On his deathbed, the great blind composer Charles Triangle composed the following remarkable piece, but his penchant for puzzles has left critics in tears searching for a deeper meaning to a piece whose significance seems, upon listening, to lie not merely in the sound.

Do you have the ability to follow where his fingers wrote, and uncover the poetic words intimated here, that occupied the mind of this immortal composer as he recalled his youth?

enter image description here


While you could not be expected to identify the meaning of each bar without some thought, I can assure you that a number of bars (e.g. the first, second, sixth, seventh, tenth, fourteenth, twenty-fourth, twenty-fifth, twenty-sixth, twenty-eighth, forty-eighth, forty-ninth, fifty-third and fifty-fourth) will yield their meaning at a glance once you get a feel for the composer's method, remembering all you have been told about him. Once you have interpreted each bar, you will be one step closer to recovering the composer's meaning.

Hint 2:

As it has now been pointed out by one particularly astute critic that the patterns formed by the outlines of each note or chord within a bar may be interpreted to deduce a letter with some level of certainty, I would prompt you to go one step further and realize the composer's infallible rule that leaves nothing to chance in the placement of each note. Bear in mind, of course, that at his advanced age the hills and valleys that mark the path of life had been transformed into one unending climb to glory!

Hint 3:

It seems that while a certain astute critic has grasped the meaning of the notes and chords within each bar (although a level of uncertainty still remains), the significance of their position on the stave remains as yet unelucidated. Bearing in mind that critic's tentative interpretation of each bar, examine more closely bars nineteen, thirty-seven, forty-two, and fifty-five, and see if you can learn something thence which might be more broadly applicable.

  • $\begingroup$ Is the (broken) link supposed to be an audio recording of the song? $\endgroup$ – Weather Vane Apr 19 '20 at 11:09
  • 1
    $\begingroup$ @WeatherVane It is. Does it work now? I don't know what happened. It's not essential to the solution - just for flavour. $\endgroup$ – Anon Apr 19 '20 at 11:34
  • 1
    $\begingroup$ Yes thanks – it's rhapsodic! $\endgroup$ – Weather Vane Apr 19 '20 at 11:36
  • 2
    $\begingroup$ Maybe we can get a hint? $\endgroup$ – sarsaparilla Apr 28 '20 at 11:52
  • 1
    $\begingroup$ @sarsaparilla That's a great idea! $\endgroup$ – Anon Apr 28 '20 at 12:24

Final answer: The secret to 'A Radiance Enduring' is:


TL;DR - There are several steps to solving this cipher:

1. Convert each bar to a letter of the alphabet through its similarity to a Braille symbol.

2. Notice that for each occurrence of the same Braille-bar letter in the cipher the first chord of notes in the bar is consistently offset from the last chord of notes in the previous bar by the same amount. Similarly, in any bar with two chords, the second chord is consistently offset from the first by the same amount. (NB For any given letter, the two offsets are not necessarily equal.)

3. Write out a table of the offsets for each letter of the alphabet, arranged in alphabetical order:
enter image description here

4. Create one long string of numbers by concatenating the offset values in alphabetical order:

5. Decode the resulting string using a Base 26 cipher to get the message "A WHAT THOUGH THE RADIANCE WHICH WAS MIND".

6. Realise that the composer is quoting an ode by William Wordsworth here (which specifically concerns reassurance to others after one's death - topical to our composer's situation), and that the word which can be substituted for "WHAT THOUGH THE RADIANCE WHICH WAS" is BRIGHT...

...giving us the final answer: A BRIGHT MIND!

Now for the full explanation, reflecting my solving process... (Including an account of the various errors of judgement I made, and how I repeatedly made this whole thing unnecessarily hard for myself!)

The first thing to note upon considering this puzzle is that from the description:

Its composer, Charles Triangle, is blind. Automatically this should set us to thinking about BRAILLE.

enter image description here

With this in mind, looking at each bar we see that:

Each bar can indeed be interpreted as a specific Braille letter, if we treat each written note as an individual Braille dot. Importantly:

- Where a bar contains two separate notes or chords (which is how I shall refer to several notes played at the same time), these correspond to the two columns of dots in a Braille symbol;

- Where a bar contains just one note or chord, this implies that the right-hand column of the Braille symbol contains no dots, i.e. it is one of A, B, K or L, depending on how many notes are present;

- Where two notes in a chord are touching, the dots in the Braille symbol are adjacent, one directly on top of the other (like the left-hand column of a Braille G), whereas when they are separated by a tone, with clear space in between them, the Braille dots similarly appear at the top and bottom of their column, with a gap between them.

By following these rules, we can soon see that the first line...

would work out as...

