Target tuning: Difference between revisions

Line 2:

== Least squares tunings ==

By assumption we have a finite set of rational intervals defining interval classes, which we may take to be octave reduced to 0 < ''r'' < 1, where the intervals are expressed logarithmically in terms of log base two. The least squares tuning T is the tuning for a particular regular temperament which minimizes the sum of the squares of the errors, ∑''i'' (T (''qi'') - log2 (''qi''))2, where ''qi'' are the rational intervals of the target set. Note that most commonly, the target set is a [[tonality diamond]] (reduced to lowest terms and duplicates removed), since these are the intervals in a [[harmonic series]] up to some odd integer ''d'', and hence may be considered the [[consonance]]s of the ''d''-odd-limit.

By assumption we have a finite set of rational intervals defining interval classes, which we may take to be octave reduced to {{nowrap|0 < ''r'' < 1}}, where the intervals are expressed logarithmically in terms of log base two. The least squares tuning, ''T'', is the tuning for a particular regular temperament which minimizes the sum of the squares of the errors, ∑''i'' ({{nowrap|''T''(''qi'') − log2(''qi''))2}}, where ''qi'' are the rational intervals of the target set. Note that most commonly, the target set is a [[tonality diamond]] (reduced to lowest terms and duplicates removed), since these are the intervals in a [[harmonic series]] up to some odd integer ''d'', and hence may be considered the [[consonance]]s of the ''d''-odd-limit.

We may find the least squares tuning in various ways, one of which is to start from a matrix R whose rows are the monzos of the target set, and a matrix U whose rows are [[val]]s spanning the temperament. From U we form the matrix V by taking the [[Normal lists|normal val list]] for U and removing the first ("period") row. A list of [[Eigenmonzo|eigenmonzos (unchanged-intervals)]] E = VRTR, where the T denotes the matrix transpose, can now be obtained by matrix multiplication, and to this we add a row for the monzo of 2. The [[Fractional monzos|projection matrix]] for the least squares tuning is then the square matrix with fractional monzo rows, the rows of E, including 2, as eigenmonzos, that is left eigenvectors with eigenvalue one, and any basis for the commas of the temperament as left eigenvectors with eigenvalues zero.

We may find the least squares tuning in various ways, one of which is to start from a matrix '''R''' whose rows are the monzos of the target set, and a matrix '''U''' whose rows are [[val]]s spanning the temperament. From '''U''' we form the matrix '''V''' by taking the [[Normal lists|normal val list]] for '''U''' and removing the first ("period") row. A list of [[Eigenmonzo|eigenmonzos (unchanged-intervals)]] {{nowrap|'''E'''' {{=}} '''VR'''T'''R'''}}, where the T denotes the matrix transpose, can now be obtained by matrix multiplication, and to this we add a row for the monzo of 2. The [[Fractional monzos|projection matrix]] for the least squares tuning is then the square matrix with fractional monzo rows, the rows of '''E''', including 2, as eigenmonzos, that is left eigenvectors with eigenvalue one, and any basis for the commas of the temperament as left eigenvectors with eigenvalues zero.

== Minimax tuning ==

Starting with the same assumptions as with least squares tuning, minimax tuning works similarly, except instead of minimizing the squares of the errors, we minimize max''i'' |T (''qi'') - log2 (''qi'')|, the maximum error over all the target intervals. This can be solved by setting the minimization up as a ~~[[Wikipedia: Linear programming~~|linear programming]] problem, but another approach leads as before to a projection matrix with fractional monzo rows.

Starting with the same assumptions as with least squares tuning, minimax tuning works similarly, except instead of minimizing the squares of the errors, we minimize {{nowrap|max''i''{{!}}''T''(''qi'') − log2(''qi''){{!}} }}, the maximum error over all the target intervals. This can be solved by setting the minimization up as a {{w|linear programming}} problem, but another approach leads as before to a projection matrix with fractional monzo rows.

For a rank-''r'' temperament, we can form a list of candidate sets of eigenmonzos by taking ''r'' - 1 elements from the list of target intervals, and adding 2 to the set, leading to an ''r''-element set. To avoid duplications and sets with rank less than ''r'', we can take the [[Normal lists|normal interval list]] defined by each of these ''r''-element sets, discarding any with less than ''r'' elements. This means we are compiling a list of rank-''r'' [[just intonation subgroup]]s, which are [[eigenmonzo subgroup]]s for the corresponding tuning. For each of these these subgroups, we may find a corresponding projection matrix as the matrix with the ''r'' generators of the subgroup as eigenmonzos and a basis for the commas as left eigenvectors with eigenvalue zero. These projection matrices define tunings, and the tuning, if unique, with the least maximum error is the minimax tuning. However, this tuning may not be unique, in which case we may break the tie by choosing the tuning, among the set of least maximum error tunings, with the smallest sum of errors squared.

