Relative errors of small EDOs: Difference between revisions

The relative error of an interval in an [[edo]] is the error approximating [[JI]] divided by the size of a single step. The formula for closest mapping: ''error''(''n'', ''r'') = round (''n'' log₂''r'') - ''n'' log₂''r'', where ''n'' is the edo number and ''r'' is the frequency ratio. With closest mapping, the relative error ranges from -0.5 to +0.5. With patent val mapping, it can be farther from zero.  

This article contains two lists. The first shows relative errors of the first 9 prime harmonies for edos up to 99. There is no point for showing higher primes because no edo under 99 is consistent up to them. For other intervals, the relative error follows the additive rule (see below), so they can be derived easily. Also by that rule, however, finding large errors with such ''p''/1 harmonies will not suffice that the edo does a poor approximation in the ''p''-limit overall. One must inspect every relevant interval to be sure of that. The second list comes in naturally for showing the root-mean-squared relative errors of a certain JI subgroup, and may be used as the criterion.   

There are two additivities of relative errors.   

For example, the errors of 3/1 for 26-edo and 27-edo are -0.2090 and +0.2060, repectively, and their sum -0.0030 is the error of 3/1 for 53-edo.   

{| class="wikitable sortable mw-collapsible" style="text-align:right"

! rowspan="2" |Edo

|}

Notes:   

# Patent val mapping is used throughout.