Tuning question ( again! )

[wirralaccordion]

If the three M reeds of a LMMM piano accordion are ( Scottish? ) tuned at +20 cents, 0, -20 cents and A4 = 440Hz then the frequencies for A4 are 435.06 Hz, 440 Hz, 445.232 Hz respectively.
For A3 the frequencies are 217.53 Hz, 220 Hz, 222.616 Hz respectively.
For A5 the frequencies are 870.122 Hz, 880 Hz, 890.466 Hz respectively.
A graph of cents v frequency would be a horizontal line. Why are all the tuning charts I have seen that show cents ( y axis ) versus frequency ( x axis ) parabolic? Is the y axis showing something other than cents?

 

[maugein96]

Phil,

Mostly way beyond my knowledge, but the following sort of stood out.

Firstly, Scottish tuning will typically be above 20 cents, maybe 25 cents or even more.

Secondly, from what I've read elsewhere it doesn't seem to be common practice to balance the sharp tuned reeds with a corresponding flat reed, i.e. you normally wouldn't get 20 flat ,0, and 20 sharp. Perhaps more likely to be 18 flat, 0, 20 sharp.

Now, I don't know why that is so, and it seems the vagaries of musette tuning will always be bordering on sorcery to those of us who struggle to understand it all.

No doubt there will be standard factory musette tunings, depending on the destination of the export and maybe even the requirements of a particular player. In my very limited experience of such things I'll reiterate that I have tried three brand new accordions off the shelf, all with factory three voice musette tuning, and no two of them sounded the same. That would be in the late 80s, and I don't know what the state of accordion tuning technology was at that time. I dare say with modern electronic tuning aids they could have perhaps got them very close, but I'd still be willing to wager that even so, no two would sound exactly alike. After the final tuning the reeds would be inclined to settle down.

Finally, with each rising (and falling) octave I've been told that tuners must actually reduce or increase the pitch of the off tuned reeds, and any graph would reflect that slight pitch change. The very few three voice musette accordions I've owned have all tended to sound very slightly out of tune the higher up the treble keyboard they were played. I presume that is because the tuner hadn't adjusted the reeds precisely, or more likely my ears just played tricks on me. If you've ever played a stringed instrument where one string is even slightly out of tune with the others you'll know what I mean, and any such occurrence tends to require prompt rectification. Yet the musette principle requires the reeds to be deliberately off tuned!

The upshot of all that was I started to go off three voice musette tuning with a vengeance, as I just couldn't find a musette tuning that sounded balanced enough. I went onto an LMM accordion with both MM reeds tuned in unison and that sorted my hearing issue out. However, with the passage of time some of the MM reeds went out of synch with each other and I started to re-acquire "musette ear". The notes where the reeds were slightly out of tune became a source of irritation.

I soon discovered that maintaining a unison tuned accordion means that any wonky reeds must be attended to immediately, so the compromise (for me) was to opt for swing tuning so that the whole show was a bit more stable. In time I got used to that tuning and stuck with it (more or less).

That aside, there are many thousands of accordionists who wouldn't dream of playing an accordion without three voice musette, or strong vibrato on two MM reeds, but their ears and mine are at odds with each other. Ideally I would go for an LMH accordion and hope everything stayed in tune!

I hope I haven't served to confuse you further. I only have O level English so had to look up the meaning of "parabolic", as I thought it may refer to airborne troops who weren't very good at parachuting! Still not sure what it means in the context that you have used it, which I took to be a "mirror image", but the logic of it all is beyond my knowledge as I mentioned above.

Don't think I've ever seen an accordion tuning graph, but it would maybe be easier if every reed had its own dedicated tuning peg!

 

[oldchemist]

I am no expert on this but I understand that the tuning at 440 (say +20, -22) applies only to that set of reeds. Higher pitched notes need a wider tuning to give a faster vibrato and lower pitched notes need a closer tuning to give slower vibrato. The change in vibrato speed is close to continuous along the keyboard so the tuning graph is not a straight line. It will be some sort of exponential shape.

 

[wirralaccordion]

I am no expert on this but I understand that the tuning at 440 (say +20, -22) applies only to that set of reeds. Higher pitched notes need a wider tuning to give a faster vibrato and lower pitched notes need a closer tuning to give slower vibrato. The change in vibrato speed is close to continuous along the keyboard so the tuning graph is not a straight line. It will be some sort of exponential shape.

Yes, that is correct. It is an exponential curve. What I am looking for is the equation that relates the cents deviation to the frequency, eg plus 20 cents at 440Hz is less than 20 cents at 880 Hz and less again at 1760Hz. I have seen this equation somewhere but cant now find it.
Once I have it I can calculate the cents deviation for every note and give these to my tuner.

