Using colour to convey musical pitches – Part 2

In Part 1, I observed how the difference between sound waves and light waves has severed the vocal practice of canntaireachd from its written forms. In this part, I explain the rationale for a colour system that could heal the rift, making canntaireachd easier to handle.

I have been using colour to communicate pitches in music education since 1995. That was when I bought rainbow marker pens and stuck several hundred spots on a fleet of xylophones and metallophones in St George’s English School, Rome. My objective was to get whole classes of children aged 3–12 making music that sounded good. I didn’t want a repeat of my own childhood experience in Glasgow – a class of thirty 7-year-olds blowing recorders sounds horrible!

Using routines like “Stick on your nose! Stick on your tummy! Stick on your head! Stick on the red note!” I found I could quickly get results with young children that not only preserved my sanity as a classroom music teacher, but got passers by to put their ears to the door. The key to success is a systematic use of colour.

In mainstream music education, colours are generally assigned to the white notes of the piano as follows:

C redD orange | E yellow | F green | G blue | A indigo | B violet

The association between the colours of the rainbow and the notes of the octave is very old. The caricature above by Charles Germain de Saint Aubin (1721–1786) is of Louis-Bertrand Castel’s ‘ocular organ’. The earliest evidence I have found is in Isaac Newton’s writings on optics; however, I suspect the idea was commonplace centuries if not millennia before Newton:

an Assistant whose Eyes for distinguishing Colours were more critical than mine, did by right Lines αβ, γδ, εζ, &c. drawn cross the Spectrum, note the Confines of the Colours, that is of the red MαβF, of the orange αγδβ, of the yellow γεζδ, of the green εηθζ, of the blue ηιχθ, of the indico ιλμχ, and of the violet λGAμ. And this Operation being divers times repeated both in the same and in several Papers, I found that the Observations agreed well enough with one another, and that the rectilinear Sides MG and FA were by the said cross Lines divided after the manner of a musical Chord. Let GM be produced to X, that MX may be equal to GM, and conceive GX, λX, ιX, ηX, εX, γX, αX, MX, to be in proportion to one another… and so to represent the Chords of the Key, and of a Tone, a third Minor, a fourth, a fifth, a sixth Major, a seventh and an eighth above that Key

The Second Book of Opticks. Part II (2nd edition, 1718), Prop. III. Prob. I

Unfortunately, a 7-colour system that repeats at the octave doesn’t transfer well to the Highland bagpipe. The pipe has nine scale degrees and Low G and high G possess very different tone colours. In about 1760, Joseph MacDonald wrote:

Joseph MacDonald's 'Compleat Theory', folio 20r

This is the beginning of a very soft Lament, dedicated to the Chief of the MacCleods of Sky. It touches upon [high] G in the Ground frequently, & in the runnings the force of the style lies upon the Lowest G, so that we may say it is a different Species of that Key.

There is another more basic reason for rejecting a 7-colour system. Part 1 identified the fundamental problem as ambiguity, noticing that ho stood for either C or B, and hin stood for either low A or low G. For a Highland pipe colour system to be fit for purpose, making Gaelic lyrics as well as canntaireachd vocables easier to handle, nine colours are essential.

The 9-colour system I introduce in this series is custom-made for Highland bagpipe music. The colours are assigned to pitches in a way that is sympathetic to long-established melodic conventions, bringing out the tonal contrasts between pieces and phrases. Rather conveying the relative height of each pitch in the scale, the colours convey their distinctive timbral characteristics. This is done by taking the drones into consideration.

The tonal quality of each scale degree is defined by its sensory dissonance, or the combination of chanter and drone spectra. This is basically the extent to which drone and chanter harmonics coincide. The following graph shows curved dissonance peaks and sharp dissonance minima across the range of the chanter:

This is my reworking of a graph published by Jonathan A. Kemp in Piping Today, 75 (2015), p. 19. The minima ranked 1–9 correlate strongly with an intonation ideal for which we have evidence spanning four centuries: from the chanter of Iain Dall MacKay (c. 1656–c. 1754), one of pibroch’s greatest composers, to the recordings of Donald MacPherson (1922–2012), the most successful competitive piper of the 20th century.

