It’s possible that it could be a significant advance in melodica tuning technology.

[David I Am]

Thank you Mr. Brinton – I will create this tuner eventually, and don’t be too surprised if I try to solicit your help to alpha test what I come up with. I value your experience and feedback.

In considering how our aerophones are distressed diff between play vs. tuned, I can think of several factors – temperature is one, as our warm breath is going over them. Also, moisture may hold some factor as well.

And that isn’t to dismiss the resonance chamber of the instrument and how that could slightly bend the effective tuning – if a reed is hitting counter to a sound compression wave that bounced off the end of the wind chamber on every couple occilations, that could slow it enough to tweak the tone a cent or two.

Perhaps a more measured differential approach option, using software, would be reasonable as a more advanced way that would address some of your concerns. Play each reed *in* the melodica while warm and moist as a ‘calibration’ pass combined with a pressure sensor inline on the blow tube, which can then be stored in memory. When you remove the plate from the melodica and place it on the bellows, it knows 1. the pressure you played it at and 2. the differential between how it was in the instrument and how it sounds now

as a for instance, initially measured 222.8 at 400 pascals vs. now outputting 223.6 at 400 pascals – if 220.0 as the desired frequency it could bend it the effective target to pitch 1.2 hz lower – naturally there’s going to be a curve of some kind involved in this effect but for a particular reed is probably linear approximation will do… At any rate, it may be possible to have a procedure which makes every effort to compensate for the alteration in circumstance.

It may also be that few people are doing tuning because it is so tedious and twitchy a process to engage in. Were there a clear sophisticated tool with videos and documentation that eliminated much of the tedium they could get more engaged. Or open a business tuning for others. That’s my hypothesis anyway. Who wants to play a lousily tuned melodica? 😀

David

[Alan Brinton]

I am very interested in hearing more about your process as you work on it, David. I tune within 3 cents plus or minus and probably average opening and closing the reed chamber about six times. I start out with readings for the melodica fully assembled, then get readings for it removed from the tray and and work from those, not fully reassembling each time. The whole process usually takes 2-3 hours. Nowadays I’m usually tuning the whole keyboard down to A=440 or 441. That issue aside, though, a melodica’s tuning can be significantly improved by determining the closest standard to where it is (often 442-443) and then just tuning the notes that are seriously out of whack, off by some looser tolerance, say +/- 6 cents or even more.

We don’t really know what kind of results your approach might produce, though, maybe better than I’m suspecting. Conceivably as good as what I get by the usual method. With your method there’s also the option of doing some fine tuning at the end. There’s also the question of how closely in tune a melodica has to be.

In my opinion, tuning a melodica is much easier than most people think it would be, especially after you’ve done a few. And tuning can be significantly improved in many cases by tuning just the few notes that are furthest out of tune. A tuning business? I’d probably charge $35-40 plus postage for a standard melodica, maybe more to tune down to A=440. But this doesn’t interest me.

[Author]

Posts