For a rank-''r'' temperament, we can form a list of candidate sets of eigenmonzos by taking {{nowrap|''r'' − 1}} elements from the list of target intervals, and adding 2 to the set, leading to an ''r''-element set. To avoid duplications and sets with rank less than ''r'', we can take the [[Normal lists|normal interval list]] defined by each of these ''r''-element sets, discarding any with less than ''r'' elements. This means we are compiling a list of rank-''r'' [[just intonation subgroup]]s, which are [[eigenmonzo subgroup]]s for the corresponding tuning. For each of these these subgroups, we may find a corresponding projection matrix as the matrix with the ''r'' generators of the subgroup as eigenmonzos and a basis for the commas as left eigenvectors with eigenvalue zero. These projection matrices define tunings, and the tuning, if unique, with the least maximum error is the minimax tuning. However, this tuning may not be unique, in which case we may break the tie by choosing the tuning, among the set of least maximum error tunings, with the smallest sum of errors squared.

== Example ==

Suppose our set of target intervals is the 5-odd-limit diamond {6/5, 5/4, 4/3, 3/2, 8/5, 5/3}. The corresponding matrix is

Suppose our set of target intervals is the 5-odd-limit diamond {{nowrap|{6/5, 5/4, 4/3, 3/2, 8/5, 5/3}<nowiki />}}. The corresponding matrix is

R = [{{monzo| 1 1 -1 }}, {{monzo| -2 0 1 }}, {{monzo| 2 -1 0 }}, {{monzo| -1 1 0 }}, {{monzo| 3 0 -1 }}, {{monzo| 0 -1 1 }}]

'''R''' = [{{monzo| 1 1 -1 }}, {{monzo| -2 0 1 }}, {{monzo| 2 -1 0 }}, {{monzo| -1 1 0 }}, {{monzo| 3 0 -1 }}, {{monzo| 0 -1 1 }}]

Then RTR = [[19 -2 -6], [-2 4 -2], [-6 -2 4]], a positive-definite symmetric matrix. If U = [{{val| 1 0 -4 }}, {{val| 0 1 4 }}], we may remove the top row and obtain V = [{{val| 0 1 4 }}]; then VRTR is [{{monzo| -26 -4 14 }}]; using this along with {{monzo| 1 0 0 }} as eigenmonzos, and using {{monzo| -4 4 -1 }} as a commatic eigenvector with eigenvalue zero, we obtain

Then {{nowrap|'''R'''T'''R''' {{=}} [[19 -2 -6], [-2 4 -2], [-6 -2 4]]}}, a positive-definite symmetric matrix. If {{nowrap|'''U''' {{=}} [{{val| 1 0 -4 }}, {{val| 0 1 4 }}]<nowiki />}}, we may remove the top row and obtain {{nowrap|'''V''' {{=}} [{{val| 0 1 4 }}]<nowiki />}}; then {{nowrap|'''VR'''T'''R''' {{=}} [{{monzo| -26 -4 14 }}]<nowiki />}}; using this along with {{monzo| 1 0 0 }} as eigenmonzos, and using {{monzo| -4 4 -1 }} as a commatic eigenvector with eigenvalue zero, we obtain

[{{monzo| 1 0 0 }}, {{monzo| 14/13 -1/13 7/26 }}, {{monzo| 4/13 -4/13 14/13 }}]

Line 22:

which is the projection matrix for the Woolhouse 7/26-comma meantone, the 5-limit least squares tuning.

Starting from the same six intervals of the 5-odd-limit diamond and adding 2, we find after computing the normal interval lists that the three subgroups are [2, 3], [2, 5], and [2, 5/3]. Computing the projection matrix and from thence the tuning in each case, we find that the minimax tuning is [{{monzo| 1 0 0 }}, {{monzo| 1 0 1/4 }}, {{monzo| 0 0 1 }}], the projection matrix for 1/4-comma meantone.

Starting from the same six intervals of the 5-odd-limit diamond and adding 2, we find after computing the normal interval lists that the three subgroups are [2, 3], [2, 5], and [2, 5/3]. Computing the projection matrix and from thence the tuning in each case, we find that the minimax tuning is [{{monzo| 1 0 0 }}, {{monzo| 1 0 1/4 }}, {{monzo| 0 0 1 }}], the projection matrix for 1/4-comma meantone.