Here are the colour names I would use in everyday speech and educational materials:

Red | Black | Orange | Indigo | Yellow  | Grey | Green | Pink | Blue

The web names in the text colour palettes of WordPress, Gmail and other applications are slightly different:

Red | Black | Amber | Blue violet | Gold  | Medium gray | Sea green | Magenta | Sky blue

The hex colour codes for designers and publishers are as follows:

#ff0000 #000000 | #ff9900 | #8A2BE2 | #ffcc00 | #999999 | #339966 | #ff00ff | #00ccff

As CSS and HTML allow for 16.777216 million colours, these values are not prescriptive. I will be tweaking them myself in efforts to polish the system further. They have the practical advantage of being currently available in the default ‘Text color’ palette of WordPress, with the exception of Blue violet. But it is easy to get any colour in WordPress: just click on ‘Custom…’ in the palette and type in the hex code.

One of the principles of this 9-colour system is to give adjacent scale degrees complementary hues. This reflects the tonal behaviour of Highland pipe music, making the contrasts within and between pieces clearer to the eye. A key signature of two sharps tells us nothing about the key or ‘taste’ of a bagpipe piece. This colour system aims to help to deepen musical understanding by allowing readers to see the tonality at a glance, revealing how the music works.

So-called Question and Answer phrases generally shift a section of melody (or a single pitch) up or down a tone. The result is a dynamic separation between two interlocking sets of pitches. They behave rather like vinegar and olive oil – sometimes blended, sometimes separate, depending on human intervention:

Set 1: low G | B | D | F | high A (vinegar)

Set 2: low A | C | E | high G (olive oil)

Melodically, the ‘vinegar’ pitches tend to clump together in opposition to the ‘oil’ pitches. Composers combine and contrast them in a multitude of ways to change the ‘taste’.

The reason for assigning the colours of the rainbow to set 1 can be seen in Kemp’s graph above: as these five pitches ascend, their sensory dissonance falls. Set 2 fills the remaining colour space with the objective of maximising two things:

  1. the ease with which all nine colours in the system can be distinguished; and
  2. the clarity with which tonal contrasts in the music become visible on the page.

The neutral colours black and grey are appropriate for the tuning notes, low A and E,  because the impact these pitches have on tonality is minimal. Their prominence in the drone spectra makes them effectively omnipresent without ever being played. All pieces contain E and over 99% contain low A; what defines a Highland bagpipe tonality is the presence or absence of the other seven pitches.

Melodically, C is most often paired with low A and High A – and contrasted with B and D. The choice of indigo responds to this traditional tonal behaviour. Although it has the web name ‘blue-violet’, it is also called ‘deep indigo’ and lies between electric indigo and pigment indigo.

The colour of high G is close to low G in hue. The tonal contrast between high G and high A also evokes the sunsets I witnessed when living on the Isle of Skye. Never have I witnessed so many spectacular skies as looking from Waternish across loch Dunvegan, past Borreraig, over to the mountains of Harris.

During the Christmas holidays, I was clearing out my filing cabinet and came across this printout from 1998:

9-colour system developed in Berlin, 1998

This 9-colour system was developed for an 8-year-old pupil in Berlin, where I was living at the time. The only revision I have made now is to increase the distance in hue between members of the ‘oil’ set, making E grey. Splitting the difference between the 1998 colours of C–E and E–G, giving the results to C and G, makes the four members of this set easier to tell apart.

I find it reassuring that a colour system I developed purely on the basis of cultural knowledge (i.e. as a result of hearing and playing Highland bagpipe music) ties in so well with Jonathan Kemp’s graph of sensory dissonance. This scientific support gives me confidence to resurrect something child-friendly as a solution for my PhD thesis.

I did not publish it 19 years ago because 1) it did not photocopy well, 2) none of the piping magazines then were in full colour, and 3) most desktop printers were black and white. Today, multifunctional and digital copiers are fast consigning black and white analogue copiers to history, all the piping magazines are in full colour, colour inkjet printers are ubiquitous and everyone has a colour screen. With these obstacles removed, I feel its time has come.

