Where/How To Learn Reed Tuning ??

[Abraxas]

I attended an excellent accordion repair course in Teddington presented by Paul Flannery. It includes reed tuning. Course is organised by Peter Le Geyt at <EMAIL email=plg@plgmarketing.com>plg@plgmarketing.com</EMAIL>

I have to look at my location settings as methinks my London in Canada has been confused for the original. The British got around :lol: :lol:

 

[boxplayer4000]

Abraxas,
There's no quick way to learn accordion tuning. Start on simple, expendable instruments.
You may get some pointers on 'talkingreeds.com'.

 

[Morne]

I asked earlier, but didnt hear back as to whether anyone knew where my link to manufacturer associated wet tuning charts was ? I lost it. It listed how many cents were found for various manufacturers. I did stumble on some more info here though: http://forum.melodeon.net/index.php?topic=2306.0 . I cant believe nobody has put together a more organized instructional for this issue. Im getting crumbs from various sources and methinks a consolidation is in order.

While I dont have any specific links, my opinion on those various numbers out there is that they dont actually say all that much. Except that X is wetter than Y. When you are given only a single value, say 16 cents, then that really only tells you what its like at that one specific note. In order to draw a curve you need one more point (or at least some other well defined parameters like gradients, curve function, etc), unless you go with the same number of cents all across the full range.

Here are some graphs I built based on a bunch of different tremolo tunings (MM only). They are intentionally messy to show that things are not as discrete as wed like.

Hertz:
<ATTACHMENT filename=Hertz.png index=0>

Cents:
<ATTACHMENT filename=Cents.png index=1>
(This graph messes up at the low end, but that is because the formula seems to break down when you extend it that far past the normal range.)

The ones labelled Tabelle and Scwebetontabelle are based on a single exponential curve with input values determined primarily from Hohners. This research was done and published by Toni Schwall. I rebuilt the formulas and used only his 2 data points (A4 and A5) to regenerate the curves.
He also mentions other curves where you could have a simple linear Hertz increase, a stepped increase (e.g. using the same Hertz for one octave and then increasing for the next), and more complicated exponential curves where you have two or three together in order to allow for a different rate of increase for each. Toni also writes a bit about MMM tuning, but I skipped that part as Im not too interested in that kind of tuning.

The next 4 are my attempted approximations using Schwalls exponential formula, but based on the rough graphs on the Accordion Revival website (based on Thierry Benetouxs book). I dont have Thierrys book, and I dont plan on getting it, so my curves should be taken with quite a few grains of salt.

Mengozzi is a relatively wet tuned accordion I have. That line was an attempt to find a simple exponential matching the current tuning, but due to some outliers it might be slightly wrong.

The Porvenkov lines are based on research done by Russians on optimal tremolo tuning. I wrote more about it on a different thread.

Which is all really to say that if youre going to be mathematical about this, then those numbers out there arent all that useful. And its also not really that simple to say Hohner tuning is X cents and Scandalli tuning is Y cents.
On the other hand, once you look at accuracy realistically, a lot of those lines will overlap neatly at various points (because you probably wont be tuning to sub-cent accuracy), but Id still maintain its not enough to have just a single number unless the tuner has those other parameters in mind and uses your single number as input. But then again, Ive done this for only a few months now and Im sure experts will call this overanalysing.

 

[Abraxas]

geoff, I'm in London, Canada not UK. The British got around.

 

[Abraxas]

OK here's my situation: I've decided to tackle a 48 bass Frontallini first. There are some sour notes on both the bass side and treble side. They don't just sound out of tune but some notes are very close but beating at quite a different frequency than others. I reason that it may be the case that these reeds are out of tune due to dirt or corrosion. So, at least on the treble side thus far I have taken the reed plates off and am soaking them in a CLR (Calcium, Lime , Rust remover)solution. I''l take the leather valves off if they don't come off in the solution. I'll give a good rinse in mineral spirits. I'll clean and inspect under magnification, to make sure they are in line and spotlessly clean. I am ordering some valve replacements from Charlie Marshal, the synthetic ones. Then what shall i do ? Wax them back onto the block and build an air box for tuning ? I am a licensed machinist with a diploma in electronics so measuring frequency and working with tools shouldn't be a problem. How does one tune these reeds though when they are waxed back in to position ? at least one reed would be inaccessible from the outside at that point. I don't get it and could really use a video showing this being done. Cheers.