A | Q | (MOU) | (CEI) | (CEI) | K | B | (PRV) | (MOU) | W | (NZ)

where unambiguous letters are indicated in bold, while ambiguities are presented in brackets. A little thought reveals that this can spell out 'A QUICK BROWN', which should instantly be recognisable to most regular puzzle solvers as the opening words of a famous pangram. Proceeding through the next two lines, we can resolve further ambiguities to continue and complete the pangram 'tribute act' (since it isn't quite a pangram here, as the lack of 'the' means it lacks T and H) - this serves as a confirmatory message that our approach is on the right track.

Seeing this through results in the first bar after the 'pangram' message being the one which is headed "Three Commands". This implies that the coded letters which follow this should provide us with three instructions to carry out to solve the puzzle.

Applying the same techniques to the rest of the coded message results in something of the form:

A | Q | U | I | C | K | B | R | O | W | N
F | O | X | J | U | M | P | S | O | V | E
R | A | L | A | Z | Y | D | O | G | (CEI) | (MOU)
(MOU) | (NZ) | (GT) | (FHS) | (FHS) | (CEI) | (FHS) | (GT) | (FHS) | (MOU) | (PRV)
(DJ) | (CEI) | (PRV) | A | L | (GT)
(CEI) | (PRV) | B | A | (FHS) | (FHS)

What can we make of this? Well, just by trying to form real (and potentially relevant) words among these letters, one potential interpretation which fits is:

COUNT SHIFTS ORDER ALTER BASS/BASH - where the last word could be one of two things in a music/cipher context. The following diagram shows my rough working in arriving at this possibility...

enter image description here

- All bars coloured green are unambiguous and have only one possible letter.

- All others are colour-coded to match other bars with the same Braille pattern, with my initial best-hypothesis prediction made bold and underlined in the letter suggestions above them.

At this point in my working the OP hinted that particular bars (boxed in blue here) might be useful in helping to pinpoint the identity of the letter in the final bar. Each of these corresponds to the letter 'S' in my projected message. The OP went on further to recommend considering the missing Braille dots also - this turned out to be a critical breakthrough in the puzzle that I might otherwise not have considered... Ignoring the missing second-column dots for A, B, L and K, I plotted these onto the same diagram:

enter image description here

It then became clear that the two columns/chords of Braille dots in bars which represented the same letter always had the same offset (or 'shift' - that word is important...), i.e. the difference in pitch between the highest notes of the two columns in that bar was always constant. For the final bar in the piece of music, the highest note of the first chord is a C and the highest of the second is a D one octave higher - a difference of 8 notes (C-D-E-F-G-A-B-C-D - count the hyphens here, not the letters). Since previous F's had a shift of only 3 notes, and H only 2, this final bar therefore had to represent an S, which always had a shift of 8 notes in a bar everywhere it was encoded.

In fact (and it took me an embarrassingly long time to realise this), every instance of a letter also has a consistent number of shifts before it, i.e. between the highest note of the last chord of the previous letter and the first chord of its own (for a bar representing an S, the first chord is always shifted by 5 from the last chord of the previous bar, for example). This means that each of the 26 letters of the alphabet can be represented by a specific pair of numbers, almost like a set of coordinates: (shift before bar, shift within bar).

Using this method also confirms that all of the other letters I had previously derived were correct, which finalises our instructions as:


This can best be split into: COUNT SHIFTS, ORDER, ALTER BASS. The question now is: How do we carry out these instructions?

1. COUNT SHIFTS - this is what we have already done. We have counted the note shifts that occur between two different bars and also those that occur within a bar. We have noticed that the shifts are consistently the same for each occurrence of the same letter.

2. ORDER - a sensible next step is to put them in a table. To that end, the shifts for each letter are as follows:
enter image description here

Conveniently (and this was confirmed by the OP to corroborate that things were moving in the right direction), in the construction of this table we have ordered the letters and shift-related numbers exactly how we are supposed to (i.e. alphabetically). What we now have is a sequence of pairs of numbers, where we have left the 'within letter' value blank if only one Braille column is required in the corresponding symbol.