I have not taken colour blindness into account and would welcome feedback from anyone who finds any of these colours difficult to distinguish. An idea offered in a comment to Part 1 was to use lightness values that would produce nine shades of grey. This would make the system backward compatible for black-and-white printing, but this advantage is outweighed by the disadvantage of poor readability due to low contrast between adjacent shades, and between lighter shades and the background.

To help make the colours easy to read, both on screen and on paper, my approach is to use boldface, a larger font size and to choose a font in which the ‘bold’ is extra-bold, like a headline (e.g. Georgia bold, point size 20).

Next week, in Part 3, I will look at Scottish Gaelic phonology, recordings of Rona Lightfoot, and the unorthodox spellings of Gaelic words used by our three earliest witnesses of canntaireachd: Colin Campbell, Niel MacLeod of Gesto and Angus MacKay. If canntaireachd is to be a useful tool, we need to be confident singing it. Rona is our greatest living exponent. What she does makes it much easier to understand the historical sources, getting closer to the pronunciation of John MacCrimmon.


4 thoughts on “Using colour to convey musical pitches – Part 2

  1. Very good. Your rationale speaks very much about making it idiomatic to the pipes. How does that sit with the idea that this repertory is shared between pipes, harp / lyre & fiddle? There’s an implication here that the notes of the pipe is the “norm” which lyre, harp or fiddle adapt themselves to. Maybe this is true but I’d like to see more discussion of this assumption.

  2. Well Barnaby, for what it is worth, I am slightly colour blind  as I discovered when I left my Nautical Training school to continue training as a deck officer in the Merchant Navy. The alternative was to switch to Engineer Officer but I did not fancy the boiler room, nae windows. So like a large proportion of the male sex I have problems in the Green spectrum. Not dangerously so, the primary colours are fine it is  more to do with ‘shades’. In that connection your two different ‘name’ examples are illuminating.

    In the first one where you have the words  ‘grey and green’  side by side I needed to effectively do a ‘double take’ to distinguish a difference. However in the second example where they are I assume exactly the same colour but now called ‘medium grey and sea green’ there is much less of a problem. This would seem to suggest that size of the coloured example is also relevant which as coloured notes will be smaller would not at least for even mild degree of colour blindness make for easy reading.

    Regarding the overall concept that you are proposing, well yes and err no. In any case Rona does not use either CC or Gesto, so looking forward to how you find that useful. But the main question I have with it is that the two written forms of canntaireachd are in fact the exceptions rather than the rule. Joseph MacDonald was musically literate so he went straight to notes on lines. (and by the way I think that his MS as we have it now was compiled from drafts after he had arrived in India, the self portrait certainly was).

    The CC was written as such because not having been taught to write music on the stave,  turning their oral form into ‘written words’ was the only option, probably as I have argued elsewhere due to the needs of a patron who was increasingly going deaf. Gesto may have been musically literate to a degree but again he required Alexander Campbell to actually turn it into a more accurate dots on lines so again writing it as per the syllables was an easier option.

    But to my point, surely with modern technology where sound clips can easily be dropped into text in any online device if anyone wants to use canntaireachd for teaching purposes then cutting out those two written forms and going back to its origin as a purely aural/oral medium should be the route?

  3. I do not mean to suggest that coloured texts are a substitute for more detailed notations, only that they can play a valuable role complementing them. An audio recording is even more prescriptive than staff notation: both shut off options, closing down the listener’s imagination.

    The case for using colour is that it makes it easier for the reader to tell editorial interventions apart from the information encoded in the raw source material. It offers superior transparency and honesty: faithfulness to what is and isn’t in the source material.

    Simon writes, “There’s an implication here that the notes of the pipe is the “norm” which lyre, harp or fiddle adapt themselves to.” That is neither my intention nor assumption, I had simply failed to write in a way that included non-pipers. What I omitted to say was that this system has a very limited application. It is intimately tailored to the Highland bagpipe and serves as a ‘light’ editorial intervention, intended to make raw source materials for pibroch easier to handle without ‘cooking’ the source as much as audio and staff interpretations. It is not designed for other instruments.

Leave a Reply