[[Category:Math]]

[[Category:Regular temperament tuning]]

@@ Line 2: / Line 2: @@
 == Least squares tunings ==
-By assumption we have a finite set of rational intervals defining interval classes, which we may take to be octave reduced to 0 &lt; ''r'' &lt; 1, where the intervals are expressed logarithmically in terms of log base two. The least squares tuning T is the tuning for a particular regular temperament which minimizes the sum of the squares of the errors, ∑<sub>''i''</sub> (T (''q<sub>i</sub>'') - log<sub>2</sub> (''q<sub>i</sub>''))<sup>2</sup>, where ''q<sub>i</sub>'' are the rational intervals of the target set. Note that most commonly, the target set is a [[tonality diamond]] (reduced to lowest terms and duplicates removed), since these are the intervals in a [[harmonic series]] up to some odd integer ''d'', and hence may be considered the [[consonance]]s of the ''d''-odd-limit.
+By assumption we have a finite set of rational intervals defining interval classes, which we may take to be octave reduced to {{nowrap|0 &lt; ''r'' &lt; 1}}, where the intervals are expressed logarithmically in terms of log base two. The least squares tuning, ''T'', is the tuning for a particular regular temperament which minimizes the sum of the squares of the errors, ∑<sub>''i''</sub> ({{nowrap|''T''(''q<sub>i</sub>'') − log<sub>2</sub>(''q<sub>i</sub>''))<sup>2</sup>}}, where ''q<sub>i</sub>'' are the rational intervals of the target set. Note that most commonly, the target set is a [[tonality diamond]] (reduced to lowest terms and duplicates removed), since these are the intervals in a [[harmonic series]] up to some odd integer ''d'', and hence may be considered the [[consonance]]s of the ''d''-odd-limit.
-We may find the least squares tuning in various ways, one of which is to start from a matrix R whose rows are the monzos of the target set, and a matrix U whose rows are [[val]]s spanning the temperament. From U we form the matrix V by taking the [[Normal lists|normal val list]] for U and removing the first ("period") row. A list of [[Eigenmonzo|eigenmonzos (unchanged-intervals)]] E = VR<sup>T</sup>R, where the <sup>T</sup> denotes the matrix transpose, can now be obtained by matrix multiplication, and to this we add a row for the monzo of 2. The [[Fractional monzos|projection matrix]] for the least squares tuning is then the square matrix with fractional monzo rows, the rows of E, including 2, as eigenmonzos, that is left eigenvectors with eigenvalue one, and any basis for the commas of the temperament as left eigenvectors with eigenvalues zero.
+We may find the least squares tuning in various ways, one of which is to start from a matrix '''R''' whose rows are the monzos of the target set, and a matrix '''U''' whose rows are [[val]]s spanning the temperament. From '''U''' we form the matrix '''V''' by taking the [[Normal lists|normal val list]] for '''U''' and removing the first ("period") row. A list of [[Eigenmonzo|eigenmonzos (unchanged-intervals)]] {{nowrap|'''E'''' {{=}} '''VR'''<sup>T</sup>'''R'''}}, where the <sup>T</sup> denotes the matrix transpose, can now be obtained by matrix multiplication, and to this we add a row for the monzo of 2. The [[Fractional monzos|projection matrix]] for the least squares tuning is then the square matrix with fractional monzo rows, the rows of '''E''', including 2, as eigenmonzos, that is left eigenvectors with eigenvalue one, and any basis for the commas of the temperament as left eigenvectors with eigenvalues zero.
 == Minimax tuning ==
-Starting with the same assumptions as with least squares tuning, minimax tuning works similarly, except instead of minimizing the squares of the errors, we minimize max<sub>''i''</sub> |T (''q<sub>i</sub>'') - log<sub>2</sub> (''q<sub>i</sub>'')|, the maximum error over all the target intervals. This can be solved by setting the minimization up as a [[Wikipedia: Linear programming|linear programming]] problem, but another approach leads as before to a projection matrix with fractional monzo rows.
+Starting with the same assumptions as with least squares tuning, minimax tuning works similarly, except instead of minimizing the squares of the errors, we minimize {{nowrap|max<sub>''i''</sub>{{!}}''T''(''q<sub>i</sub>'') − log<sub>2</sub>(''q<sub>i</sub>''){{!}} }}, the maximum error over all the target intervals. This can be solved by setting the minimization up as a {{w|linear programming}} problem, but another approach leads as before to a projection matrix with fractional monzo rows.
