Another Idea For A Just-tuned Anglo

[Chris Allert]

i've heard second-hand from someone that paul groff believes anglos would have historically been tuned differently for the push notes than the pull notes. i also know this would be consistent with how many harmonicas are tuned, with a pure major triad for blowing and a 7-limit tuning for the draw.

 

since i've also gotten a few e-mails from different people about ideas for just tuning anglos, i thought i'd post my current thoughts on the matter in the hopes that these thoughts can be corrected by those who know more about this than i.

 

so here's an idea for tuning an c/g anglo:

 

on the push, use a pure 5-limit just tuning that will fit all the notes on the c and g row:

 

http://www.morelater.com/anglo/push.html

 

and on the pull use this tuning:

 

http://www.morelater.com/anglo/pull.html

 

i've rendered some sound files of what the c-row would sould like here

 

http://www.morelater.com/anglo/

 

(the different mp3 files are what it sounds like with different waveforms so you can get an idea of how the upper harmonics will interact with eachother)

 

so to summarize, the c and g row are tuned a pure fifth apart, you get a pure major triad on the push for each row, and you get a very pure sounding diminished minor chord on the pull of either row. in addition D is the same in either direction (c-row pull or g-row push), so you can use that as a reference pitch. since G, D, and A are pure fifths apart in both the push and pull temperament, you can also use A as the reference pitch for both pushing and pulling.

 

the accidental row is less straightforward. my instinct would be to just extend the system for each direction to all 12 notes. see the above push.html and pull.html files for my thoughts on what the notes should "logically" be in either system. the file http://www.morelater.com/anglo/push7.html is another option for tuning the push accidentals.

 

to try different settings, you can use the following utilities i wrote (and which i used to generate the above tuning tables):

 

5-limit just scale calculator:

http://www.morelater.com/scales/just.html

 

7-limit calculator:

http://www.morelater.com/scales/just75.html

 

to browse my hastily written and undocumented source code, go here:

 

http://www.morelater.com/scales

 

and click on any of the .txt files to see the source for the corresponding cgi program

[Dana Johnson]

After hearing the Mp3's, I think that while chordal accompaniment for singing might work well with this scheme. ( the out of tuneness of the drow notes lending color to the chords and not being too far out) Where I would have difficulty is in the contrast between the press and draw notes. I often will play them in close proximity depending on phrasing of the, but here they would stick out like a sore thumb. My experience of peoples tone memory is that for short periods it is very good, and would not want to hear the same note many cents apart. Even a few cents is often more than they would like to tolerate. If the whole thing is tempered the same throughout, it doesn't matter so much if the intervals are good and "bad" if at least they are the same. This kind of tuning would effectively eliminate the usefullness of the duplicate notes, though it would create a different kind of usefullness I suppose. Seems like the usefullness of th e set up would be limited to certain kinds of music enough that you might as well stick with the full just tuning or some other temperment specifically designed for that music and not try to make a temperament do too many things.

Dana

[Chris Allert]

i might agree that this system wouldn't work very well for the way many people play irish music today, myself included. but it seems like it might be better for someone who plays up and down the rows and wants to use a lot of adjacent button thirds and fifths for accompinament. i've heard that william kimber played this way, and i wonder if the style he developed had something to do with the way his concertina was tuned.

 

i was actually thinking that it might be nice to have some of the duplicate notes sounding very different on the push than the pull since it might give one more options for giving a tune a different flavor. for most irish music played in d and g, it would only be the c natural that would be different depending direction. but the difference would be very noticable. the c natural is a very important note in a lot of tunes and the way a lot of tunes use c sharp in some phrases and c natural in others is an important part of what gives the music its character.

 

but maybe in practice having two c naturals wouldn't be as useful since you would also have to phrase things differently depending on whether you wanted the push c or the pull c.

 

i guess there's only one way to find out...

[Dana Johnson]

the c natural is a very important note in a lot of tunes and the way a lot of tunes use c sharp in some phrases and c natural in others is an important part of what gives the music its character.

