Well, I haven't used this knowledge in about three years, but here goes: In the Pre-Baroque/Baroque/Classical era, not all keys are equal. Some math to follow. A scale is divided into twelve chromatic intervals from do to do' (octave higher). There are several ways to determine how to space these intervals; the term for the way the pitches in a scale are defined is 'temperament.' There are three basic systems to be discussed: Even, odd, and just. Odd-tempering is the oldest system; the term is a catchall, but generally refers to Pythagorean tuning. To build an odd-tempered scale: Start with a base frequency for C. Define C' as 2C, define C'' as 4C, etc. following C^n = C2^n. Now things get more difficult. Pythagorean tuning is based around the interval C-G, the perfect fifth, defined such that G = 3C/2 (in other words, the frequency of G is 1.5 times that of C). By the same token, F below C can be defined 'F (F an octave lower) = 2C/3. Therefore, by the octave definition established earlier, G' = 2G = 3C, ''F = 1/3 C, etc. You have now defined the "tonic," "dominant," and "subdominant." To define the other intervals, rely on the definition of the perfect fifth. Thus: C = 1C G = 3C/2 D' = a fifth above G = (3C/2) = 9C/4 A' = 27C/8 E'' = 81C/16 And so on until you've defined at least one occurence of the "dominant series," C-G-D-A-E-B-F#. Divide their frequencies using the octave definition to bring them back into the range of the scale (thus D is 9C/8, A is 27C/4, etc.). Now, establish the "subdominant series" by multiplying the tonic by the subdominant definition 'F = 2C/3. You will obtain the series C-F-Bb-Eb-Ab-Db-Gb. Raise those tones to insert them into the octave. Those of you following along will have already noticed a problem -- F# is supposed to be enharmonic to Gb, yet violates the definition of 'Gb = 1F#/2. You will furthermore note that the discrepancy increases the more iterations from your base C you are. However, all notes are defined relative to C (actually, all notes are defined relative to A, but that's a separate issue). [more]
This archive was generated by hypermail 2.4.0: Sat 12 Feb 2022 12:30:42 AM EST EST