-For a rank-''r'' temperament, we can form a list of candidate sets of eigenmonzos by taking ''r'' - 1 elements from the list of target intervals, and adding 2 to the set, leading to an ''r''-element set. To avoid duplications and sets with rank less than ''r'', we can take the [[Normal lists|normal interval list]] defined by each of these ''r''-element sets, discarding any with less than ''r'' elements. This means we are compiling a list of rank-''r'' [[just intonation subgroup]]s, which are [[eigenmonzo subgroup]]s for the corresponding tuning. For each of these these subgroups, we may find a corresponding projection matrix as the matrix with the ''r'' generators of the subgroup as eigenmonzos and a basis for the commas as left eigenvectors with eigenvalue zero. These projection matrices define tunings, and the tuning, if unique, with the least maximum error is the minimax tuning. However, this tuning may not be unique, in which case we may break the tie by choosing the tuning, among the set of least maximum error tunings, with the smallest sum of errors squared.
+For a rank-''r'' temperament, we can form a list of candidate sets of eigenmonzos by taking {{nowrap|''r'' − 1}} elements from the list of target intervals, and adding 2 to the set, leading to an ''r''-element set. To avoid duplications and sets with rank less than ''r'', we can take the [[Normal lists|normal interval list]] defined by each of these ''r''-element sets, discarding any with less than ''r'' elements. This means we are compiling a list of rank-''r'' [[just intonation subgroup]]s, which are [[eigenmonzo subgroup]]s for the corresponding tuning. For each of these these subgroups, we may find a corresponding projection matrix as the matrix with the ''r'' generators of the subgroup as eigenmonzos and a basis for the commas as left eigenvectors with eigenvalue zero. These projection matrices define tunings, and the tuning, if unique, with the least maximum error is the minimax tuning. However, this tuning may not be unique, in which case we may break the tie by choosing the tuning, among the set of least maximum error tunings, with the smallest sum of errors squared.
 == Example ==
-Suppose our set of target intervals is the 5-odd-limit diamond {6/5, 5/4, 4/3, 3/2, 8/5, 5/3}. The corresponding matrix is
+Suppose our set of target intervals is the 5-odd-limit diamond {{nowrap|{6/5, 5/4, 4/3, 3/2, 8/5, 5/3}<nowiki />}}. The corresponding matrix is
-R = [{{monzo| 1 1 -1 }}, {{monzo| -2 0 1 }}, {{monzo| 2 -1 0 }}, {{monzo| -1 1 0 }}, {{monzo| 3 0 -1 }}, {{monzo| 0 -1 1 }}]
+'''R''' = [{{monzo| 1 1 -1 }}, {{monzo| -2 0 1 }}, {{monzo| 2 -1 0 }}, {{monzo| -1 1 0 }}, {{monzo| 3 0 -1 }}, {{monzo| 0 -1 1 }}]
-Then R<sup>T</sup>R = [[19 -2 -6], [-2 4 -2], [-6 -2 4]], a positive-definite symmetric matrix. If U = [{{val| 1 0 -4 }}, {{val| 0 1 4 }}], we may remove the top row and obtain V = [{{val| 0 1 4 }}]; then VR<sup>T</sup>R is [{{monzo| -26 -4 14 }}]; using this along with {{monzo| 1 0 0 }} as eigenmonzos, and using {{monzo| -4 4 -1 }} as a commatic eigenvector with eigenvalue zero, we obtain
+Then {{nowrap|'''R'''<sup>T</sup>'''R''' {{=}} [[19 -2 -6], [-2 4 -2], [-6 -2 4]]}}, a positive-definite symmetric matrix. If {{nowrap|'''U''' {{=}} [{{val| 1 0 -4 }}, {{val| 0 1 4 }}]<nowiki />}}, we may remove the top row and obtain {{nowrap|'''V''' {{=}} [{{val| 0 1 4 }}]<nowiki />}}; then {{nowrap|'''VR'''<sup>T</sup>'''R''' {{=}} [{{monzo| -26 -4 14 }}]<nowiki />}}; using this along with {{monzo| 1 0 0 }} as eigenmonzos, and using {{monzo| -4 4 -1 }} as a commatic eigenvector with eigenvalue zero, we obtain
 [{{monzo| 1 0 0 }}, {{monzo| 14/13 -1/13 7/26 }}, {{monzo| 4/13 -4/13 14/13 }}]
@@ Line 22: / Line 22: @@
 which is the projection matrix for the Woolhouse 7/26-comma meantone, the 5-limit least squares tuning.
-Starting from the same six intervals of the 5-odd-limit diamond and adding 2, we find after computing the normal interval lists that the three subgroups are [2, 3], [2, 5], and [2, 5/3]. Computing the projection matrix and from thence the tuning in each case, we find that the minimax tuning is [{{monzo| 1 0 0 }}, {{monzo| 1 0 1/4 }}, {{monzo| 0 0 1 }}], the projection matrix for 1/4-comma meantone.
+Starting from the same six intervals of the 5-odd-limit diamond and adding 2, we find after computing the normal interval lists that the three subgroups are [2,&nbsp;3], [2,&nbsp;5], and [2,&nbsp;5/3]. Computing the projection matrix and from thence the tuning in each case, we find that the minimax tuning is [{{monzo| 1 0 0 }}, {{monzo| 1 0 1/4 }}, {{monzo| 0 0 1 }}], the projection matrix for 1/4-comma meantone.
 [[Category:Math]]
 [[Category:Regular temperament tuning]]