 

but maybe in practice having two c naturals wouldn't be as useful since you would also have to phrase things differently depending on whether you wanted the push c or the pull c.

 

i guess there's only one way to find out...

This is an interesting point. I've heard a number of good fiddlers be less distinct about the C / C# note and a few others as well. It isn't as though they can't play the "right" note, rather that the music allows and even incourages some ambiguity here. Pulling the note sharp lends some tension which is a little disconcerting at first if you are used to the standard notes, but after a little listening, it adds a lot to the tunes. The only way I've found to approximate this on a concertina is to sort of slide between the two notes, be it the C/ C#, F/ F#or G/G#. There are probably others I'd do it with as well, and it only half accomplishes the goal. I think in the wrong hands the tuning would be a disaster, but in the right hands it could open up a whole new world. Pity reeds don't have little screw tuners on them

Dana

[Paul Groff]

Dana wrote: "This is an interesting point. I've heard a number of good fiddlers be less distinct about the C / C# note and a few others as well. It isn't as though they can't play the "right" note, rather that the music allows and even incourages some ambiguity here. "

 

 

Dana, I think your language may lead some readers to a conclusion you probably don't share: that is, that the equal-tempered values of "C" or "C#" are somehow right, and that when a great fiddler, whistle-player, fluter or piper uses a different value (or a whole series of different values) for these notes he or she is somehow in an "ambiguous" zone between these "normal" or "correct" values. That point of view (which as I noted, you probably don't share but which your language suggests) is exactly backward.

 

In many kinds of traditional music, but for example in Irish music as played by the greatest players I have heard, musicians are very deliberately and skillfully going for the RIGHT pitches that they want to hear, but very often those correct notes cannot be found on an equal tempered instrument, or by playing a fiddle (etc.) in perfect equal temperament.

 

Dana wrote: "I think in the wrong hands the tuning would be a disaster, but in the right hands it could open up a whole new world."

 

Rather, this is getting closer to the older, original, and traditional sound of the early german concertinas and anglo concertinas.

 

Chris, I'm not going to publish much more of my unpublished work on tunings here, but your own experiments will doubtless lead you to tunings you will find very useful in your goal to make harmonies on the anglo sound more like those of the Irish pipes.

 

I will tell you that I have coined a new term (never yet published), "decoupled tunings" to refer to the class of tunings for bisonoric instruments in which the "same named note" (e.g. D#) has a different pitch value on the press vs. the draw of the bellows. Thus there is a large family of "just-decoupled tunings" and another of "meantone-decoupled" tunings in each of which the press and draw values for the same note are decoupled in pitch, allowing the harmony among the family of notes in each bellows direction to be optimized.

 

Now are you starting to understand my objection in a previous thread when you listed a set of 12 pitch values and called it "the chromatic scale in just intonation." Again, there is NO set of 12 pitches/octave that will guarantee the music stays in just intonation.

 

Paul

Edited January 28, 2007 by Paul Groff

[Dan Worrall]

i might agree that this system wouldn't work very well for the way many people play irish music today, myself included. but it seems like it might be better for someone who plays up and down the rows and wants to use a lot of adjacent button thirds and fifths for accompinament. i've heard that william kimber played this way, and i wonder if the style he developed had something to do with the way his concertina was tuned.

 

i was actually thinking that it might be nice to have some of the duplicate notes sounding very different on the push than the pull since it might give one more options for giving a tune a different flavor. for most irish music played in d and g, it would only be the c natural that would be different depending direction. but the difference would be very noticable. the c natural is a very important note in a lot of tunes and the way a lot of tunes use c sharp in some phrases and c natural in others is an important part of what gives the music its character.

 

but maybe in practice having two c naturals wouldn't be as useful since you would also have to phrase things differently depending on whether you wanted the push c or the pull c.

 

i guess there's only one way to find out...

Chris,

This is all very interesting. I think Dana is right in saying that this might work better for a chorded style of playing than for Irish style. I'm intrigued mainly by its possibilities with the chorded style. Robin Harrison, as you know, recently had a concertina tuned 'just' and, last I heard, was quite happy with it. He plays in a Kimber style (like me, he is taken with all Kimber's simple third interval chords, which sound awful in equal temperament).

