I suspected you were =) so i sent a PM. Your choice of colours isn't loosely based on the science. I understand the need to 'massage' the colours around so that all 12 tones have distinguishable colours, but all your correspondances are offset by at least 2 semitones. Because it doesn't matter if you shift them all one way or the other, you might as well have something that is somewhat based on the science, instead of choosing B for blue, which is completely arbitrary. I suggest this which mostly roughly lines up with the science: Gs reds As oranges B yellow Cs greens Ds blues E indigo Fs violets The commonly known rainbow colours are ROYGBIV but indigo is not actually particularly distinguishable from blue. Isaac Newton seemed to add it because he felt a need to have 7 main spectral colours. Indigo is very different from purple. Purple is not a spectral colour, that is, it cannot be the result of a single wavelength, it is actually a mixture of red and violet wavelengths, and is the section of the colour wheel that 'joins' the 2 ends of the spectral colours. In your system, purple should lie between violet and red, if you want to use it, but your system seems to use only spectral colours. ////////////////////////////// On that subject and referring to my results, it occurred to me that instead of choosing between violet and red for F, F#, we could combine the colours to add purple to the system, making the system suitably cyclical and matching the cyclical colour wheel. Usefully, this adds another colour, to help with the 'multiple reds' / 'mulitple violets' problem. F# appears with very roughly equal amounts of red and violet, so would be close to purple.