 

[jozz]

 

[Abraxas]

OK, I deduce from the video that after I thoroughly clean and inspect all reeds, I ,
2) replace valves - in this case with synthetics
3) rewax the reeds back onto their blocks
4) here's where I get a bit lost also, some suggest the reed blocks need to be strapped back into the accordion and the reassembled box is sounded and disassembled for tuning... I'll call this method "in vivo". That means a small box should take about a life time ....others are taking the reed blocks out and dragging them across an air table and tuning the reeds and of course I'll call this "in vitro"

LASTLY ... :hb :hb :hb :hb I'm still completely confused about the reed relationships for musette. I take it one tunes a reed spot on but the second how many cents out for a nice sweet musette ? 5 cents ? 10 ? Is there some sample sound clips online ? Do I follow the same number from bass to treble ? Is this second reed tuned sharp above the nominal or flat below it ? I've heard the reeds in there now beating and I sure as hell would hate to have a tremolo that beats at a different rate from one reed to another. How did we get from a nice musette to something that sounds like a moog synth through a tremolo unit ? An excess of separation in frequencies ?

Thank you so much for your time and patience with me.

 

[jozz]

Most of these questions are answered here.

 

[Abraxas]

By the way ...the sound Im after is at about 1:58 here

 

[debra]

OK, I deduce from the video that after I thoroughly clean and inspect all reeds, I ,
2) replace valves - in this case with synthetics
3) rewax the reeds back onto their blocks
4) heres where I get a bit lost also, some suggest the reed blocks need to be strapped back into the accordion and the reassembled box is sounded and disassembled for tuning... Ill call this method in vivo. That means a small box should take about a life time ....others are taking the reed blocks out and dragging them across an air table and tuning the reeds and of course Ill call this in vitro
...

What I do:
1) take off all reed plates, remove valves, clean and soak in naphta overnight, then clean again;
2) glue on new valves on the side of the plate that is on the inside of the reed block;
3) rewax the reed plates onto the blocks;
4) carefully clean any possible wax spills;
5) glue on new valves on the outside; (when you do this before waxing the waxing job becomes way more delicate)
6) check whether all notes play as they should, whether valves do not catch on the side wall inside the reed block... and this check is where that tuning table with a hole for a single reed comes in handy;
7) important: wait 2 to 3 weeks to let the wax settle (dry)
8) start tuning: you always measure with the reed blocks inside the instrument, but if the accordion has no cassotto you can place half the accordion on the tuning table (this time with a large rectangular hole, a bit smaller than the body of the accordion)
9) you can work on the reeds with the reed blocks still in the instrument or with reed blocks removed but you should never check the frequency on a reed block on the tuning table, only with the reed blocks inside the instrument.

 

[Abraxas]

Thanks Paul. I guess what you are saying is that the frequency in vivo is different than in vitro. What I don’t understand then is why we have an air table ? If I sound the reed on that table, why am I doing that if I am going to do it again in the instrument and use the frequency measurement from the instrument ? I’m missing something.

 

[jozz]

7) important: wait 2 to 3 weeks to let the wax settle (dry)

I dont doubt your methods, but why is this?

Ive been to my guy for a wax repair jobs (certain spots), in and out and playing the same evening twice. And it wasnt a problem (i think).

 

[debra]

Thanks Paul. I guess what you are saying is that the frequency in vivo is different than in vitro. What I don’t understand then is why we have an air table ? If I sound the reed on that table, why am I doing that if I am going to do it again in the instrument and use the frequency measurement from the instrument ? I’m missing something.

The air table as you call it (actually a better name than tuning table as it is not used for tuning) is mainly used to check whether reeds are functioning (sounding) properly. You can check whether the valves open will enough, whether the valve on the inside isnt catching on the side or top wall of the resonance chamber, whether there isnt any unwanted vibration like a reed or booster valve vibrating at low volume, etc. You can do a rough check of the frequency, for instance to see that the right note is in the right position on the reed block, but thats about it. To do real measurement of the frequency for small deviations the reed block needs to be in the accordion as the combination of reed, reed block and accordion body together determines at what frequency the reed will sound.

 

[debra]

7) important: wait 2 to 3 weeks to let the wax settle (dry)

I dont doubt your methods, but why is this?

Ive been to my guy for a wax repair jobs (certain spots), in and out and playing the same evening twice. And it wasnt a problem (i think).

As the wax settles there may be small deviations still. You can start playing the same evening indeed, but the accordion may be ever so slightly out of tune again after a few weeks. Note that the same is true (typically just the next day or so) after tuning a reed (a reed tuned down tends to go up in frequency again all by itself) or after replacing a leather valve.