I am just (no pun intended) finishing building a GD 40 button, with a keyboard very slightly adjusted (barely, really) along its periphery to augment a chorded style of playing, and have been considering giving it a 'just' tuning when I get to that stage in a few weeks. I must admit that your idea of changing the tuning for push vs pull is a new thought for me, and right now I don't have the time it would take to give it a good hard think. By your statement 'there's only one way to find out' I guess you are thinking of tuning an instrument this way yourself? The proof will be in such a pudding, of course. If you are doing that, please let me and the other thread readers know how it goes. I certainly applaud the initiative!

[Chris Allert]

i think robin harrison's new dipper is tuned to the "push" temperament i described above but tuned lower. i'm glad to hear he's happy with it. i wonder if this means my name is slowly making its way up colin dipper's wait list...

 

i'm not sure if i'll be able to try out this tuning idea any time soon, but maybe someone else will. there's probably some way i can hook my midi concertina up to the computer to try it (i think it allows different midi channels for the push and pull, but getting the computer or synthesizer to use a different temperament for each channel will take a little more work).

 

dan, if you're interested, here are settings that might work well on a g/d anglo:

 

push: ("best key" of d, low note of g, a=440)

http://www.morelater.com/scales/just.cgi?r...&low_note=7

 

pull: ("best key" of a, low note of a, a=440)

http://www.morelater.com/scales/just75.cgi...&low_note=9

 

paul, i'm happy to see you are still around to comment on this and i hope your work gets published soon, though i suppose something will have been lost once the mystery surrounding your research has been removed. but please let me know when you begin taking preorders for your book and i'll be at the top of the list.

[Paul Groff]

[

though i suppose something will have been lost once the mystery surrounding your research has been removed.

 

 

{whispers} psssst, Chris.....

 

Yes, you have uncovered my secret scheme. Did you know the story of how the tomato became popular in Europe? This may be apocryphal, but they say the peasants were suspicious of this New World plant that so closely resembled the poisonous, European, members of the nightshade family....until it was decided to closely guard the tomato garden. That gave evidence that the tomato must be valuable, so it was promptly stolen and popularized.

 

Hope no one overheard that....

 

Actually I am now being paid for my work in another area of research, so that work will be published promptly as is my obligation. If anyone had ever paid me for the time I put into concertina research, I would feel the same obligation there.

 

Meanwhile you all are coming along nicely with only modest prodding from me. Grow, my little temperament researchers, grow.....

[Dan Worrall]

Chris,

 

Here is an interesting and very thorough treatment of temperament as used in harmonicas, by Pat Missin:

http://www.patmissin.com/tunings/tunings.html . The site has about a dozen articles on the subject, as well as very detailed instructions on how a professional tuner (himself) approaches the reed tuning process.

 

According to Missin, 7-limit Just Intonation is "the traditional tuning for diatonic harmonicas". He provides some very nice audio examples of both scales and chords in JI and ET...easier to hear than the ones you generated on computer, and of course those reeds are a lot closer to the concertina. He points out that the 4th note (F in key of C) is a bit flat in 7th limit JI...so he tunes this note to a 19th limit (which makes it very close to ET) to make the scale sound natural, thus giving a little roughness to the G7 chord. His audio clips show this nicely. What would be your take on that as you think about it for anglo use?

 

He notes that most harmonicas sold today are in ET, like everything else...but that traditionally things were much different (Just)....and he points out which professional harmonica players today use which temperament. Apparently, most pros who play chords a lot get theirs fine tuned to JI. With harmonicas being less expensive, players have the luxury of owning several in various intonations, which they use as the piece requires. But as he notes, his version of JI with a 19th limit F does not sound appreciably different to the ear than an ET scale...and gives much better chords (of course, the chord selection on the harmonica is a bit limited, to say the least!).