3. ALTER BASS - to carry out this instruction, we have to note the use of a pun here. When the OP writes 'alter bass', what they intend is for us to 'alter the base' of a number in this puzzle! (In fact, this was something I had suggested in an earlier partial answer, but when push came to shove I had little success deriving this correctly and needed some very patient and persistent encouragement from the OP to point me in the right direction - more on that shortly...) What number can we change the base of? Well, why not concatenate our ordered pairs of shifts to form one long number:


So what base do we convert this number to? In hindsight, the obvious answer when trying to convert numbers into letters is to decrypt this using base 26 (and in my defence, I did try this initially but ended up with gibberish that led me nowhere - it turned out this was because the online tool I used to decrypt it only permitted a certain number of digits in the number to be decoded and truncated my input! I recommend this tool for your decryption purposes...). Doing so provides a string of letters:


This is the message that has been encoded in the sheet music by Charles Triangle. But it needs some interpretation. Noticing the tag, we need to realise that:

Here, Charles Triangle is quoting a section from a William Wordsworth ode, Ode on Intimations of Immortality:

What though the radiance which was once so bright
Be now for ever taken from my sight,
Though nothing can bring back the hour
Of splendour in the grass, of glory in the flower;
We will grieve not, rather find
Strength in what remains behind;
In the primal sympathy
Which having been must ever be;
In the soothing thoughts that spring
Out of human suffering;
In the faith that looks through death,
In years that bring the philosophic mind.”

(Note the use of 'intimated' and 'immortal' in the puzzle's original flavourtext both clue the title of this poem!)

If we therefore parse our encoded phrase as "A 'WHAT THOUGH THE RADIANCE WHICH WAS' MIND", we can infer that the central phrase here should be substituted by the word at the end of the first sentence in the Wordsworth poem: BRIGHT.

Altogether, this means that the secret to Charles Triangle's piece, 'A Radiance Enduring' is A BRIGHT MIND, likely suggesting the poem should be taken as comfort to others after his own death and implying that a bright mind will leave behind creations whose radiance will endure far longer than one lifetime... Deep words!

One final note about a part of the puzzle I did not use in my solving process:

Both Charles Triangle's name and the musical instruction 'echeggiante' were intended as subtle pointers to part of the solution. 'Echeggiante' is an Italian word meaning 'echoing' and the Greek capital letter 'delta' is shaped like a triangle. 'Echo', 'Delta' and 'Charlie' (a more familiar name for 'Charles') are keywords in the NATO Phonetic Alphabet and were intended by the OP as a little hint towards step 2 of the 'Three Commands', indicating that the Braille letter shifts once decoded should be arranged alphabetically. In fact, (OP's intention from comments) even the title could be seen to be cluing further codewords if you link 'Alpha' to 'Radiance' (alpha radiation) and 'Bravo' to 'Enduring' (bravo --> 'braving'), which then means the first 5 codewords all appear in alphabetical order...

Final remarks: This puzzle was nicely put together. The idea of encoding a message in a musical score was inspired, and doing so using Braille was a real novelty. The encoded commands should have been sufficient for me to solve the puzzle with some trial-and-error, although something inbuilt to help confirm when the second command was complete would have been very useful. I am glad I persisted with the solving process, as it is great to see this satisfying puzzle brought to its clever conclusion. Part of me wishes other puzzlers had joined the fray (I was really flailing about at some times!) but in everybody's defence it was a difficult puzzle to get into and required some real time investment. Rewarding though - it's a good feeling to bring this one home! :)

  • $\begingroup$ Your initial decoded phrase contains "a" instead of "the" and so doesn't actually have the desired property. $\endgroup$ – AxiomaticSystem Apr 30 '20 at 17:17
  • $\begingroup$ @AxiomaticSystem True, it doesn't have the property I mention but there's no denying that's what the OP os referencing... $\endgroup$ – Stiv Apr 30 '20 at 17:20
  • $\begingroup$ Extended discussion between Stiv and Anon has been moved to chat. $\endgroup$ – Stiv May 9 '20 at 13:28
  • $\begingroup$ @Stiv Well congratulations on finally solving this puzzle. I can go to bed now... Nsgre nyy gung jbex, lbh npghnyyl cebirq gur oevtugarff bs lbhe bja zvaq ol tyrnavat n zrnavat sebz gur dhbgr orlbaq jung V vagraqrq. V vagraqrq "jung gubhtu gur enqvnapr juvpu jnf... zvaq" gb fvzcyl pyhr gur irel cnffntr lbh dhbgrq - "jung gubhtu gur enqvnapr juvpu jnf" gb tvir gur fgneg bs gur dhbgr (gb ybat gb rapbqr nf n jubyr) naq "zvaq" gb fubj jurer gb svavfu, ohg lbh chyyrq nabgure zrnavat bhg bs vg. Lbhe cuvybfbcuvpny ersyrpgvbaf ner fcbg ba gubhtu, naq ersyrpg gur pbagrag bs guvf ornhgvshy cnffntr... $\endgroup$ – Anon May 9 '20 at 14:39
  • 1
    $\begingroup$ @Anon You're very welcome :) It genuinely kept me up at night sometimes - and that's a compliment, not a complaint! $\endgroup$ – Stiv May 9 '20 at 14:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.