There is a useful explanation at
http://accordionman.web-ring.info/TUNING/

 

[StargazerTony]

Once I have it I can calculate the cents deviation for every note and give these to my tuner.

Just curious, but isn't that something that your tuner would/should already know?

 

[wirralaccordion]

Once I have it I can calculate the cents deviation for every note and give these to my tuner.

Just curious, but isn't that something that your tuner would/should already know?

Well, not exactly. According to the link I have provided there are at least three possibilities, and I think "reference tuning" ( option 2 ) is what I mean.
Unfortunately I don't think Andy Bakke is on this forum to ask.

 

[boxplayer4000]

I find it difficult to believe that good tuning can be done simply using mathematics or computers. A computer or tuner may be useful to take you into the area where you want to be but I believe the final adjustment needs to be made by ear. If all reeds were perfectly made and displayed similar harmonic content and response then it might be possible to rely on mechanical means but of course that is not the case.
I also find it unhelpful to try and glibly describe musette tuning as simply + and - 20 cents (for example) when in fact tunings of, say, -20 0 +20, -18 0 +20 and -16 0 +20 , though close will all sound quite different. To my ear there are very precise points where the low reed needs to be to sound in tune. (Adjustment can of course be made with the + reed as well.)
As my background was in mechanics and engineering I did study the cents/Hz frequency information and found the charts which Peterson Tuners used to provide helpful.

 

[maugein96]

Whilst the technical and mathematical elements of this are way beyond my knowledge or experience, it seems always to have been the case that tuners carry out the final adjustments by ear, as mentioned by Boxplayer. You are virtually reliant on the tuner's perception of what sounds right and what doesn't, regardless of how many "cents" or Hertz they are working with. 

I have read the odd article in various accordion magazines and media that infers that each and every accordion will be so unique that no other can be made to sound like a clone of it, regardless of any chart or formulae used by the tuner. 

It seems accordion reeds are prone to go out of tune due to various environmental factors, so the best you are ever likely to get is an approximation of what you perceive to be your requirements. On the very few occasions when I've heard discussion about musette tuned accordions, certain people prefer one type of sound over another, and tuners are probably no different in that respect. One tuner will do the same box differently from another. 

I've only been dabbling in accordions since the mid 80s, but given the number of reeds required to provide three voice musette, has the tuner realistically got any chance at all of getting all the reeds bang on? That's probably the attraction of three voice musette. No two boxes will ever sound identical, almost like the human voice. 

Verchuren's tuner was inundated with requests to tune Crosios the same as Verchuren's box. Buy his own admission, even with over 50 years of tuning experience he was never able to achieve that. 

Better than that, get two players to play the same box, one after another. I'll virtually guarantee that they'll not get the same sounds, even if they play exactly the same tunes from musical notation. 

Personally speaking, my ears sometimes start to object when I play past the middle of the treble keyboard, especially on a musette tuned box, and given the choice I'll often move to an other key so that I'm not playing so many high notes. Yet most accordionists revel in those very notes that I can find irritating. 

I think if I was a tuner all I could cope with would be bass accordions. I'd better not complain too much as there aren't many of them left!

 

[maugein96]

After I'd put the above post on something else sprang to mind, and whilst it doesn't 100% cover the subject of "tuning" as such, it might be worth considering.

You'll need to forgive me for my constant reference to French accordions, but I have virtually zero knowledge outside of that field.

On the subject of being at the mercy of the tuner, the very well respected accordion repairer and tuner, Georges Pellegrini, maintains a website. He supplies new accordions, repairs boxes, and engages in buying and selling various trade ins he receives. He writes a summary of any work he has carried out on an accordion prior to offering it for sale.

He got his hands on a used top of the range Cavagnolo Vedette 10. These boxes cost upwards of 12,000 Euros to buy new. Whilst carrying out the refurb he decided that the double bassoon reeds were not performing to their best potential. His solution was to replace the Cavagnolo reed blocks with heavier wood so that they achieved a better sound quality. Cavagnolo have been making that model of accordion (or variations of it) for maybe 60 years, but with Georges' experience of repairing many different makes, he decided the accordion construction wasn't adequate. It may come as no surprise that he doesn't deal with Cavagnolo unless by special request.

Now that is the viewpoint of one man. Few other accordion professionals would post on a website and suggest that Cavagnolo's top of the range boxes are of inferior build quality.

My point is, whose ear can we use as a benchmark when it comes to tuning and sound quality?

No amount of graphs and charts are likely to be followed to the letter if the tuner ultimately decides they don't produce the correct sound (in his opinion).