 

Dan

[Dana Johnson]

Dana, I think your language may lead some readers to a conclusion you probably don't share: that is, that the equal-tempered values of "C" or "C#" are somehow right, and that when a great fiddler, whistle-player, fluter or piper uses a different value (or a whole series of different values) for these notes he or she is somehow in an "ambiguous" zone between these "normal" or "correct" values. That point of view (which as I noted, you probably don't share but which your language suggests) is exactly backward.

 

In many kinds of traditional music, but for example in Irish music as played by the greatest players I have heard, musicians are very deliberately and skillfully going for the RIGHT pitches that they want to hear, but very often those correct notes cannot be found on an equal tempered instrument, or by playing a fiddle (etc.) in perfect equal temperament.

Paul

 

Thanks for the clarification and for assisting me in saying what I wanted to more clearly. I actually may diverge even more from this idea of "correct" pitches. The interesting phenomenon of pitches related by integral ratios has a great deal of "music" built around it, but it is hardly the only "music" out there. Even so, the ambiguity of the C/C# was pointed out to me by your old music partner, Sara B. And in the cases we were talking about, it wasn't just a deviation from the Tempered pitch but a whole hearted, dragging away from an interval that was close or actually beat free, to something that was farther afield. In all cases for the music the notes whatever the pitch they were, were correct according to the person playing them, and correct or otherwise sounded wonderful to me. Mostly my definition of correct is : Does it work? Whenever your temperament work gets published, put me in line for a copy!

Dana

[Paul Groff]

(of course, the chord selection on the harmonica is a bit limited, to say the least!).

 

Dan

 

This is the heart of the issue with the traditional german concertina and anglo tunings and why the harmonica comparison is not relevant to Chris' initial question at the top of this thread -- even though he himself made the comparison! Most systems of diatonic harmonicas (mouth organs) have at most one note that is available both on the blow and draw directions. Even a typically-tuned 2 row 20-key german concertina, though it adds only one new named note that would not be present on a 1-row concertina (or harmonica) has many notes duplicated in both directions. THis provides the option for many more harmonized intervals (thirds, etc.) as well as multiple major triads, minor triads, diminished triads, suspended fourth triads, and even some 4-note "seventh" and "sixth" chords. It is really a wonder how many of these can be played on a 2-row (especially if the bottom left key of the LH inside row is tuned efficiently), and adding even a 3-key third row (as in 26 key anglos) increases their number more dramatically still.

 

To optimize the harmony within the set of draw notes, and again within the set of press notes, and taking into account that you never will be able to play at the same instant a note in one direction together with a note in the other direction, the german concertinas (and many early multiple- row button accordions) employed various tunings in which the pitch value for the "same named note" on the press direction was allowed to deviate from the pitch value for that "same named note" on the draw direction. Those are the tunings I call "decoupled." The original tunings of most vintage anglos were different from the traditional tunings of 2 row german concertinas, but again were "press/draw decoupled."

 

This option is scarcely available on a diatonic mouth organ.

 

Paul

Edited January 30, 2007 by Paul Groff

[Chris Allert]

This is the heart of the issue with the traditional german concertina and anglo tunings and why the harmonica comparison is not relevant to Chris' initial question at the top of this thread -- even though he himself made the comparison! Most systems of diatonic harmonicas (mouth organs) have at most one note that is available both on the blow and draw directions. Even a typically-tuned 2 row 20-key german concertina, though it adds only one new named note that would not be present on a 1-row concertina (or harmonica) has many notes duplicated in both directions. THis provides the option for many more harmonized intervals (thirds, etc.) as well as multiple major triads, minor triads, diminished triads, suspended fourth triads, and even some 4-note "seventh" and "sixth" chords. It is really a wonder how many of these can be played on a 2-row (especially if the bottom left key of the LH inside row is tuned efficiently), and adding even a 3-key third row (as in 26 key anglos) increases their number more dramatically still.

