[Cat]

Here is why there are 26 shapes in the Dora Bella cipher, I think.  Basically, we are just dealing with embelishments around a a "+" and an "x".  That's it.

Now if we just knew what the darn thing said . . .

[Aaron]

Yeah, the general consensus seems to be a mix between substitution and cipher, since no matter which way I've written the letters into a substitution solver I didn't get back much of interest. Of course, those solvers assume it's in English...

Also, I can only see 24 shapes in the Dorabella Cipher, which leads me to wonder what 25 and 26 could be, or if a couple letters were simply "left out" of the substitution.

[Cat]

You are right -- I should rotate the last shapes properly, and  in a certain light while 24 shapes are possible, only 22 are used.  I will take a better photo and post to show what I am talking about.

(I am using this frame because it presumes orientation might matter)

For example, in an octive, a note may have 2 names.  The black key C-Sharp is also D-Flat.  So perhaps we are looking at a cipher where several shapes mean the same thing, only adjusted 8 steps?  That would still leave us with a 24 letter Alphabet (8 * 3)

Therefore, we might have a written variation of the following notes starting with a-("a Flat" in the bottom left corner) until we go full circle:

a# B# C# D# E# F# G# A# <--the top level has 3 curls at 8 different *possible* angles
a    B   C    D    E    F   G    A   <-- the middle level has 2 curls at 8 different angles
a-  B-  C-   D-  E-  F-  G-  A-  <-- this bottom level has only 1 curl at 8 different angles

*possible because only 6 curls of three are used


So now we start counting down how many notes there really are because *some* notes equal others, and hitting a number like 13 actual would be sweet, because that would mean there was room for 8 white keys and 5 black on a piano.

a#  *   C# D#   *  F# G# A#
a    B   C    D    E    F   G    A   = 15 (sniffle! 2 over --unless the last 2 (A and A#) are nulls)
a-  *    *     *    *      *    *     *


Edit:

On second thought, 15 may be just about perfect as well -- because that is 3 ways to symbolize the numbers 1 through 5 for a "5x5", and all the copies as noted when hoping for 8+5 still fill their function . . .

[Aaron]

7 white keys, 5 black keys per octave means the whole cipher could be 2 octaves on a piano precisely. Which fits with Elgar being a composer and all. I kinda wonder if this was never meant to be a message at all in the first place, but *just* a musical piece (albeit a pretty one I'm sure). The question is how do you map the symbols to the keys? I'd wager that it's either all single loops, all double loops, and then all triple loops in some set rotation order or single-double-triple as rotated in a specific order.

If it is a cipher on top of being a musical piece, I would guess at it being some kind of a rotary cipher vs a plain substitution cipher (aka, 1-1-1-1 would be ABCD).

Based on the way it was written, I'm fairly certain it's intended to be read left to right or perhaps right to left.

[Phil_The_Rodent]

I did look into the music idea a while ago.

Some notes: yes, Cat, you're right that C# and D- are the same note, but a composer would use either one or the other -- they don't mix sharps and flats in a key (it's one or the other).

Second, we resolve to root, whatever that happens to be. And we should find the third, fourth, fifth and root (or octave of these) most prominent in a frequency analysis.

There are some very specific rules for melody in counterpoint, but I don't know these at the moment. I am taking a course in this soonish.

If this is a melody, it is a melody with the note lengths removed. Perhaps that is the puzzle part (after decryption).

[Aaron]

Or perhaps it's a very simple melody in which the number of curls denotes note length?

[Cat]

I am taking notes on what both of you say, and am grateful for the help.  I just had a wild idea -- what happens when the lines are extended so that they cross or just intersect other lines?  Would we have an elaborate dot-to-dot picture that then resolves into a message spelled with the regular alphabet, or would there be a picture?  How would it look if we did experiment with putting in the "note lengths"?

[Aaron]

I doubt playing connect the dots would reveal anything... that requires too much creative interpretation and is prone to "filling in the gaps" with our own preconceptions. If that were really intended I think the symbols would have been more regularly spaced in a grid-like fashion.

[Cat]

True . . . (After figuring that out the "hard" way