 

i agree that harmonica tunings aren't really relevant to concertinas designed for chromatic playing, but it seems that basing the tuning of a 20 button anglo on two harmonicas a pure fifth apart would give the purest sounding chords of most of the chords that are available, even taking into account playing across the rows. but i can see how there are some problems on the pull. (most obviously the bad E-B fifth). and once you start adding a third row things will only get worse. but i still think it would be a good system for players that never leave the c and g rows.

 

maybe it's not a great system for the 40 button concertina dan is building if he plans to use many of those extra buttons.

 

 

to clarify my proposal a little more (i realize the pages i linked to are not well documented):

 

in both the push and pull tunings, the following notes are the same:

 

G B D F# A C#

 

you get these notes by setting D and F# a pure 5/4 major third apart, and the you get G, B, A, and C# as pure fifths above and below above and below D and F#.

 

in the push tuning, C and E are obtained as pure fifths down from G and B, so you end up with pure C-E-G and G-B-D triads.

 

on the pull you get F natural and C as pure 7/6 narrow minor thirds up from D and A respectively. because D and A are a pure fifth apart, so are F and C. E is a pure fifth up from A, and also a pure 9/7 up from C. the reason for using the 7/6 is to make the two chords B-D-F-A and F#-A-C-E sound better. this also means C and E on the pull are quite a bit off from C and E on the push.

 

so on the push on the c and g row, you have C-E-G-B-D, all in pure intervals to one another.

 

on the pull you have B-D-F-A, and tuned a pure fifth above that F#-A-C-E. the F#-F interval will sound harsh, and E-B will be a very bad 27/20 instead of a pure 3/2 as it is on the push. F-A-C will sound a little wierd since it is made up of the 9/7 F-A interval and the 7/6 A-C interval. but F-C on the pull is still a pure fifth, so the triad will still sound nice, and will add a little variety. the best minor triads will be D-F-A and A-C-E. E-G-B on the push and B-D-F# on the pull will sound good, though a little harsher since they are 5-limit minor triads instead of 7-limit. still, they'll sound better than equal temperament. D-F#-A on the pull is a pure major triad. if you add a C# to the pull, then A-C#-E is also a pure major triad, as is F-A-C#. B-D-F#-A and F-A-C#-E will also both sound equivalent and smooth. D-F-A-C should sound good too since D-F is a pure 7/6 interval and A-C are pure fifths above that.

 

as a final disclaimer, when i started this thread i didn't intend to suggest that my tuning proposal reflects how concertinas were actually tuned before equal temperament. it is simply a tuning proposal that i think would be musically interesting and well suited to at least a few styles of playing, in particular the morris style of william kimber as described to me by dan worrall. i'm not sure what to suggest to those interested in historical accuracy. paul, are you taking orders for retuning? how long is the waiting list?

[Dan Worrall]

 

This is the heart of the issue with the traditional german concertina and anglo tunings and why the harmonica comparison is not relevant to Chris' initial question at the top of this thread -- even though he himself made the comparison! Most systems of diatonic harmonicas (mouth organs) have at most one note that is available both on the blow and draw directions. Even a typically-tuned 2 row 20-key german concertina, though it adds only one new named note that would not be present on a 1-row concertina (or harmonica) has many notes duplicated in both directions. THis provides the option for many more harmonized intervals (thirds, etc.) as well as multiple major triads, minor triads, diminished triads, suspended fourth triads, and even some 4-note "seventh" and "sixth" chords. It is really a wonder how many of these can be played on a 2-row (especially if the bottom left key of the LH inside row is tuned efficiently), and adding even a 3-key third row (as in 26 key anglos) increases their number more dramatically still.

 

i agree that harmonica tunings aren't really relevant to concertinas designed for chromatic playing, but it seems that basing the tuning of a 20 button anglo on two harmonicas a pure fifth apart would give the purest sounding chords of most of the chords that are available, even taking into account playing across the rows. but i can see how there are some problems on the pull. (most obviously the bad E-B fifth). and once you start adding a third row things will only get worse. but i still think it would be a good system for players that never leave the c and g rows.

 

maybe it's not a great system for the 40 button concertina dan is building if he plans to use many of those extra buttons.

 

 

to clarify my proposal a little more (i realize the pages i linked to are not well documented):

 

in both the push and pull tunings, the following notes are the same:

 

G B D F# A C#

 

you get these notes by setting D and F# a pure 5/4 major third apart, and the you get G, B, A, and C# as pure fifths above and below above and below D and F#.

 

in the push tuning, C and E are obtained as pure fifths down from G and B, so you end up with pure C-E-G and G-B-D triads.

 

on the pull you get F natural and C as pure 7/6 narrow minor thirds up from D and A respectively. because D and A are a pure fifth apart, so are F and C. E is a pure fifth up from A, and also a pure 9/7 up from C. the reason for using the 7/6 is to make the two chords B-D-F-A and F#-A-C-E sound better. this also means C and E on the pull are quite a bit off from C and E on the push.

 

so on the push on the c and g row, you have C-E-G-B-D, all in pure intervals to one another.

 

on the pull you have B-D-F-A, and tuned a pure fifth above that F#-A-C-E. the F#-F interval will sound harsh, and E-B will be a very bad 27/20 instead of a pure 3/2 as it is on the push. F-A-C will sound a little wierd since it is made up of the 9/7 F-A interval and the 7/6 A-C interval. but F-C on the pull is still a pure fifth, so the triad will still sound nice, and will add a little variety. the best minor triads will be D-F-A and A-C-E. E-G-B on the push and B-D-F# on the pull will sound good, though a little harsher since they are 5-limit minor triads instead of 7-limit. still, they'll sound better than equal temperament. D-F#-A on the pull is a pure major triad. if you add a C# to the pull, then A-C#-E is also a pure major triad, as is F-A-C#. B-D-F#-A and F-A-C#-E will also both sound equivalent and smooth. D-F-A-C should sound good too since D-F is a pure 7/6 interval and A-C are pure fifths above that.

 

as a final disclaimer, when i started this thread i didn't intend to suggest that my tuning proposal reflects how concertinas were actually tuned before equal temperament. it is simply a tuning proposal that i think would be musically interesting and well suited to at least a few styles of playing, in particular the morris style of william kimber as described to me by dan worrall. i'm not sure what to suggest to those interested in historical accuracy. paul, are you taking orders for retuning? how long is the waiting list?

 

Chris,

Many thanks for all your ideas on this! I'll give it a good think, which for me goes verrrryy slowly. I have an old spare instrument that I may even try it out on.

 

For the way I play, the most used of the extra buttons on a 40 button instrument (and for that matter the top row of the 30 button) are the ones that either duplicate notes in a different direction, give me an extra bass note that is the same as one I already have an octave higher, or allows me to make a full chord out of one that would otherwise just be partial. 95% of the time I find that I am playing one of the chords of the three chord trick anyway, or some inversion of it. I know there are players who travel all over the instrument playing in keys far from the home rows (usually melodically only, but sometimes with chords), but I find that 80-90% of the dance music I play is in G or D (hence a GD concertina seems best to play in a chorded style). I stretch myself with occasionally going to key of A (one more sharp) but rarely go the other way to C (one less sharp) on a GD instrument. This is a lengthy way of saying that a system such as you propose is not necessarily worse on a 40 button than for a 20. Obviously, such would be tailored specifically to a chorded style, and would not be everybody's cup of tea; I suspect ET will always be more popular for folks who want to be able to do it all...ie also play in multiple keys, or play a lot of chromatic music like jazz.

 

By the way, most who play in a chorded style do NOT play 'along the rows', but spend almost all their time cross row fingering between the G and D rows on a GD instrument. This style of cross row fingering goes all the way back to Hoeselbarth and Minasi in the middle 1840s (see Concertina Library). While considering these early styles, it is worth mentioning that they were masters at getting interesting chords out of a humble 20 button instrument. Paul listed quite a few chords that are possible on a twenty....and both these early players used them (for example check out 'God Save the King' on Minasi, which Minasi plagiarized from Hoeselbarth...it has a different title in German).

 

Keep thinking...the ideas are great!

Dan